
==== Front
101573691
39703
Cell Rep
Cell Rep
Cell reports
2211-1247

39159044
10.1016/j.celrep.2024.114649
nihpa2019561
Article
Optogenetic control of kinesin-1, -2, -3 and dynein reveals their specific roles in vesicular transport
Nagpal Sahil 1
Swaminathan Karthikeyan 1
Beaudet Daniel 1
Verdier Maud 12
Wang Samuel 1
Berger Christopher L. 3
Berger Florian 4
Hendricks Adam G. 15*
1 Department of Bioengineering, McGill University, Montreal, QC H3A 0E9, Canada
2 Department of Biomedical Engineering and Health, Episen, Université Paris-Est Créteil, 94010 Créteil Cedex, France
3 Department of Molecular Physiology and Biophysics, University of Vermont, Burlington, VT 05405-0075, USA
4 Cell Biology, Neurobiology, and Biophysics, Department of Biology, Utrecht University, Utrecht, the Netherlands
5 Lead contact
AUTHOR CONTRIBUTIONS

The project was conceptualized by A.G.H. and S.N. S.N. and A.G.H. designed the experiments. S.N., M.V., and S.W. cloned the inhibitors. S.N. performed all the live cell experiments and analysis. C.L.B. provided the kinesin-2 motor for the motility assay. S.N. and D.B. performed the motility assays. K.S. wrote the code for velocity analysis. S.N., F.B., and A.G.H. developed the mathematical model. S.N. and A.G.H. wrote the manuscript. A.G.H. obtained funding for this project.

* Correspondence: adam.hendricks@mcgill.ca
15 9 2024
27 8 2024
18 8 2024
22 9 2024
43 8 114649114649
https://creativecommons.org/licenses/by-nc-nd/4.0/ This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0/).
SUMMARY

Each cargo in a cell employs a unique set of motor proteins for its transport. To dissect the roles of each type of motor, we developed optogenetic inhibitors of endogenous kinesin-1, -2, -3 and dynein motors and examined their effect on the transport of early endosomes, late endosomes, and lysosomes. While kinesin-1, -3, and dynein transport vesicles at all stages of endocytosis, kinesin-2 primarily drives late endosomes and lysosomes. Transient optogenetic inhibition of kinesin-1 or dynein causes both early and late endosomes to move more processively by relieving competition with opposing motors. Kinesin-2 and -3 support long-range transport, and optogenetic inhibition reduces the distances that their cargoes move. These results suggest that the directionality of transport is controlled through regulating kinesin-1 and dynein activity. On vesicles transported by several kinesin and dynein motors, modulating the activity of a single type of motor on the cargo is sufficient to direct motility.

In brief

Leveraging the native autoinhibitory domains of different transport motors, Nagpal et al. developed their optogenetic inhibitors to investigate the effects of transient inhibition on the motility of endocytic vesicles. Their results suggest that, while kinesin-1 and dynein control the direction of the cargo, kinesin-2 and -3 aid in long-range transport.

Graphical Abstract
==== Body
pmcINTRODUCTION

Kinesin and dynein drive the long-range transport of vesicular cargoes, mRNA, and proteins along microtubules. Their dysfunction often leads to neurodegenerative disease.1,2 The activation and recruitment of motors are precisely coordinated to control the intracellular transport.3 Multiple regulatory mechanisms direct cargoes to their destinations in the cell, including scaffolding and adaptor proteins, microtubule-associated proteins (MAPs), and mechanical interactions between different motors on the same cargo. In this study, we seek to mimic the endogenous mechanisms of regulation to better understand how perturbations in motor activity control the direction and duration of cargo transport.

Conventional approaches to inhibit motor proteins, such as function-blocking antibodies,4 genetic inhibition,5 chemical-genetic inhibition,6 and dominant-negative expression,7 are irreversible or lack temporal control. Long-term inhibition of either kinesins or dynein often stops motility in both directions.4,5,8 Optogenetics offers the potential to overcome these limitations and provide spatiotemporal control over motor activity (reviewed in reference 9). Existing optogenetic applications to control motor activity have focused on recruiting exogenous motors to vesicular cargoes.10–12 While these systems provide insight into how the localization of organelles is affected by recruiting motors of different types to their surface, they are limited in their ability to probe the dynamics of the native sets of motors bound to intracellular cargoes.

Here, we developed optogenetic tools to reversibly inactivate endogenous kinesin-1, -2, -3, and cytoplasmic dynein. We used the autoinhibitory domains of kinesins and dynein to develop optogenetic inhibitors based on the LOVTRAP system.13 LOVTRAP employs a light-sensitive LOV2 protein and a Zdark (or Zdk) protein that selectively binds to the dark-adapted conformational state of LOV2. We fused inhibitory domains for each motor to Zdk. LOV2 is anchored to the mitochondria membrane. In the absence of blue light (dark state), the inhibitory domain is sequestered on mitochondria. Upon blue light excitation, Zdk is freed from the mitochondria due to conformational changes in LOV2, allowing the attached inhibitory domain to diffuse in the cytoplasm and interact with endogenous motor proteins (Figures 1A, 1B, 1D, and 1E). This LOV2-based trapping of inhibitory domains provides reversible control of motor activity with diffusion-limited activation kinetics. We used these optogenetic tools to dynamically control the localization of peptides that selectively inhibit kinesins 1–3 and dynein. We examined the effect of transiently inhibiting these motors on the motility of early endosomes, late endosomes, and lysosomes to understand how teams of different motors work together to transport cargoes along microtubules. We find that optogenetic inhibition of kinesin-1 and dynein enhances transport by relieving competition between opposing motors, while the inhibition of kinesin-2 and -3 reduces motility. Together, these results suggest that kinesin-1 and dynein control the direction of movement, while kinesin-2 and -3 aid in long-range transport.

RESULTS

Optogenetic inhibitors of kinesin and dynein motors

Motor proteins adopt autoinhibitory states to prevent unnecessary ATP hydrolysis and congestion of microtubule tracks and remain inactive when not bound to a cargo. Autoinhibition is often achieved by intramolecular interactions between different domains of the motor protein.14–17 We have exploited these interactions to design optogenetic tools that inhibit endogenous motor proteins in living cells upon illumination with blue light (Figure 1).

When no cargo is bound to kinesin-1, it adopts a compact, autoinhibited conformation through a series of intramolecular interactions.16–18 In this autoinhibited state, the C-terminal globular tail domain is positioned near the enzymatically active heads.17–20 The highly conserved QIAKPIRP amino acid sequence in the tail domain interacts with the Switch I region of the head domain and inhibits the initial microtubule-stimulated ADP release step of the motor.19 While interactions among several domains mediate kinesin-1 autoinhibition,16,17 peptides with a central QIAKPIRP sequence are sufficient to inhibit kinesin-1 motility in vitro.19,21 For the kinesin-1 optogenetic inhibitor (K1OI), we cloned the tail domain (amino acids 823–944) from the KIF5B motor onto the Zdk protein of the LOVTRAP system. However, the resulting protein bound strongly with microtubules due to the presence of basic residues N-terminal to the critical QIAKPIRP sequence,19 preventing its association with LOV2. Deletion of these basic residues (amino acids 904–918) reduced the affinity of K1OI to microtubules such that its localization could be controlled using LOVTRAP while retaining its ability to inhibit kinesin-1.

A similar autoinhibited state is observed in kinesin-2 (KIF3A/B) in the absence of cargo.22 However, in kinesin-2, the tail domain blocks the interaction of the motor with microtubules, whereas the coiled-coil segment inhibits processive motility of the motor domain.14,23,24 We used the minimal inhibitory domain, amino acids 601–702 of KIF3A, to construct the kinesin-2 optogenetic inhibitor (K2OI). Although previous studies indicate that overexpression of the KIF3A tail acts as a dominant negative to the heterotrimeric KIF3A/KIF3B/KAP complex, the mechanism of autoinhibition has not been characterized in detail.

In contrast to kinesin-1 and -2, autoinhibition in kinesin-3 motors is mediated by interactions between the coiled-coil 1 region and the neck coiled coil, which prevents the formation of dimers required for processive motility.25,26 To construct the kinesin-3 optogenetic inhibitor (K3OI), we used the coiled-coil 1 domain, along with the β finger domain of KIF1A (amino acids 391–486), both of which are essential for the inhibition of motor activity.27 The amino acid sequences of the inhibitory domains are homologous for KIF1A, KIF1B, and KIF1C (Figure S1A), suggesting that K3OI should interact with KIF1A, -B, and -C isoforms. Live-cell experiments with KIF1A showed a decrease in strong localization of the motor along the cell periphery upon overexpression of the KIF1A inhibitor (Figure S1B). Additionally, we developed optogenetic inhibitors for KIF1B and KIF16B motors (Figure S1C). We used a similar design for the KIF1B inhibitor, whereas for the KIF16B motor, we used the CC2 and CC3 domains, based on a previous study that reported their role in KIF16B autoinhibition.28

Cytoplasmic dynein 1 exists in an autoinhibited phi state in the absence of its binding partners. While recent studies have shed light on the mechanism of autoinhibition,15,29 the characterization of the intramolecular interactions and binding partners mediating autoinhibition remains incomplete. Thus, we adopted a different approach to optogenetically inhibit dynein motors. Instead of relying on a domain that is involved in an intramolecular interaction-based inhibition, we used the residues of p150 that are required for the assembly of the dynein-dynactin complex. Overexpression of the CC1 domain of p150 competes with endogenous p150 for binding with the intermediate chain, thereby disrupting functional dynein-dynactin complexes.30,31 For the dynein optogenetic inhibitor (DOI), we cloned the CC1 domain of p150 onto the Zdk component of the optogenetic system, mimicking its overexpression during blue light illumination.

