The Effect of Temperature on Microtubule-Based Transport by Cytoplasmic Dynein and Kinesin-1 Motors

Abstract

Cytoplasmic dynein and kinesin are both microtubule-based molecular motors but are structurally and evolutionarily unrelated. Under standard conditions, both move with comparable unloaded velocities toward either the microtubule minus (dynein) or plus (most kinesins) end. This similarity is important because it is often implicitly incorporated into models that examine the balance of cargo fluxes in cells and into models of the bidirectional motility of individual cargos. We examined whether this similarity is a robust feature, and specifically whether it persists across the biologically relevant temperature range. The velocity of mammalian cytoplasmic dynein, but not of mammalian kinesin-1, exhibited a break from simple Arrhenius behavior below 15°C-just above the restrictive temperature of mammalian fast axonal transport. In contrast, the velocity of yeast cytoplasmic dynein showed a break from Arrhenius behavior at a lower temperature (∼8°C). Our studies implicate cytoplasmic dynein as a more thermally tunable motor and therefore a potential thermal regulator of microtubule-based transport. Our theoretical analysis further suggests that motor velocity changes can lead to qualitative changes in individual cargo motion and hence net intracellular cargo fluxes. We propose that temperature can potentially be used as a noninvasive probe of intracellular transport.

Figure 1

Temperature impacts the velocities of kinesin and mammalian cytoplasmic dynein differently. (A) Kinesin velocity and Arrhenius fit (solid line) down to 8°C. (B) Dynein velocity and fit to a piecewise Arrhenius trend (solid line) with a crossover at ∼15°C. (C) Direct comparison of the trends in (A) (dashed gray line) and (B) (solid black line) on the logarithmic Arrhenius plot. Inset: ratio of the fit curves obtained in (A) and (B) plotted on a linear scale. (D) Yeast cytoplasmic dynein velocity and fit to a piecewise Arrhenius trend (solid line) with a crossover at ∼8°C (activation energy 50.5 kJ/mol and 151.5 kJ/mol above and below 8°C, respectively). (E) Arrhenius plot of (D). Error bars, mean ± SE.

Figure 2

Temperature dependence of kinesin and mammalian cytoplasmic dynein processivity and stall force. (A and B) Kinesin processivity at 5°C (A) is 497.7 ± 0.09 μm/s, which is significantly lower than the 841.5 ± 0.2 μm/s obtained at room temperature (B). (C) Dynein processivity is 0.69 ± 0.09 μm/s at 10°C. Error bars, mean ± SE. (D and E) Force production by single kinesin (D) and mammalian cytoplasmic dynein (E) motors (kinesin: 5.3 ± 0.2 pN at 295 K vs. 5.2 ± 0.2 pN at 280.5 K; dynein: 1.2 ± 0.1 pN at 286.5 K vs. 1.2 ± 0.1 pN at 302.5 K; error bars: mean ± SE).

Figure 3

Simulations of cargo transported by a team of one kinesin and four dyneins (5 pN and 1.25 pN stall force, respectively). Motors were simulated using the anisotropic force-detachment relationship (Fig. S4 and Supporting Materials and Methods). (A) Independent traces of simulated bead motion at 275 K (black) and 286 K (dark gray) are shown superimposed (100 traces for each temperature). (B and C) Probability that a simulated trace will have a positive final location (i.e., the probability of kinesin winning) for (B) baseline and (C) 10× higher dynein processivity values at high, intermediate, and low temperatures (310 K, 286 K, 275 K). (D) Transport velocity histograms illustrate that the directional preference reverses sign as a function of temperature. Note that the velocity undergoes large changes with temperature, necessitating the rescaling of the x axis.

Figure 4

Simulations of cargo transported by a team of one kinesin and four dyneins (5 pN and 1.25 pN stall force, respectively). Motors were simulated using the isotropic force-detachment relationship (model B, Fig. S5). (A) Independent traces of simulated bead motion at 275 K (dark gray) and 286 K (light gray) are shown superimposed (100 traces for each temperature). (B) Probability that a simulated trace will have a positive final location (i.e., the probability of kinesin winning) for baseline (1×: solid line) and 10× (dashed line) higher dynein processivity values. (C) Transport velocity histograms reveal that the directional preference reverses sign as a function of temperature. Note that the velocity undergoes large changes with temperature, necessitating the rescaling of the x axis. (D) Model: when all motors are active at high temperatures (top), some ensembles of kinesin and dynein motors will exhibit motility with an overall bias in the minus-end direction on MTs. However, at low temperatures (bottom), dynein steps dramatically more slowly than kinesin, leading to an overall transport bias in the plus-end direction, as well as other potentially observable effects (Movies S1, S2, and S3).