Kinetochores specifically capture plus ends and grip more strongly when pulled toward plus ends. (A) Tip-captured, polarity-marked microtubules (magenta), polymerized with dim plus ends and bright minus ends, nearly always bound assembled Ndc80-GFP kinetochores (cyan) by their plus ends. The bar graph shows the percentage of tip-attached microtubules that were bound by their plus ends (mean ± SD from N = 8 experiments examining a total of 196 tip-captured microtubules). (B) Schematic of rupture force assay. Native kinetochore particles isolated from yeast were conjugated sparsely to polystyrene microbeads. A laser trap was used to attach a kinetochore bead to either the plus or minus end of an individual dynamic microtubule and then to measure the rupture strength of the attachment. (C) Left: Distribution of rupture strengths for yeast kinetochores attached to plus and minus ends (N = 43 and 26 events, respectively). Open circles represent individual strength measurements. Triangles show censored data when rupture strength exceeded the maximum force of the trap or the microtubule broke away from the coverslip surface. Boxes extend from first to third quartiles with medians indicated by central horizontal solid lines. Whiskers extend ± one SD from means, which are indicated by dashed black lines. Right: Rupture strengths for individual kinetochore beads measured sequentially at both ends of the same microtubule, either minus end first or plus end first as indicated. (D) Schematic of bidirectional sliding assay. (E) Record of force and position versus time for a kinetochore-decorated bead (yeast KT) attached to the side of a coverslip-anchored microtubule and pulled alternately toward the plus (red trace) and minus end (blue trace). Additional records are shown in Fig. S3. (F) Distribution of bidirectional sliding speeds for side-attached yeast kinetochores (KTs) measured at 1 ± 0.5 pN of force applied toward plus and minus ends (N = 105 and 123 events, respectively). Dots represent the speeds of individual sliding events. Boxes extend from first to third quartiles with medians indicated by central horizontal solid lines. Whiskers extend ± one SD from means, which are indicated by dashed black lines. (G) Kinetochore beads were tested on both long and short extensions, to confirm that the speed differential arises from microtubule polarity rather than asymmetric anchorage of the microtubule to the coverslip. Frictional drag coefficients for individual kinetochore beads sliding toward plus (red symbols) and minus ends (blue symbols) on short and long microtubule extensions, as indicated. Beads 1–4 were measured sequentially on both extensions of the same microtubule. Symbols represent mean frictional drag coefficient ± SEM (from N > 5 sliding events per bead–microtubule pair). (H) Record of force and position versus time for a bead coated with human Ndc80 complex attached to the side of a coverslip-anchored microtubule and pulled alternately toward the plus (red trace) and minus end (blue trace). (I) Distribution of bidirectional sliding speeds for human Ndc80c-coated beads measured at 1 ± 0.5 pN of force applied toward plus and minus ends (N = 39 and 47 events, respectively), plotted as in panel F.