Optogenetic kinesin inhibitors specifically target their respective motors

To determine the specificity of the optogenetic inhibitors we designed, we tested the effect of inhibitor-containing cell extracts on the in vitro motility of kinesin-1 (KIF5C), −2 (KIF3A), and −3 (KIF1A) motors (Figure 2). We observed a 5-fold decrease in run length and a 10-fold decrease in landing rate of KIF5C in the presence of K1OI compared to control cell lysate (Figure 2A). The inhibitory domains of KIF5A, -B, and -C are highly similar, suggesting the KIF5B inhibitor likely inhibits all three kinesin-1 isoforms (Figure S2). However, lysates containing kinesin-2 or -3 inhibitors did not alter kinesin-1 motility. Similarly, for KIF3A, we observed a 3-fold decrease in run length and landing rate of the motor upon addition of the respective inhibitor, with no significant change in cases of other kinesin inhibitors (Figure 2B).

For KIF1A, we observed a decrease in run lengths from (5.33 ± 0.22) μm for the control to (3.00 ± 0.30) and (3.07 ± 0.43) μm for KIF1A and KIF1B inhibitors, respectively (Figure 2C). We also observed a drastic decrease in landing rates from (4.33 ± 0.17) events μm−1 min−1 for the control to (0.44 ± 0.13) and (0.77 ± 0.23) events μm−1 min−1 for KIF1A and KIF1B inhibitors, respectively. The autoinhibitory domains of KIF1A, -B, and -C share 85%–88% sequence homology (Figure S2), which is in line with our observations that both KIF1A and KIF1B inhibitors block KIF1A motility. Co-expression of the KIF1A inhibitor reduces peripheral punctae of KIF1A motors in living cells (Figure S1B). In contrast, the autoinhibitory domain of KIF16B does not have strong sequence similarity with KIF1A (Figure S2) and only mildly affects KIF1A motility (Figure 2B). The inhibitory domains of kinesin-1 (KIF5B) and kinesin-2 (KIF3A) had no strong effects on KIF1A motility (Figures 2A–2D). These results suggest that KIF1 inhibitors likely inhibit KIF1A, -B, and -C, while there is less cross-reactivity among KIF1s and KIF16B. Furthermore, our results support the idea that CC1 domain along with the β finger domain of KIF1 motor mediates its autoinhibition.27

Taken together, in vitro motility assays and sequence analysis show that the autoinhibitory domains act as specific inhibitors to target kinesins -1, -2, -3, and suggests that this strategy could be extended to other kinesin family members.

Optogenetic inhibition of kinesin-1 and dynein enhances early endosome motility by relieving competition between opposing motors

Signaling molecules, macromolecules, and particles are transported in endosomes into and out of cells. Early endosomes are formed upon internalization of material from the cell’s environment, and are characterized by the presence of specific membrane markers such as EEA1 and Rab5.32 Early endosomes move in short, fast runs interspersed with frequent pauses.33 Kinesin-1,34,35 kinesin-3,36–38 kinesin-14,34 and dynein 8,35,38 motors have been reported to contribute to early endosome motility.

To understand the change in motion of cargoes upon inhibition, we characterized their motility over time using three parameters: the radius of gyration (Rg) is an indicator of the displacement, mean squared displacement (MSD) indicates processivity, and the running-mean velocity indicates the direction of movement (see STAR Methods). Positive velocities correspond to motility toward the cell periphery, while negative velocities correspond to inward movement.

We first examined cargo motility upon expressing the inhibitory peptide alone, which served as a non-optogenetic control to study the dominant-negative effect of the inhibitors. We observed a significant decrease in the Rg and processivity of early endosome in cells expressing the inhibitors in comparison to non-transfected cells, with the most profound drop in the case of dynein and kinesin-1 inhibition (Figures S3B and S3G). Reduced processivity can result from both a reduction in motor-driven processive movements and constrained movement due to an increase in bi-directional motility with equal frequency of switches in both directions. Concomitantly, there was an increase in stationary motility upon kinesin-1 inhibition, and more outward motility in the case of dynein inhibition (Figure S3H), with a shift from peripheral localization to juxtanuclear localization of early endosomes in all the inhibitory conditions (Figure S3F).

Short-term, optogenetic inhibition resulted in varied responses depending on the motor type, in contrast to the general decrease in motility observed upon long-term expression of the inhibitory peptides. When we inhibited kinesin-1 and dynein motors, we saw an increase in mean displacement (measured using Rg) and processivity (quantified using the MSD) (Figures 3E and 3F). We grouped the endosomes as moving outward (velocity ≥0.01 μm/s), inward (velocity ≤0.01 μm/s), or stationary (0.01 μm/s < velocity < −0.01 μm/s) during the 20-s period before the inhibition and analyzed the mean velocity at each time point of the cargoes in every category that we could track over the entire time-lapse video (see STAR Methods). When kinesin-1 is inhibited, the average velocity of cargoes moving toward the cell periphery becomes more variable, suggesting that kinesin-1 inhibition increases the competition between opposing motors. This induced tug-of-war resulting in frequent and large directional switches could be responsible for enhanced processivity. Dynein inhibition leads to a directional switch where minus-end directed cargoes reverse and move to the cell periphery (Figure 3D). This change in the direction of movement suggests that inhibiting kinesin-1 and dynein motors relieves the competition between opposing motors.

Kinesin-2 is not a primary motor on early endosomes.37,39 Accordingly, early endosomes were not strongly affected by kinesin-2 optogenetic inhibition (Figure 3E), consistent with long-term overexpression of the inhibitory peptide (Figure S3). In contrast, inhibiting kinesin-3 motors resulted in decreased displacement and processivity (Figures 3E and 3F) and a shift toward inward movement (Figure 3D), in agreement with previous studies identifying kinesin-3 as a contributor to the motility of early endosomes.36,37 Interestingly, our results suggest that KIF1A and KIF1B inhibition alters the motility of early endosomes and lysosomes, whereas KIF16B inhibition affects only the early endosomes and not lysosomes (Figures S1D and S1E).

Late endosome motility is enhanced by optogenetic inhibition of kinesin-1 and dynein and disrupted by inhibition of kinesin-2 and -3

As endosomes progress in the endocytic pathway, they undergo morphological and biochemical changes as they mature into late endosomes (the process is reviewed in detail in reference 40). Maturation also involves luminal acidification (with pH dropping from around 6.5 to 5.541 and movement toward the perinuclear region with more frequent directional switches.42 This transition is accompanied by a switch from Rab5 to Rab7, a dynamic process initiated by the recruitment of Rab7-GEF by Rab5-GTP and subsequent release of Rab5-GEF, preventing further Rab5 activation.43 Rab7 recruits other adaptor proteins such as RILP and FYCO, which enable interaction with dynein and kinesin motors, respectively, thus allowing late endosomes to move bidirectionally.44,45

Upon the overexpression of motor inhibitors, we again observed the most significant change in mean Rg and MSD in the case of kinesin-1 and dynein inhibition. However, the inhibition of kinesin-2 showed a more pronounced effect on the motility of late endosomes compared to early endosomes (Figures S3B and S3G). Apart from dampened motility, we also observed clustering of vesicles, which is in concert with a previous study that showed clustering of late endosomes upon kinesin-2 inhibition (Figure S4C).39 For all inhibitory conditions, we observed an increase in stationary runs and a slight shift from perinuclear enrichment of vesicles to increased juxtanuclear and peripheral enrichment (Figures S3F and S3H). Therefore, sustained inhibition of motors impairs the transport kinetics of these vesicles, steering them to improper locations within the cell.

Rab7+ vesicles responded similarly to Rab5+ vesicles to optogenetic motor inhibition. The Rg of late endosomes increased upon the transient inhibition of kinesin-1 and dynein motors, indicative of relieving competition between opposing motors (Figure 4E). The effect on processivity is less apparent (Figure 4F) because longer trajectories are weighed more heavily in the MSD analysis, while each trajectory is weighed equally in the Rg calculation, suggesting that kinesin-1 and dynein inhibition has a stronger effect on short trajectories. Upon kinesin-1 inhibition, we observe that the vesicles that were moving toward the plus end switch to dynein-driven minus-ended motility upon blue light illumination. When dynein is inhibited, inward trajectories switch to outward movement, while outward trajectories continue to move outward (Figure 4D). Kinesin-2 and -3 inhibition results in a clear decrease in displacement and processivity, indicating that these motors drive long-range transport of late endosomes.37,39

Lysosome motility is enhanced upon optogenetic inhibition of dynein and suppressed by the inhibition of kinesins -1, -2, and -3

The majority of late endosomes fuse with lysosomes, generating a hybrid organelle called an endolysosome that provides a controlled acidic environment for the degradation of the endocytosed macromolecules. These endolysosomes further mature to form dense lysosomes, which is the final compartment of the endocytic pathway.46 Lysosomes have a characteristically low pH of around 4.5, owing to the presence of degradative acid hydrolases.41,46 Lysosome motility involves kinesin-1, -2, and -3 for anterograde transport and dynein for retrograde transport (reviewed in reference 47).

Overexpression of motor inhibitors showed markedly different effects on lysosome motility compared to Rab5- and Rab7-enriched vesicles. Kinesin-3 and dynein inhibition reduced the mean Rg and MSD with an almost two and three times drop in the fraction of high Rg lysosomes, respectively (Figures S5B and S5G). Furthermore, there was an increase in the fraction of stationary lysosomes upon kinesin-3 and dynein inhibition. Kinesin-1 and -2 inhibition did not significantly affect the mean Rg or directionality of the cargoes; however, there was a decrease in processivity, as shown by the MSD. Similar to late endosomes, we observe a shift from perinuclear localization of lysosomes to a more peripheral localization for all inhibitors, with the most significant change in the case of kinesin-1 inhibition (Figure S4F).

