Computational protocol: AMPK activity regulates trafficking of mitochondria to the leading edge during cell migration and matrix invasion

Similar protocols

Protocol publication

[…] ATP and ADP levels were measured using either the ATPlite assay kit from PerkinElmer (Waltham, MA) or the respective kits from Abcam (Cambridge, MA) according to the manufacturers’ instructions. Samples were read on a Synergy HT plate reader (BioTek, Winooski, VT). After luminescence values were obtained, protein concentrations for each sample were measured; the nucleotide levels are expressed as relative luminescence/[protein].To determine actual nucleotide concentrations (mM, rather than nucleotide concentration relative to protein concentration), volumes for cell bodies (CBs) and pseudopodia (Pd) for SKOV-3 cells were calculated in two ways. The first was a literature-based method in which the values for the average amount of CB or Pd protein yielded per Transwell insert (472.4 and 49.2 μg, respectively) were divided by 100 mg/ml, an approximate value for the concentration of soluble protein in the cytoplasm of mammalian cells (e.g., The result was divided by the number of cells plated per insert (i.e., 1.5 × 106) to yield a nominal CB volume of ∼3149 μm3 (femtoliters) and a per-cell Pd volume of ∼328 μm3. Given that the average number of Pd per cell in our preparations is 3.6, the volume of an individual Pd is ∼91 μm3. The second calculation involved physical measurement of CB and Pd volumes from photomicrographs of cells cultured on Transwell inserts and then fixed and stained with fluorescent phalloidin or CellTracker Green BODIPY. A total of 90 cells (∼5 cells/insert from six different inserts from three different Pd preparations) were imaged by wide-field or confocal microscopy, and CB and Pd volumes were measured by hand using the MeasureStack plug-in for ImageJ ( This approach yielded a per-cell Pd volume of ∼385 μm3 and a CB volume of ∼3485 μm3. The average nuclear volume (∼215 μm3) was subtracted from the measured CB volume to give a cytoplasmic CB volume of ∼3270 μm3 (this is an appropriate adjustment, as the nuclei are spun out during the harvesting of CBs from a typical preparation). Thus the volumes of 3149 and 3270 μm3 for the CB and 328 and 385 μm3 for total Pd per cell measured by two very distinct methods are in excellent agreement with each other and within the range of volumes reported for other mammalian cell types. Furthermore, dividing the average yield of CB or Pd protein from a preparation by these values gives CB and Pd protein concentrations of 96.3 and 85.3 mg/ml, respectively; these values are also in fair agreement with the aforementioned assumption of 100 mg/ml. The values derived from direct volumetric measurements were subsequently used as nominal volumes for determining actual concentrations of ATP and ADP in CBs and Pd. ATP and ADP concentrations in extracts (5 μg of protein) were determined from separate aliquots of the same extracts using the described ATPlite assay. ADP levels were determined by conversion of ADP to ATP by inclusion of pyruvate kinase and phosphoenolpyruvate in the extract (using reagents and protocols from the ADP Assay Kit [Abcam]) and subtraction of the parallel values for unconverted ATP. Assay concentrations (μM) were converted into empirical concentrations (mM) by the formula ((([assay]/1000) × 200)/5) × [protein]CB or Pd, where division by 1000 converts μM to mM, 200 is the assay volume (in μl), and 5 is the amount of extract protein in the reaction (in μg), which reduces to ([assay]/25) × [protein]CB or Pd. [...] Time-coded stacks of mitochondria over time were generated using ImageJ. Briefly, images were uniformly adjusted for brightness/contrast, and a “top-hat” filter was applied to isolate bright mitochondria from background (). Filtered images were used create binary masks, which were then multiplied by the original fluorescence image. The resulting image stack was color-coded as a function of time using the Temporal-Color Code function and a modified Rainbow Smooth look-up table. Stack projections were rendered as either maximum intensity projections or progressive overlays using the Z Code Stack plug-in (Supplemental Movie S1).Mitochondrial tracks were calculated using the MTrack2 ImageJ plug-in, and vectors were calculated by plotting the final position as a function of normalizing the starting position of each mitochondrion to (0, 0). We have cos θ = 1 when the angle of mitochondrial motility vectors (θ) is equal to the angle from the center of the cell to the cell periphery proximal to the path of migration (leading edge). Conversely, cos θ = −1 when the angle of mitochondrial motility vectors (θ) is opposite the angle from the center of the cell to the cell periphery distal to the path of migration (trailing edge). The instantaneous velocity for an individual mitochondrion was defined as the Euclidean distance traveled between any two successive frames of a time-lapse series divided by the acquisition interval. For each mitochondrion analyzed, the largest of these values was designated the maximum instantaneous velocity, and the average of these values represented the mean velocity. The same approach was used to determine the instantaneous and mean velocities of proximal membrane segments. Mitochondrial flux was determined using the standard mass flux equation, jm = (Δm/Δt)/A, where m is the thresholded mitochondrial fluorescence area (a surrogate for mitochondrial mass), t is time (specifically, Δt = 3× the image acquisition interval [typically 30–90 s]), and A is the area (specifically, a region of interest that is ≤20% of the total cellular area and encompasses the leading edge, trailing edge, cell body, or periphery, as indicated). […]

Pipeline specifications

Software tools ImageJ, Mtrack2
Applications Laser scanning microscopy, Microscopic phenotype analysis
Diseases Ovarian Neoplasms
Chemicals Adenosine Triphosphate, Ampicillin