On the use of in vivo cargo velocity as a biophysical marker

Abstract

Molecular motors move many intracellular cargos along microtubules. Recently, it has been hypothesized that in vivo cargo velocity can be used to determine the number of engaged motors. We use theoretical and experimental approaches to investigate these assertions, and find that this hypothesis is inconsistent with previously described motor behavior, surveyed and re-analyzed in this paper. Studying lipid droplet motion in Drosophila embryos, we compare transport in a mutant, Delta(halo), with that in wild-type embryos. The minus-end moving cargos in the mutant appear to be driven by more motors (based on in vivo stall force observations). Periods of minus-end motion are indeed longer than in wild-type embryos but the corresponding velocities are not higher. We conclude that velocity is not a definitive read-out of the number of motors propelling a cargo.

Figure 1

Normalized velocity as a function of relative load per motor due to viscous load (ℓ, Equation 3) from three studies of saltatory motion of intracellular cargos (Supplementary Table 1). Sustained velocities as a function of relative load per motor for plus and minus-end transport of Drosophila peroxisomes (left) [3] and Xenopous pigment organelles (center) [4] and plus-end vesicle transport in PC12 neurites [22]. In all cases, the load per motor given by the VENoM model does not appear to correlate with cargo velocity. The peroxisome and pigment organelle velocities were normalized by the velocity attributed to a single motor (v(1)). The neuronal vesicle transport was reported [26] already normalized on a per-run basis by the lowest motion velocity of each run (this was reported to be broadly distributed around v=0.44±0.009 μm/s).

Figure 2

(A) A brightfield image of a developing Drosophila embryo with a high magnification DIC image showing the lipid droplets (arrow). Lipid droplets move along microtubules surrounding the nuclei (N, outlined in black). The direction of motion of a lipid droplet along such a microtubule is shown in green. (B) An example of a lipid droplet track with a long plus and a long minus run separated by a sequence of shorter runs and pauses. The detailed view in (C) shows the start and end of the plus and minus runs. The start and end points are highlighted by filled points. Minus runs exhibit distinct slow-down at the ends of runs.

Figure 3

Average run-lengths and velocities in the minus-end directions. Δ(halo) minus runs are longer and slower than wild-type (two sided Student’s t test, p-value ≤ 10⁻⁴). These velocities are higher than previously reported because in this analysis we only used long runs ( ≥ 0.5 μm) [23].

Figure 4

Thorough quantification of velocities at the start and end of runs reveals that velocities at the ends of minus runs are consistently lower up to 0.33 s from the endpoint. (A) The start and end velocities were calculated for a window 0.3 sec in duration that was made to slide into the run for a total of 0.3 sec as illustrated. (B) Mean start, end, and overall run velocities are compared for minus and plus runs. This procedure was followed to avoid artifacts resulting from imprecise determination of endpoints. In general, end velocities are lower than start and overall average run velocities. All end velocities of minus runs are statistically different from corresponding group velocities (one sided t-test, p-value ≤ 0.01). All start and end plus velocities are within measurement error of mean run velocity. Error bars are drawn on one side for clarity. (The number of runs used for this Figure ranged from 72 to 59 for minus-end motion and from 71 to 67 for plus-end motion). (C) Velocity distributions for the start and end of minus runs show a shift towards lower velocities at the ends of the runs.