Lysosomes also responded differently to optogenetic inhibition of motors compared to early and late endosomes (Figures 5 and S5). For lysosomes, kinesin-1 inhibition resulted in reduced motility (Figures 5E and 5F), in contrast to early and late endosomes (Figures 2E, 2F, 3E, and 3F). The cargo velocities provide insight into the effect of the inhibitor, where we observe a relatively mild effect (Figures 5C and 5D), even with prolonged optogenetic inhibition (Figure S6B). These results suggest kinesin-1 motors play a different role in lysosome motility compared to endosomes. Kinesin-2 and -3 inhibition led to a decrease in mean Rg and MSD, with a clear shift toward minus-ended motility in the lit state, with some cell-to-cell variability in the extent of induced minus-ended motility (Figures 5D–5F and S5). For dynein inhibition, we see enhanced motility with a directional switch toward the plus end of the microtubules, although the effect is weaker than that observed for early and late endosomes (Figures 5D–5F and S6). Interestingly, lysosomes that were not motile in the dark state started moving toward the cell periphery upon dynein inhibition (Figure 5D).

Mathematical modeling reveals the estimated fraction of inhibited motors

To gain more insight into how the mechanochemical properties of different motors determine their roles in vesicle transport, we tested the effect of inhibiting specific motors in a mathematical model of transport by kinesins and dynein. We extended the mathematical model developed by Müller et al.48 to simulate the motility of a cargo driven by any number of different types of motors (see STAR Methods). We estimated motor parameters such as binding rate, detachment rate, detachment force, stall force, and forward and backward velocities based on previous single-molecule experiments and published models, following closely the analysis in Allison et al.49 (see STAR Methods). To simulate different vesicle populations, we used estimates of the number of motors on each of these cargoes from previous single-molecule fluorescence and quantitative photobleaching experiments.50 Using these parameters, the model captured the motility characteristics of early endosomes, late endosomes, and lysosomes, and generated different results for them according to the number of estimated motors used to define each cargo type. We then mimicked the effect of optogenetic inhibition of the motors in the model by decreasing the binding rates of the motors.

Simulated trajectories show similar changes in displacement and directionality in response to motor inhibitions as observed experimentally for most of the cases (Figure 6A). We note that the distribution of modeled trajectories in DOI is influenced by outliers in the positive range that are not plotted in the figure, which made the mean of data shift upward. We used the model to estimate the degree to which optogenetic inhibitors decreased the activity of different motors in our experiments. We estimate that the optogenetic inhibitors developed here reduced the binding rate by ~40%–60% for kinesin-1, 30%–40% for kinesin-2, 30%–50% for kinesin-3, and 25%–50% for dynein, depending on the cargo (Figure 6A).

We also examined the change in directionality of simulated trajectories upon the inhibition of different motors in Figure 6B, where closed shapes represent the experimental data, and open symbols represent the modeling results for percentage change in outward, inward, and stationary motility (shown in blue, green, and black, respectively) for each inhibitory condition. Overall, we observed a similar trend in change of directionality as our experimental data. However, in some cases, the modeled change in directionality was far greater than the experimental data. For instance, in the case of lysosomes, the mean increase in percentage of inward motility for kinesin-1 inhibition was 5.7% for the experimental data, whereas the estimated increase by the model was 14%. Similarly, for dynein inhibition, the mean increase in percentage of outward motility was 8.4% for the experimental data, whereas the estimated increase by the model was more than 20%. This indicates that there could be some missing parameters in our model for cargo transport. It is also important to note that in vitro results do not always align with in vivo observations, as there are multiple additional factors that are difficult to quantify, such as possible roles of microtubule posttranslational modifications (PTMs) and cargo adaptors, and are thus not accurately captured by the in vitro reconstitution and modeling.

DISCUSSION

Cargo transport is regulated through multiple mechanisms, including the cytoskeletal tracks, motor adaptors and scaffolding proteins, and mechanical interactions between motors on the same cargo. To examine the specific roles of kinesin-1, -2, -3, and dynein on organelle motility, we developed optogenetic inhibitors that target endogenous kinesins and dynein.

Using optogenetics to control kinesin and dynein activity, we addressed the following questions:

What motors contribute to the motility of different endocytic cargoes?

While the types of motors associated with different vesicle populations have been characterized using genetic approaches and live cell imaging,35,37,50 determining which motors are active has been challenging as long-term inhibition via dominant-negative expression or genetic manipulation often results in reduced transport in both directions.4,5,8 By comparing transport in the same cell in control and under acute inhibition of specific motors, we can map which motors contribute to the motility of early endosomes, late endosomes, and lysosomes (Figure 7A). We find that early endosomes are transported by kinesin-1, kinesin-3, and dynein. Kinesin-1, -2, -3, and dynein contribute to late endosome and lysosome motility (Figure 7A).

Why are multiple types of motors bound to the same cargo? Kinesins -1, -2, -3 are processive and drive transport toward microtubule plus ends. However, many intracellular cargoes are bound by multiple types of kinesins and dynein.37,50–52 What advantage is provided by multiple kinesin types acting on the same cargo? Our results suggest each motor type has a distinct role in cargo transport. Optogenetic inhibition of kinesin-1 and dynein results in longer trajectories and a switch in the direction of movement (Figures 6 and 7). Thus, kinesin-1 and dynein play regulatory roles to determine the direction of motility. In contrast, optogenetic inhibition of kinesin-2 and -3 results in reduced displacement (Figure 6A) and more stationary trajectories (Figure 6B), suggesting that kinesin-2 and -3 act as long-range haulers. Optogenetically inhibiting multiple kinesins at the same time showed a more prominent directional switch, with the majority of cargoes shifting to dynein-driven motility (Figure S7), indicating that multiple types of kinesins work together to transport cargoes.53 Interestingly, MAPs like tau and MAP7 preferentially target kinesin-1.51,54 Kinesin-1 and -3 prefer microtubules marked with different sets of PTMs.55,56 Furthermore, kinesin-1 requires multiple interactions to fully relieve autoinhibition.16,17 Together, these results suggest that many mechanisms that regulate intracellular transport might specifically target kinesin-1, where kinesin-2 and -3 could aid in long-range transport (Figure 7B).

We found that sustained and acute inhibitions had remarkably different effects on cargo motility. While the overexpression of inhibitors often led to a halt in motility with more stationary runs, optogenetic inhibition upregulated cargo transport in some cases of motor inhibition. This indirect effect of acute inhibition suggests close coupling between different motors that are on a cargo, whereby motors can directly activate and influence one another. This crosstalk between opposite polarity motors has been previously established in other studies, where it was shown that the slowest teams of motors mechanically communicate with other motors through the membrane of the cargo to drive transport.57,58 Based on our results, we hypothesize that selectively inhibiting kinesin-1 and dynein motors in the case of early and late endosomes relieves competition with opposing motors. Turning off kinesin-1 or dynein might enable more processive motors, such as kinesin-3, to take over transport. This hypothesis is in line with computational and in vitro studies that suggest that the number of engaged motors governs motility.59–61

To summarize, we developed optogenetic inhibitors of kinesin-1, -2, -3, and dynein motors to mimic regulation by motor adaptors and effectors in the cell. We examined the change in motility of different endocytic vesicle populations upon acute, optogenetic inhibition. Acute inhibition has different effects on motility compared to long-term inhibition using dominant negatives. Short-term inhibition with optogenetics mimicked how these motors might be regulated during cargo transport, where kinesin-1 and dynein inhibition leads to enhanced motility, owing to strong directional switches to the opposite direction, whereas kinesin-2 and -3 inhibition results in reduced motility. We propose that the activity of kinesin-1 and dynein motors determines the net direction of movement, while kinesin-2 and -3 aid in long-range transport. In conclusion, different vesicles of the endocytic pathway rely on unique sets of transport motors, where modulating the activity of a single type of motor dictates the overall motility of the cargo, indicating an intricate underlying interplay of motors to drive the vesicle to its correct destination.

Limitations of the study

The optogenetic tools developed in this study were utilized to study the transport of vesicles in the endocytic pathway with precise temporal control over the inhibition of motor proteins; however, we did not explore the potential of spatial control over the inhibitors in this study. It would be difficult to spatially restrict the inhibitors to a specific region of the cell—for example, to study its effect on motors enriched only at the axons or dendrites of a neuron. With the current design, inhibitors are released into the entire cytoplasm upon blue light illumination, and it would require additional changes in the design to restrict the release of the inhibitors to a particular region of the cell. This would expand the utility of these inhibitors, for instance, to study the mechanics of motor interactions on different microtubule subpopulations in a cell.55 Furthermore, the model developed in this study does not accurately capture the complex process of cargo transport. Cargo transport involves regulation at multiple levels, such as at the cell-signaling level and at the molecular level by MAPs, adaptor, and scaffolding proteins, PTMs of the cytoskeleton and motor proteins, and so on. This would entail coding multiple parameters that are further dependent on various other parameters, such as the stage of cell cycle, cell health, and tissue type, to name a few. The model developed in this study is a simple case using some basic parameters that are well characterized from previous studies. Finally, optogenetic inhibition of motors requires the transfection of cells with the plasmids developed in this study, which makes it difficult to control the expression levels of the inhibitors. This could be overcome by generating stable cell lines that have steady expression levels of the inhibitors.

STAR★METHODS

RESOURCE AVAILABILITY

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Adam Hendricks (adam.hendricks@mcgill.ca).

Materials availability

Plasmids generated in this study will be shared by the lead contact upon request.

