The 8 nm regular stepping of kinesin differs substantially from the cytoplasmic dynein motor, which takes a wide range of forward (4–24 nm), backward (20%), as well as sideways steps (Reck-Peterson et al., 2006 and Gennerich et al., 2007). However, by extending the kinesin neck linker, we were able to produce a motor with a stepping behavior that is more similar to dynein than to WT kinesin. This result reveals that the regular pattern of stepping can be modulated by varying the length and flexibility of the elements interconnecting the two motor domains. In some cases, it may be advantageous to design highly efficient and regular motors (kinesin and myosin V), whereas in other cases, reducing motor efficiency but increasing its ability to move around the track (dynein and myosin VI) may help to avoid obstacles (Dixit et al., 2008). It will be interesting to explore how modulating a motor's stepping behavior affects its ability to execute transport functions in living cells.
Experimental Procedures
Single Molecule Fluorescence Motility Assays
The construct design is described in the Supplemental Experimental Procedures.
For single-molecule motility assays, sea urchin axonemes were immobilized on a glass surface, and 200 pM kinesin motor was then perfused into the chamber in motility buffer (12 mM PIPES [pH 6.8], 1 mM MgCl2, 2 mM EGTA, and 1 mg/ml casein) containing 2 mM DTT, 2% glucose, and 3 μl “gloxy” to remove free oxygen in solution (Yildiz et al., 2003). Speed and run length measurements were performed with GFP-labeled motors at 1 mM ATP using total internal reflection fluorescence microscopy, as described elsewhere (Reck-Peterson et al., 2006).
For high-precision fluorescent tracking, a unique reactive cysteine was introduced in the catalytic core (E215C), dialyzed motors were labeled with biotin-maleimide (EZ-link PEO2, Pierce) at 0.5 biotins per head (2 hr, 4°C), the reaction was quenched with 2 mM DTT, and excess of biotin was removed through microtubule affinity purification. Microtubule-bound, biotinylated kinesin motors were incubated with 400 nM streptavidin-coated quantum dots (655 nm, Invitrogen) for 5 min within the flow chamber in the absence of ATP (reaction was performed on microtubule-bound kinesin to avoid aggregation from multivalent quantum dots). The sample was washed with imaging buffer containing 140 mM 2-mercaptoethanol. Motor velocity was kept in the range of 10–15 nm/s by adding different concentrations of ATP for each construct (see Figure 3 legend). At this speed, we could minimize scoring two successive steps as one “large step,” since the motor's average dwell time (950 ms) was long, compared with the temporal resolution (70 ms). Our step detection program scores a dwell period consisting of three data points (210 ms). At this time resolution, we estimate that only 2% of our scored steps will be due to the rapid succession of two steps. We also compared stepping of different constructs at similar average dwell times to minimize artifacts (such as selectively scoring two steps as one large step for one construct).
To crosslink the C termini of the neck-linker region, a cysteine was introduced to position 337 of the 13P cysteine-free mutant (Cys-13P). Purified Cys-13P (2 μM) were treated with 2 mM TCEP for 30 min and then were passed through a desalting column (NAP5, GE Healthcare), and a bifunctional maleimide crosslinker (BMOE, Pierce) was added at an equimolar ratio to motor for 4 hr in 80 mM Pipes (pH 7.0), 100 mM NaCl, 1 mM MgCl2, and 1 mM EGTA at 4°C. The remaining crosslinker was quenched with 10 mM DTT for 30 min.