Data and code availability

All data reported in this paper will be shared by the lead contact upon request.

All original code has been deposited at GitHub and Zenodo (DOI: https://doi.org/10.5281/zenodo.13146134)

Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

EXPERIMENTAL MODEL AND SUBJECT DETAILS

Plasmid preparation

The optogenetic module was derived from plasmids pTriEx-NTOM20-LOV2 and pTriEx-mCherry-Zdk1 (Addgene plasmid #81009 and #81057 respectively), gifts from Klaus Hahn. Markers for endosomes were derived from plasmids iRFP-FRBRab5 and iRFP-FRB-Rab7 (Addgene plasmid #51612 and #51613 respectively), gifts from Tamas Balla. Dominant negative constructs for KIF5B, KIF3A and dynein were kindly gifted by Erika Holzbaur. Dominant negative constructs for kinesin-3 were constructed from plasmids pBa-FRB-3myc-KIF1A tail 391–1698 (Addgene plasmid #64286), pBa-FRB-3myc-KIF1B beta tail 387–1770 (Addgene #64287) and pBa.GFP-KIF16B tail 458–1317 (Addgene #134624), gifts from from Gary Banker and Marvin Bentley. KIF1A(1–393)-LZ-3xmCitrine plasmid was a gift from Kristen Verhey, University of Michigan, Ann Arbor, MI.

A peptide linker - GSGGSGSGGT - was added to the inhibitory peptide constructs before fusing them with mCherry-Zdk1. Optogenetics constructs were created with the overlap extension PCR cloning method69 and the primers were designed on an online platform for restriction-free cloning.64 Briefly, an amplicon containing the fragment of interest from the donor plasmid and flanked by a nucleotide sequence that overlaps with the insertion site of the recipient plasmid was generated from the primary PCR step. This amplicon was gel extracted and was used as an oversized primer for the secondary PCR reaction. The recombinant plasmid obtained at the end of secondary PCR reaction was incubated with Dpn1 enzyme, followed by its transformation into bacteria. Finally, colony PCR was done to screen the colonies and positive clones were confirmed by Sanger sequencing.

Cell culture

U-2 OS and COS-7 cells were obtained from American Type Culture Collection, Manassas, VA. All cell lines were routinely screened for mycoplasma contamination. Cells were grown on 100mm polystyrene cell culture dishes (Gibco, Thermo Fisher Scientific, Waltham, MA) for maintenance and on 35mm glass-bottom dishes - No. 1.0 coverslip (MatTek Corporation, Ashland, MA) for imaging, with DMEM basal media (Gibco), supplemented with 10% (v/v) FBS (Gibco), 1% (v/v) Glutamax (Gibco) and Penstrep(Gibco). Cells were passaged using 0.25% trypsin-EDTA (Gibco) and PBS (Wisent, St-Jean-Baptiste, QC, Canada) and incubated at 37°C with 5% CO2.

METHOD DETAILS

Transfection and imaging

Cells were transiently transfected with ~400ng total plasmid DNA prepared in OPTI-MEM Reduced Serum (Gibco), using Lipofectamine LTX with Plus-Reagent (Invitrogen, Thermo Fisher Scientific), according to the manufacturer’s instructions. Cells were incubated with transfection solution for 4 h before replacement with fresh DMEM media. For staining lysosomes, cells were treated with 50nM Lysotracker Deep Red (Invitrogen) for 10 min in complete DMEM media. Live-cell imaging was performed on a TIRF set-up built on an Eclipse Ti-E inverted microscope (Nikon, Melville, NY) with an attached EMCCD camera (iXon U897, Andor Technology, South Windsor, CT), 1.49 numerical aperture oil-immersion 100x objective and 100 mW diode lasers (Coherent). Time-lapse movies were taken using 561nm laser (1 mW) to excite the mCherry fluorophore, 647nm laser (1 mW) to excite the endosome marker, and 488nm laser (1 mW) for blue-light illumination, along with appropriate emission filter sets.

Cells were imaged at 37°C in Leibovitz’s L-15 media (Gibco), supplemented with 10% (v/v) FBS. For overexpression studies, cells were co-transfected overnight with dominant-negative constructs at a concentration of 200ng of DNA per dish and endosome marker constructs at 50ng of DNA per dish. Cargoes were imaged for 120s, with exposure set to 300ms per frame.

For optogenetic studies, cells were co-transfected overnight with pTriEx-NTOM20-LOV2 plasmid at a concentration of 300ng of DNA per dish, inhibitor construct at 50ng of DNA per dish (with the exception of K1OI, which was at 15ng of DNA per dish) and endosome marker constructs at 50ng of DNA per dish. Optogenetic cells were first identified based on the mitochondrial localization of the mCherry fluorophore in the absence of blue light. After taking a 45 s movie to document the presence of optogenetic constructs in the cell, with an exposure of 800ms per frame and a short 5s blue light pulse, a 255s movie tracking the cargo was taken in the same cell. Cargoes were imaged for 120s in the dark-state followed by 135s period in the lit-state (out of which, the initial 15s period was assumed to be the inactivation period, where the inhibitors diffuse around the cytoplasm and interact with motors to deactivate them, and was not considered for MSD & Rg analysis), with an exposure set to 300ms per frame.

Tracking of cargoes

Endosomes were tracked with the ImajeJ plugin TrackMate,70 with the Laplacian of Gaussian filter, and sub-pixel localization for cargo detection, and the simple Linear Assignment Problem tracker that generates the trajectories. Based on the known differences in maximal velocity for different endosomes, tracking parameters were set to linking distance (μm):gap-closing maximum distance (μm):gap-closing maximum frame gap of 1:1:1 for Rab5 and Rab7 vesicles and 1.5:1.5:1 for lysosomes. However, changing these parameters did not significantly affect the analysis (Figure S8).

Using Trackmate output, X and Y coordinates, and time duration of each cargo trajectory in a cell were obtained. 2D position of individual trajectories was calculated by taking the square root of the sum of the squares of the X and Y coordinates of the trajectory with respect to the origin of the trajectory. This position data of trajectories was then used for further analysis.

Mathematical modeling

Motor parameters from Gickling et al.49 generated cargo trajectories similar to the observed trajectories in our data (Table S1), with the exception of the binding rates which were reduced 100-fold to match the frequency of directional switches in our experiments. To simulate different endosome populations, we changed the number of attached motors. For early endosomes, we decreased the number of attached dynein motors based on our reported data.50 Moreover, we omitted kinesin-2 in our model to recapitulate motors reported to be present on early endosomes.39,71

Microtubule preparation

Bovine tubulin isolation and fluorescently labeled microtubule polymerization were conducted as previously described.72

Microtubule solution was prepared by mixing 25% Alexa 647 labeled tubulin and 75% unlabeled tubulin in BRB80 (80mM K-PIPES, 1mM MgCl2, 1mM EGTA, pH 6.8) to a final concentration of 5 mg/mL, supplemented with 1 mM GTP (Sigma Aldrich). Microtubules were then polymerized at 37°C for 30 min, and later stabilized with 20 μM Taxol (Cytoskeleton) in DMSO, and incubated for an additional 25 min at 37°C. Microtubules were cleared 2X by pelleting them at 10,600g for 5 min at RT, then resuspending with T-BRB80 (BRB80 supplemented with 20 μM Taxol).

Cell lysate protein expression

KIF1A(1–393)-LZ-3xmCitrine motors (a gift from Kristen Verhey, University of Michigan, Ann Arbor, MI) were expressed in U-2 OS cells for cell lysate protein expression, as previously described.72 Cells were washed with PBS and collected using a cell scrapper in motility assay buffer (MAB; 10 mM PIPES, 5 mM K-Acetate, 4 mM MgCl2, and 1mM EGTA, pH 7.0.), supplemented with protease inhibitor cocktail (PIC; BioShop), 10 mM dithiothreitol (DTT) and 1 mM MgATP. Cell were then transferred to a tight-fitting Dounce cell homogenizer and homogenized on ice by hand with 20–30 up-and-down strokes to ensure adequate disruption of cell membranes. Following, the cell homogenate was spun at 650g for 10 min at 4°C to remove cell debris, followed by another centrifugation step of the supernatant at 10,600g for 10 min at 4°C to remove large macromolecular aggregates. Final supernatant containing motors was then aliquoted into eppendorf tubes, flash frozen, and stored at −80°C. To control for motor concentration, all experiments were done from the same batch of motors. Similarly, cell extracts from different inhibitors were extracted and stored in −80°C.

In-vitro motility assay

KIF1A in vitro motility assays were conducted as previously described.72 Briefly, flow chambers were made by mounting silanized coversplips on glass slides using vacuum grease and double-sided tape. After incubating the flow chamber with 2:50 diluted anti-β-tubulin (T4026 clone TUB2.1, Sigma) in BRB80 for 5 min, chambers were washed with BRB80, and then flowed with F-127 (Sigma) for 10 min to block any non-specific binding. This was followed by 2X washes with T-BRB80. Diluted microtubules (2:50 in T-BRB80) were then flowed through the chamber and incubated for 5 min at RT, with subsequent 2X T-BRB80 washes to remove unbound microtubules. Cell lysate supplemented with 0.2 mg/mL BSA, 10 mM DTT, 1 mM MgATP, 20 μM taxol, 1 mg/mL casein, and an oxygen scavenger system (15 mg/mL glucose, ≥2000 units/g glucose oxidase, ≥6 units/g catalase). For control experiments, 30 μL of cell lysates with motors was diluted in 20 μL motility assay buffer, whereas for inhibitory experiments, 30 μL of cell lysates with motors was mixed with 20 μL of cell lysates with respective inhibitors. Using TIRF, microtubules were imaged for 1 frame with the 640 nm laser at 1–10% power, and time-lapse movies of motors were taken with 200 msec exposure using 488 nm laser set at 2% power for 5 min. Kymographs were generated using Multi Kymograph plugin in FIJI, which were then analyzed with KymoButler.

Purified recombinant rat kinesin-1 430 GFP protein (a gift from Dr. Gary Brouhard, Department of Biology, McGill University) and recombinant biotin-tagged kinesin-2 (KIF3A/A) were expressed in bacteria and purified. Kinesin-1 and kinesin-2 motors were incubated in the presence of 5 mM MgATP and taxol-stabilized microtubules in T-BRB80 buffer for 5 min following centrifugation at 16,000g for 5 min. Motors that were not able to release microtubules in saturating ATP were removed. Aliquots of purified motors were flash frozen in liquid nitrogen and stored at −80°C.

Kinesin-2 aliquots were thawed and incubated on ice with streptavidin conjugate Quantum dots (Qdot, Invitrogen) for 1hr in the dark at a ratio of 40:1 (v/v), and used immediately. For in vitro motility assays with kinesin-1 and kinesin-2-Qdots, motors were diluted in motility assay buffer that contained 30 μL of mock cell lysate or lysates that contained the inhibitors and supplemented with 0.2 mg/mL BSA, 10 mM DTT, 1 mM MgATP, 20 μM taxol, 1 mg/mL casein, and an oxygen scavenger system (15 mg/mL glucose, ≥2000 units/g glucose oxidase, ≥6 units/g catalase). Motility event recordings were obtained using the same acquisition parameters as those used for KIF1A.

QUANTIFICATION AND STATISTICAL ANALYSIS

Quantification of radius of gyration

Radius of gyration (Rg) is the radius of a circle that can be drawn around the trajectory, in such a way that it encompasses half of the points in the trajectory, thus parameterizing the size of the trajectory. It is indicative of the distance covered by the motor-driven cargo, indirectly estimating the run length of the motor. Radius of gyration is a scalar quantity (thus directionless), calculated using the following formula: Rg=1n∑inxi−X¯2+yi−Y¯2

where n is the total number of detected points in the trajectory at consecutive time frames, xi and yi are the X and Y coordinates of the trajectory point at i time point, and X¯ and Y¯ are the mean position of all the points in the trajectory. In our study, we took the mean of calculated Rg values for all the trajectories in a cell.

Quantification of mean-squared displacement

MSD is the average of the squared (therefore directionless) displacement of the cargo from its starting position over period of time and it is analogous to cargo’s processivity, that is, the distance that the motors drive the cargo before detachment from microtubules. In general, the larger the MSD, the higher the active transport of the cargo and the higher will be the processivity. MSD was calculated using the following formula, where τ is the delay or sliding time, T is the total time, t is the current time point, and x is the position: MSD(τ)=1T−τ∑t=0T−τxt+τ−xt∧2

Quantification of velocity

Velocity is the rate of change in position of a cargo in motion from a frame of reference, and since it is a vector quantity, it has a directional component in it. Velocity of the trajectories provides information on the directional switches that the motor proteins undergo when carrying cargoes in a cell. The trajectories which span less than 400 time frames were discarded. Sliding window algorithm was employed in analyzing the data, which reduces the computational power by breaking a large array into smaller sub-arrays. Using this algorithm, the average velocity of individual trajectories was calculated with a window size of 15 s. The average of all the trajectories was calculated based on the 20 s period before inhibition, which represents the state of motor motility right before inhibition. Average velocity was then categorized into three types: Positive velocity (velocity ≥ 0.01 μm/s), negative velocity (velocity ≤ 0.01 μm/s) and neutral velocity (0.01 μm/s < velocity < −0.01 μm/s).

Statistical analysis

Distribution of data was checked on MATLAB using histograms and quantile-quantile plots. Wilcoxon signed-rank test was used to test for statistical significance between dark and lit states for Rg analysis. Data for differences in radius of gyration for experimental case was tested for difference from zero, using Student’s t-test. Run lengths, velocity, and landing rates were assessed with one-way ANOVA followed by Dunnett’s test for comparing means from multiple groups to a control. A p-value below 0.05 was considered statistically significant. The mean and standard error are described in the main text and/or the figures. The number of values examined, the experimental replicates, and the statistical tests applied are described in the figure legends.

Supplementary Material

1

2

ACKNOWLEDGMENTS

We thank other students in the lab, namely Abdullah Chaudhary, Linda Balabanian, and Ora Cohen, for help with data analysis and thoughtful discussions. We also thank the Genome Quebec Innovation Centre at McGill for sequencing our DNA samples. The work was supported by Canadian Institutes of Health Research (CIHR) grants PJT-159490 and PJT-185997 to A.G.H.

Figure 1. Native autoinhibitory domains guide the design of optogenetic inhibitors

For a Figure360 author presentation of Figure 1, see https://doi.org/10.1016/j.celrep.2024.114649.

(A) Schematic illustration of the mechanism of action for optogenetic inhibitors. In the dark state, the inhibitory peptide is sequestered on mitochondria. Upon blue light illumination, it is released into the cytoplasm, where it interacts with endogenous motor proteins.

(B) Overview of plasmid constructs co-transfected for the optogenetic experiments. Plasmid construct on top encodes the protein that is tethered to the mitochondria, and the bottom plasmid construct encodes the fluorophore containing inhibitory peptide.

(C) Ab initio-based protein-prediction results from the I-TASSER server (top) for different optogenetic inhibitors where the segment in red shows the cloned inhibitory peptide. Domain maps of different proteins used for creating optogenetic inhibitors (bottom), with the cloned segment highlighted in red. CC, coiled-coil; CG, CAP-Gly; FHA, forkhead-associated; MD, motor domain; NC, neck coil; PH, pleckstrin homology; SP, serine-proline-rich region.

(D) Snapshots from time-lapse imaging of a cell expressing NTOM20-LOV2 and mCherry-Zdk1-K2OI constructs. The yellow box shows the cytoplasmic region used for generating the intensity trace shown in (E). Scale bar, 12 μm.

(E) Fluorescence intensity trace over time for the region of interest in cytoplasm for the cell shown in (D), demonstrating the reversible aspect of optogenetic inhibitors.

Figure 2. Kinesin motors are sensitive to their respective inhibitors

Example kymographs from in vitro motility assays on Taxol-stabilized microtubules with motor alone, and in combination with different kinesin inhibitors (left) for KIF5C-eGFP (A), KIF3A-QD (B), and KIF1A-mCitrine (C). Scale bars: vertical 10 s, horizontal 5 μm. Run length and landing rate plots of trajectories from kymographs (right). The number of kymographs used for the analysis are as follows: For kinesin-1 motor (A): control: 32 kymographs; Kin-1 inhibition: 21 kymographs; Kin-2 inhibition: 34 kymographs; KIF1A inhibition: 27 kymographs. For kinesin-2 motor (B): control: 31 kymographs; Kin-1 inhibition: 32 kymographs; Kin-2 inhibition: 34 kymographs; KIF1A inhibition: 31 kymographs. For kinesin-3 motor (C): control: 28 kymographs; Kin-1 inhibition: 30 kymographs; Kin-2 inhibition: 32 kymographs; KIF1A inhibition: 37 kymographs; KIF1B inhibition: 36 kymographs; KIF16B inhibition: 31 kymographs. Black horizontal lines indicate means, and black vertical lines indicate SEMs. Statistical analysis was done using one-way ANOVA, followed by Dunnett’s test, and asterisks indicate significance as follows: *p ≤ 0.05; ****p ≤ 0.0001; n.s., not significant.

Figure 3. Early endosome motility is modulated by optogenetic inactivation of kinesin-1 and -3 and dynein motors

(A) Scheme showing the inhibition of motors that are driving early endosomes in the lit state (left), with the inhibitory peptide labeled with an orange fluorophore, while the early endosome marker, Rab5, is labeled with a far-red fluorophore. On the right is the summary of change in motility upon inhibition of different transport motors, shown by differently colored arrows, where the length of the arrow indicates the run length of the cargo.

(B) The MSD plot of early endosomes in untransfected U2OS cells, which do not express optogenetic inhibitors, without and with blue light illumination, shown in black and blue, respectively (mean ± SEM). Each cell was first imaged without shining any blue light, and then with blue light illumination. This blue light control shows that blue light itself does not affect the motility of early endosomes.

(C) Polar plot projections of early endosomes trajectories from time-lapse images, centered around the cell nucleus, showing the directionality of Rab5-enriched endosomes in a U2OS cell under dark-state (top) and lit-state (bottom) conditions. The four panels correspond to cells that were transiently transfected with different optogenetic inhibitors. The net directionality was categorized as inward (magenta), outward (green), or stationary (gray) based on Rg values, and rho values in the first and the last points of the trajectories.

(D) Plot shows the changes in average velocity for all the trajectories in a cell (corresponding to the cell shown in C) upon blue light illumination. For velocity analysis, average velocity was first categorized into three types, namely, positive velocity, negative velocity, and neutral velocity. It was then normalized to the average velocity in the time window just before inhibition, allowing us to compare changes at the time of inhibition. The color scheme is also based on the average velocity of the trajectories right before the inhibition, where green represents positive average velocity before inhibition, magenta represents negative average velocity before inhibition, and gray represents stationary vesicles that were not moving before inhibition.

(E and F) Rg and MSD plots for motility of early endosomes upon optogenetic inhibition of different motors. Each dot in the Rg plot indicates a cell, with a line connecting the same cell under the two conditions. A yellow line indicates an increase in Rg, whereas a purple line indicates a decrease, and a black line indicates no change. Black horizontal line shows mean while vertical gray line indicates SEM. For the MSD plot, dark state and lit state are shown in gray and blue, respectively (mean ± SEM). The number of cells, trajectories, and experiments used for the plots are as follows: K1OI: 41 cells, 7,713 trajectories over 5 experiments; K2OI: 48 cells, 8,560 trajectories over 3 experiments; K3OI: 47 cells, 8,976 trajectories over 5 experiments; DOI: 26 cells, 4,859 trajectories over 3 experiments. Statistical analysis for Rg experiments was done using the Wilcoxon signed rank test, and asterisks indicate significance as follows: *p ≤ 0.05; **p ≤ 0.01; *** p ≤ 0.001.

Figure 4. Late endosome motility is enhanced by optogenetic inhibition of kinesin-1 and dynein and reduced by optogenetic inhibition of kinesin-2 and -3

(A) Scheme showing the inhibition of motors that are driving late endosomes in the lit state (left), with the inhibitory peptide labeled with an orange fluorophore, while late endosome marker, Rab7, is labeled with a far-red fluorophore. On the right is the summary of change in motility upon inhibition of different transport motors, shown by differently colored arrows, where the length of the arrow indicates the run length of the cargo.

(B) The MSD plot of late endosomes in untransfected U2OS cells, that do not express optogenetic inhibitors, without and with blue light illumination, shown in black and blue, respectively (mean ± SEM). Each cell was first imaged without shining any blue light, and then with blue light illumination. This blue light control shows that blue light itself does not affect the motility of late endosomes.

(C) Polar plot projections of late endosome trajectories from time-lapse images, centered around the cell nucleus, showing the directionality of Rab7-enriched endosomes in a U2OS cell under dark-state (top) and lit-state (bottom) conditions. The four panels correspond to cells that were transiently transfected with different optogenetic inhibitors. The net directionality was categorized as inward (magenta), outward (green), or stationary (gray) based on Rg values, and rho values in the first and the last points of the trajectories.

(D) Plot shows the changes in average velocity for all the trajectories in a cell (corresponding to the cell shown in C) upon blue light illumination. For velocity analysis, average velocity was first categorized into three types, namely, positive velocity, negative velocity, and neutral velocity. It was then normalized to the average velocity in the time window just before inhibition, allowing us to compare changes at the time of inhibition. The color scheme is also based on the average velocity of the trajectories right before the inhibition, where green represents positive average velocity before inhibition, magenta represents negative average velocity before inhibition, and gray represents stationary vesicles that were not moving before inhibition.

(E and F) Rg and MSD plots for the motility of late endosomes upon optogenetic inhibition of different motors. Each dot in the Rg plot indicates a cell, with a line connecting the same cell under the two conditions. A yellow line indicates an increase in Rg, whereas a purple line indicates a decrease, and a black line indicates no change. Black horizontal line shows mean while vertical gray line indicates SEM. For the MSD plot, dark state and lit state are shown in gray and blue, respectively (mean ± SEM). The number of cells, trajectories, and experiments used for the plots are as follows: K1OI: 34 cells, 5,638 trajectories over 4 experiments; K2OI: 37 cells, 6,189 trajectories over 3 experiments; K3OI: 28 cells, 4,473 trajectories over 3 experiments; DOI: 27 cells, 4,512 trajectories over 3 experiments. Statistical analysis for Rg experiments was done using the Wilcoxon signed rank test and asterisks indicate significance as follows: *p ≤ 0.05; **p ≤ 0.01; *** p ≤ 0.001.

Figure 5. Lysosome motility is reduced by acute inhibition of kinesin motors and enhanced upon acute inhibition of dynein

(A) Scheme showing the inhibition of motors that are driving lysosomes in the lit state (left), with the inhibitory peptide labeled with an orange fluorophore, while the lysosome is labeled with far-red LysoTracker. On the right is the summary of change in motility upon inhibition of different transport motors, shown by differently colored arrows, where the length of the arrow indicates the run length of the cargo.

(B) The MSD plot of lysosomes in untransfected U2OS cells, which do not express optogenetic inhibitors, without and with blue light illumination, shown in black and blue, respectively (mean ± SEM). Each cell was first imaged without shining any blue light, and then with blue light illumination. This blue light control shows that blue light itself does not affect the motility of lysosomes.

(C) Polar plot projections of lysosome trajectories from time-lapse images, centered around the cell nucleus, showing their directionality in a U2OS cell under dark-state (top) and lit-state (bottom) conditions. The four panels correspond to cells that were transiently transfected with different optogenetic inhibitors. The net directionality was categorized as inward (magenta), outward (green), or stationary (gray) based on Rg values, and rho values in the first and last points of the trajectories.

(D) Plot shows the changes in average velocity for all the trajectories in a cell (corresponding to the cell shown in C) upon blue light illumination. For velocity analysis, average velocity was first categorized into three types, namely, positive velocity, negative velocity, and neutral velocity. It was then normalized to the average velocity in the time window just before inhibition, allowing us to compare changes at the time of inhibition. The color scheme is also based on the average velocity of the trajectories right before the inhibition, where green represents positive average velocity before inhibition, magenta represents negative average velocity before inhibition, and gray represents stationary vesicles that were not moving before inhibition.

(E and F) Rg and MSD plots for motility of lysosomes upon optogenetic inhibition of different motors. Each dot in the Rg plot indicates a cell, with a line connecting the same cell under the two conditions. A yellow line indicates an increase in Rg, whereas a purple line indicates a decrease, and a black line indicates no change. The black horizontal line shows mean, while vertical gray line indicates SEM. For the MSD plot, dark state and lit state are shown in gray and blue, respectively (mean ± SEM). The number of cells, trajectories, and experiments used for the plots are as follows: K1OI: 34 cells, 3,791 trajectories over 4 experiments; K2OI: 34 cells, 3,843 trajectories over 4 experiments; K3OI: 34 cells, 3,798 trajectories over 5 experiments; DOI: 27 cells, 3,114 trajectories over 3 experiments. Statistical analysis for Rg experiments was done using the Wilcoxon signed rank test, and asterisks indicate significance as follows: *p ≤ 0.05; **p ≤ 0.01; *** p ≤ 0.001.

Figure 6. Mathematical modeling indicating the unique mechanochemical properties of each motor determines its role in transport

(A) Plots for difference in Rg upon optogenetic inhibition of different motors for early endosomes, late endosomes, and lysosomes. The solid horizontal line indicates mean, and the dashed horizontal line indicates median. Top summarizes the Rg results reported in figures above and bottom indicates the normalized Rg values obtained from the modeled trajectories under different inhibitory conditions. The simulated trajectories exhibit a broader distribution as results for each simulated trajectory are plotted, compared to the mean of the trajectories for each cell in the experimental data. Binding rates are estimated to be reduced by ~40%–60% for K1OI, 30%–40% for K2OI, 30%–50% for K3OI, and 25%–50% for DOI depending on the cargo. p values were calculated from Student’s t test. (B) Plots indicating changes in directionality upon inhibition of different motors for early endosomes, late endosomes, and lysosomes. Trajectories were categorized into outward, inward, or stationary based on the difference in positions for end and start points of the trajectory. Filled circles, triangles, and squares indicate experimental data for three different cells, and empty circles indicate the modeling results.

Figure 7. Different endocytic vesicles have varying responses to optogenetic inactivation of kinesins and dynein

(A) Schematic for transport of different endocytic vesicles based on optogenetic inhibition of different motor proteins. The length of the arrow indicates the change in motility of the cargo upon optogenetic inhibition.

(B) Proposed model where cargo motility is more sensitive to the activity of kinesin-1 and dynein, while kinesin-2 and -3 may play a more supportive role in directing the cargo to its correct destination. The length of the arrow indicates the sensitivity of the respective motor in determining the net motility of the cargo.

KEY RESOURCES TABLE REAGENT or RESOURCE	SOURCE	IDENTIFIER	
	
Antibodies	
	
Mouse Monoclonal Anti-β-Tubulin	Sigma-Aldrich	T4026 clone TUB2.1; RRID:AB_477577	
	
Chemicals, peptides, and recombinant proteins	
	
Lipofectamine LTX	Invitrogen, Thermo Fisher Scientific	15338100	
Lysotracker Deep Red	Invitrogen, Thermo Fisher Scientific	L12492	
	
Experimental models: cell lines	
	
U-2OS	American Type Culture Collection, Manassas, VA	RRID: CVCL_0042	
COS-7	American Type Culture Collection, Manassas, VA	RRID: CVCL_0224	
	
Recombinant DNA	
	
pTriEx-mCherry-Zdk1	Wang et al.13	Addgene #81057; RRID:Addgene_81057	
pTriEx-NTOM20-LOV2	Wang et al.13	Addgene #81009; RRID:Addgene_81009	
pBa-FRB-3myc-KIF1A tail 391-1698	Bentley et al.37	Addgene # 64286; RRID:Addgene_64286	
iRFP-FRB-Rab5	Hammond et al. (2014)62	Addgene #51612; RRID:Addgene_51612	
iRFP-FRB-Rab7	Hammond et al. (2014)62	Addgene #51613; RRID:Addgene_51613	
pBa-FRB-3myc-KIF1B beta tail 387-1770	Bentley et al.37	Addgene #64287; RRID:Addgene_64287	
pBa.GFP-KIF16B tail 458-1317	Bentley et al.37	Addgene #134624; RRID:Addgene_134624	
KIF1A(1–393)-LZ-3xmCitrine	Gift from Kristen Verhey	Lab plasmid	
	
Software and algorithms	
	
MATLAB	Mathworks	https://mathworks.com/	
KymoButler	Jakobs et al. (2019)63	https://gitlab.com/deepmirror/kymobutler	
RF Cloning	Bond and Naus64	https://rf-cloning.org/index.php	
I-TASSER	Roy et al. (2010)65	https://zhanggroup.org/I-TASSER/	
AlphaFold	Jumper et al. (2021)66	https://colab.research.google.com/github/deepmind/alphafold/blob/main/notebooks/AlphaFold.ipynb	
PyMOL	The PyMOL Molecular Graphics System, Version 2.0 Schrödinger, LLC	http://www.pymol.org/pymol	
P-COILS	Gabler et al. (2020)67	https://toolkit.tuebingen.mpg.de/tools/pcoils	
Original code for trajectory and velocity analysis	This paper	https://doi.org/10.5281/zenodo.13146134	
FIJI	Schindelin et al. (2012)68	https://fiji.sc/	
	
Other	
	
TIRF Microscope	Eclipse Ti-E inverted microscope	Nikon, Melville, NY	
EMCCD camera	iXon U897	Andor Technology, South Windsor, CT	

Highlights

Optogenetic inhibition of KIF5B, KIF3A, KIF1A, and dynein motors using the LOVTRAP system

Kinesin-1 and dynein inhibition changes the direction of movement of endosomes

Kinesin-2 and -3 inhibition reduces the long-range motility of endosomes and lysosomes

DECLARATION OF INTERESTS

The authors declare no competing interests.

SUPPLEMENTAL INFORMATION

Supplemental information can be found online at https://doi.org/10.1016/j.celrep.2024.114649.
==== Refs
REFERENCES

1. Hirokawa N , Niwa S , and Tanaka Y (2010). Molecular motors in neurons: transport mechanisms and roles in brain function, development, and disease. Neuron 68 , 610–638.21092854
2. Franker MAM , and Hoogenraad CC (2013). Microtubule-based transport–basic mechanisms, traffic rules and role in neurological pathogenesis. J. Cell Sci. 126 , 2319–2329.23729742
3. Fu M.-meng , and Holzbaur ELF (2014). Integrated regulation of motor-driven organelle transport by scaffolding proteins. Trends Cell Biol. 24 , 564–574.24953741
4. Brady ST , Pfister KK , and Bloom GS (1990). A monoclonal antibody against kinesin inhibits both anterograde and retrograde fast axonal transport in squid axoplasm. Proc. Natl. Acad. Sci. USA 87 , 1061–1065.1689058
5. Martin M , Iyadurai SJ , Gassman A , Gindhart JG Jr. , Hays TS , and Saxton WM (1999). Cytoplasmic dynein, the dynactin complex, and kinesin are interdependent and essential for fast axonal transport. Mol. Biol. Cell 10 , 3717–3728.10564267
6. Engelke MF , Waas B , Kearns SE , Suber A , Boss A , Allen BL , and Verhey KJ (2019). Acute inhibition of heterotrimeric kinesin-2 function reveals mechanisms of intraflagellar transport in mammalian cilia. Curr. Biol. 29 , 1137–1148.e4.30905605
7. Hendricks AG , Holzbaur ELF , and Goldman YE (2012). Force measurements on cargoes in living cells reveal collective dynamics of microtubule motors. Proc. Natl. Acad. Sci. USA 109 , 18447–18452.23091040
8. Valetti C , Wetzel DM , Schrader M , Hasbani MJ , Gill SR , Kreis TE , and Schroer TA (1999). Role of dynactin in endocytic traffic: effects of dynamitin overexpression and colocalization with CLIP-170. Mol. Biol. Cell 10 , 4107–4120.10588646
9. Passmore JB , Nijenhuis W , and Kapitein LC (2021). From observing to controlling: Inducible control of organelle dynamics and interactions. Curr. Opin. Cell Biol. 71 , 69–76.33706236
10. Edward RB , Ayloo S , Chenoweth DM , Lampson MA , and Holzbaur ELF (2015). Optogenetic control of organelle transport using a photocaged chemical inducer of dimerization. Curr. Biol. 25 , R407–R408.25989077
11. Van Bergeijk P , Adrian M , Hoogenraad CC , and Kapitein LC (2015). Optogenetic control of organelle transport and positioning. Nature 518 , 111–114.25561173
12. Nijenhuis W , van Grinsven MMP , and Kapitein LC (2020). An optimized toolbox for the optogenetic control of intracellular transport. J. Cell Biol. 219 , e201907149.32328628
13. Wang H , Vilela M , Winkler A , Tarnawski M , Schlichting I , Yumerefendi H , Kuhlman B , Liu R , Danuser G , and Hahn KM (2016). LOVTRAP: an optogenetic system for photoinduced protein dissociation. Nat. Methods 13 , 755–758.27427858
14. Hammond JW , Blasius TL , Soppina V , Cai D , and Verhey KJ (2010). Autoinhibition of the kinesin-2 motor KIF17 via dual intramolecular mechanisms. J. Cell Biol. 189 , 1013–1025.20530208
15. Torisawa T , Ichikawa M , Furuta A , Saito K , Oiwa K , Kojima H , Toyoshima YY , and Furuta Ken’ya (2014). Autoinhibition and cooperative activation mechanisms of cytoplasmic dynein. Nat. Cell Biol. 16 , 1118–1124.25266423
16. Chiba K , Ori-McKenney KM , Niwa S , and McKenney RJ (2022). Synergistic autoinhibition and activation mechanisms control kinesin-1 motor activity. Cell Rep. 39 , 111016.35767955
17. Tan Z , Yue Y , Leprevost F , Haynes S , Basrur V , Nesvizhskii AI , Verhey KJ , and Cianfrocco MA (2023). Autoinhibited kinesin-1 adopts a hierarchical folding pattern. bioRxiv. 10.1101/2023.01.26.525761.
18. Weijman JF , Yadav SKN , Surridge KJ , Cross JA , Borucu U , Mantell J , Woolfson DN , Schaffitzel C , and Dodding MP (2022). Molecular architecture of the autoinhibited kinesin-1 lambda particle. Sci. Adv. 8 , eabp9660.36112680
19. Hackney DD , and Stock MF (2000). Kinesin’s IAK tail domain inhibits initial microtubule-stimulated ADP release. Nat. Cell Biol. 2 , 257–260.10806475
20. Dietrich KA , Sindelar CV , Brewer PD , Downing KH , Cremo CR , and Rice SE (2008). The kinesin-1 motor protein is regulated by a direct interaction of its head and tail. Proc. Natl. Acad. Sci. USA 105 , 8938–8943.18579780
21. Yonekura H , Nomura A , Ozawa H , Tatsu Y , Yumoto N , and Uyeda TQP (2006). Mechanism of tail-mediated inhibition of kinesin activities studied using synthetic peptides. Biochem. Biophys. Res. Commun. 343 , 420–427.16546134
22. Jiang X , Ogawa T , Yonezawa K , Shimizu N , Ichinose S , Uchihashi T , Nagaike W , Moriya T , Adachi N , Kawasaki M , (2023). The two-step cargo recognition mechanism of heterotrimeric kinesin. EMBO Rep. 24 , e56864.37575008
23. Brunnbauer M , Mueller-Planitz F , Kösem S , Ho TH , Dombi R , Gebhardt JCM , Rief M , and Ökten Z (2010). Regulation of a heterodimeric kinesin-2 through an unprocessive motor domain that is turned processive by its partner. Proc. Natl. Acad. Sci. USA 107 , 10460–10465.20498083
24. Ichinose S , Ogawa T , and Hirokawa N (2015). Mechanism of activity-dependent cargo loading via the phosphorylation of kif3a by pka and camkiia. Neuron 87 , 1022–1035.26335646
25. Hammond JW , Cai D , Blasius TL , Li Z , Jiang Y , Jih GT , Meyhofer E , and Verhey KJ (2009). Mammalian kinesin-3 motors are dimeric in vivo and move by processive motility upon release of autoinhibition. PLoS Biol. 7 , e1000072.19338388
26. Ren J , Zhang Y , Wang S , Huo L , Lou J , and Feng W (2018). Structural delineation of the neck linker of kinesin-3 for processive movement. J. Mol. Biol. 430 , 2030–2041.29752968
27. Huo L , Yue Y , Ren J , Yu J , Liu J , Yu Y , Ye F , Xu T , Zhang M , and Feng W (2012). The CC1-FHA tandem as a central hub for controlling the dimerization and activation of kinesin-3 KIF1A. Structure 20 , 1550–1561.22863567
28. Farkhondeh A , Niwa S , Takei Y , and Hirokawa N (2015). Characterizing kif16b in neurons reveals a novel intramolecular “stalk inhibition” mechanism that regulates its capacity to potentiate the selective somato-dendritic localization of early endosomes. J. Neurosci. 35 , 5067–5086.25810535
29. Zhang K , Foster HE , Rondelet A , Lacey SE , Bahi-Buisson N , Bird AW , and Carter AP (2017). Cryo-EM reveals how human cytoplasmic dynein is auto-inhibited and activated. Cell 169 , 1303–1314.e18.28602352
30. Quintyne NJ , Gill SR , Eckley DM , Crego CL , Compton DA , and Schroer TA (1999). Dynactin is required for microtubule anchoring at centrosomes. J. Cell Biol. 147 , 321–334.10525538
31. Pfister KK (2015). Distinct functional roles of cytoplasmic dynein defined by the intermediate chain isoforms. Exp. Cell Res. 334 , 54–60.25576383
32. Christoforidis S , McBride HM , Burgoyne RD , and Zerial M (1999). The Rab5 effector EEA1 is a core component of endosome docking. Nature 397 , 621–625.10050856
33. Nielsen E , Severin F , Backer JM , Hyman AA , and Zerial M (1999). Rab5 regulates motility of early endosomes on microtubules. Nat. Cell Biol. 1 , 376–382.10559966
34. Villari G , Enrico Bena C , Del Giudice M , Gioelli N , Sandri C , Camillo C , Fiorio Pla A , Bosia C , and Serini G (2020). Distinct retrograde microtubule motor sets drive early and late endosome transport. The EMBO journal 39 , e103661.33215754
35. Qiu R , Zhang J , and Xiang X (2023). Kinesin-1 autoinhibition facilitates the initiation of dynein cargo transport. J. Cell Biol. 222 , e202205136.36524956
36. Hoepfner S , Severin F , Cabezas A , Habermann B , Runge A , Gillooly D , Stenmark H , and Zerial M (2005). Modulation of receptor recycling and degradation by the endosomal kinesin KIF16B. Cell 121 , 437–450.15882625
37. Bentley M , Decker H , Luisi J , and Banker G (2015). A novel assay reveals preferential binding between Rabs, kinesins, and specific endosomal subpopulations. J. Cell Biol. 208 , 273–281.25624392
38. Bielska E , Schuster M , Roger Y , Berepiki A , Soanes DM , Talbot NJ , and Steinberg G (2014). Hook is an adapter that coordinates kinesin-3 and dynein cargo attachment on early endosomes. J. Cell Biol. 204 , 989–1007.24637326
39. Brown CL , Maier KC , Stauber T , Ginkel LM , Wordeman L , Vernos I , and Schroer TA (2005). Kinesin-2 is a motor for late endosomes and lysosomes. Traffic 6 , 1114–1124.16262723
40. Cullen PJ , and Steinberg F (2018). To degrade or not to degrade: mechanisms and significance of endocytic recycling. Nat. Rev. Mol. Cell Biol. 19 , 679–696.30194414
41. Casey JR , Grinstein S , and Orlowski J (2010). Sensors and regulators of intracellular pH. Nat. Rev. Mol. Cell Biol. 11 , 50–61.19997129
42. Yi JY , Ori-McKenney KM , McKenney RJ , Vershinin M , Gross SP , and Vallee RB (2011). High-resolution imaging reveals indirect coordination of opposite motors and a role for LIS1 in high-load axonal transport. J. Cell Biol. 195 , 193–201.22006948
43. Poteryaev D , Datta S , Ackema K , Zerial M , and Spang A (2010). Identification of the switch in early-to-late endosome transition. cell 141 , 497–508.20434987
44. Jordens I , Fernandez-Borja M , Marsman M , Dusseljee S , Janssen L , Calafat J , Janssen H , Wubbolts R , and Neefjes J (2001). The Rab7 effector protein RILP controls lysosomal transport by inducing the recruitment of dyneindynactin motors. Curr. Biol. 11 , 1680–1685.11696325
45. Pankiv S , Alemu EA , Brech A , Bruun J-A , Lamark T , Øvervatn A , Bjørkøy G , and Johansen T (2010). FYCO1 is a Rab7 effector that binds to LC3 and PI3P to mediate microtubule plus end–directed vesicle transport. J. Cell Biol. 188 , 253–269.20100911
46. Huotari J , and Helenius A (2011). Endosome maturation. The EMBO journal 30 , 3481–3500.21878991
47. Pu J , Guardia CM , Keren-Kaplan T , and Bonifacino JS (2016). Mechanisms and functions of lysosome positioning. J. Cell Sci. 129 , 4329–4339.27799357
48. Müller MJI , Klumpp S , and Lipowsky R (2008). Tug-of-war as a cooperative mechanism for bidirectional cargo transport by molecular motors. Proc. Natl. Acad. Sci. USA 105 , 4609–4614.18347340
49. Allison MG , Ma T-C , Feng Q , Jiang R , Badieyan S , Cianfrocco MA , and Hancock WO (2022). Kinesin-1,-2, and-3 motors use family-specific mechanochemical strategies to effectively compete with dynein during bidirectional transport. Elife 11 , e82228.36125250
50. Beaudet D , Berger CL , and Hendricks AG (2024). The types and numbers of kinesins and dyneins transporting endocytic cargoes modulate their motility and response to tau. J. Biol. Chem. 300 , 107323.38677516
51. Chaudhary AR , Berger F , Berger CL , and Hendricks AG (2018). Tau directs intracellular trafficking by regulating the forces exerted by kinesin and dynein teams. Traffic 19 , 111–121.29077261
52. Serra-Marques A , Martin M , Katrukha EA , Grigoriev I , Peeters CAE , Liu Q , Hooikaas PJ , Yao Y , Solianova V , Smal I , (2020). Concerted action of kinesins kif5b and kif13b promotes efficient secretory vesicle transport to microtubule plus ends. Elife 9 , e61302.33174839
53. Tjioe M , Shukla S , Vaidya R , Troitskaia A , Bookwalter CS , Trybus KM , Chemla YR , and Selvin PR (2019). Multiple kinesins induce tension for smooth cargo transport. Elife 8 , e50974.31670658
54. Chaudhary AR , Lu H , Krementsova EB , Bookwalter CS , Trybus KM , and Hendricks AG (2019). MAP7 regulates organelle transport by recruiting kinesin-1 to microtubules. J. Biol. Chem. 294 , 10160–10171.31085585
55. Guardia CM , Farías GG , Jia R , Pu J , and Bonifacino JS (2016). BORC functions upstream of kinesins 1 and 3 to coordinate regional movement of lysosomes along different microtubule tracks. Cell Rep. 17 , 1950–1961.27851960
56. Tas RP , Chazeau A , Cloin BMC , Lambers MLA , Hoogenraad CC , and Kapitein LC (2017). Differentiation between oppositely oriented microtubules controls polarized neuronal transport. Neuron 96 , 1264–1271.e5.29198755
57. Ally S , Larson AG , Barlan K , Rice SE , and Gelfand VI (2009). Oppositepolarity motors activate one another to trigger cargo transport in live cells. J. Cell Biol. 187 , 1071–1082.20038680
58. De Rossi MC , Wetzler DE , Benseñor L , Rossi MED , Sued M , Rodríguez D , Gelfand V , Bruno L , and Levi V (2017). Mechanical coupling of microtubule-dependent motor teams during peroxisome transport in drosophila S2 cells. Biochim. Biophys. Acta. Gen. Subj. 1861 , 3178–3189.28935608
59. Arpag G , Norris SR , Iman Mousavi S , Soppina V , Verhey KJ , Hancock WO , and Tüzel E (2019). Motor dynamics underlying cargo transport by pairs of kinesin-1 and kinesin-3 motors. Biophys. J. 116 , 1115–1126.30824116
60. Ohashi KG , Han L , Mentley B , Wang J , Fricks J , and Hancock WO (2019). Load-dependent detachment kinetics plays a key role in bidirectional cargo transport by kinesin and dynein. Traffic 20 , 284–294.30809891
61. Monzon GA , Scharrel L , DSouza A , Henrichs V , Santen L , and Diez S (2020). Stable tug-of-war between kinesin-1 and cytoplasmic dynein upon different ATP and roadblock concentrations. J. Cell Sci. 133 , jcs249938.33257498
62. Hammond GR , Machner MP , and Balla T (2014). A novel probe for phosphatidylinositol 4-phosphate reveals multiple pools beyond the Golgi. J. Cell Biol. 205 , 113–126.24711504
63. Jakobs MA , Dimitracopoulos A , and Franze K (2019). KymoButler, a deep learning software for automated kymograph analysis. elife 8 , e42288.31405451
64. Bond SR , and Naus CC (2012). RF-cloning. org: an online tool for the design of restriction-free cloning projects. Nucleic Acids Res. 40 , W209–W213.22570410
65. Roy A , Kucukural A , and Zhang Y (2010). I-TASSER: a unified platform for automated protein structure and function prediction. Nat. Protoc. 5 , 725–738.20360767
66. Jumper J , Evans R , Pritzel A , Green T , Figurnov M , Ronneberger O , Tunyasuvunakool K , Bates R , Žídek A , Potapenko A , and Bridgland A (2021). Highly accurate protein structure prediction with AlphaFold. Nature 596 , 583–589.34265844
67. Gabler F , Nam SZ , Till S , Mirdita M , Steinegger M , Söding J , Lupas AN , and Alva V (2020). Protein sequence analysis using the MPI bioinformatics toolkit. Curr. Protoc. Bioinformatics 72 , e108.33315308
68. Schindelin J , Arganda-Carreras I , Frise E , Kaynig V , Longair M , Pietzsch T , Preibisch S , Rueden C , Saalfeld S , Schmid B , and Tinevez JY (2012). Fiji: an open-source platform for biological-image analysis. Nat. Methods 9 , 676–682.22743772
69. Anton VB , and Matsumura I (2010). Overlap extension PCR cloning: a simple and reliable way to create recombinant plasmids. Biotechniques 48 , 463–465.20569222
70. Tinevez J-Y , Perry N , Schindelin J , Hoopes GM , Reynolds GD , Laplantine E , Bednarek SY , Shorte SL , and Eliceiri KW (2017). Trackmate: An open and extensible platform for single-particle tracking. Methods 115 , 80–90.27713081
71. Bananis E , Nath S , Gordon K , Satir P , Stockert RJ , Murray JW , and Wolkoff AW (2004). Microtubule-dependent movement of late endocytic vesicles in vitro: requirements for dynein and kinesin. Mol. Biol. Cell 15 , 3688–3697.15181154
72. Lessard DV , Zinder OJ , Hotta T , Verhey KJ , Ohi R , and Berger CL (2019). Polyglutamylation of tubulin’s c-terminal tail controls pausing and motility of kinesin-3 family member kif1a. J. Biol. Chem. 294 , 6353–6363.30770469
