The C-terminal transmembrane region of mouse mitochondrial outer membrane protein of 25 kDa (OMP) cDNA and mouse VAMP2 cDNA were cloned by polymerase chain reaction. The sequences were verified by DNA sequencing. Human amyloid precursor protein 695 (APP) -venus plasmid was provided by Dr Sakurai (Juntendo University; Sakurai et al., 2008). EGFP-OMP, EGFP-VAMP2 and APP-EGFP were generated by inserting the coding region into Enhanced Green Fluorescent Protein (EGFP) vectors (Clontech, Mountain View, CA, USA). The mCherry-OMP and APP-mCherry were generated by replacing the EGFP coding region with the coding region of mCherry (Shaner et al., 2004). The DNA fragments coding for EGFP and mCherry fusion proteins were inserted into the expression plasmids containing β-actin promoter sequences (Ebihara et al., 2003). G-CaMP6 plasmid was provided by Dr Nakai (Saitama University; Ohkura et al., 2012).
Hippocampal neurons were fixed in 2% paraformaldehyde in phosphate-buffered saline for 25 min, permeabilised with 0.2% Triton X-100 for 5 min, blocked with 5% normal goat serum for 30 min and reacted with mouse monoclonal antibody to cytochrome c (Promega, Madison, WI, USA). The first antibody was visualised by secondary antibody staining using goat anti-mouse IgG conjugated to Alexa 647 (Molecular Probes, Eugene, OR, USA).
All procedures were performed at room temperature (set at 24 °C). FM1-43 (Molecular Probes) loading was performed by exposing neurons to the dye (15 μm) in high-K+ saline solution (75 mm NaCl, 70 mm KCl, 2 mm CaCl2, 2 mm MgCl2, 5 mm HEPES and 20 mm glucose, pH 7.4) for 2 min followed by washing in low-Ca2+ saline solution (140 mm NaCl, 5 mm KCl, 0.1 mm CaCl2, 4 mm MgCl2, 5 mm HEPES and 20 mm glucose, pH 7.4) three times for 2 min. After taking the first images in low-Ca2+ saline solution, neurons were exposed to high-K+ saline solution for 2 min and then switched to low-Ca2+ saline solution again for washing. Second images were taken at the same axonal regions. The difference of fluorescence intensity between the first and second images was used for analysis as FM1-43(Δ).
Images were obtained by using a Fluoview confocal laser-scanning microscope with ×60 1.4 NA oil-immersion lenses (Olympus, Tokyo, Japan). A confocal aperture was set at a diameter of 600–700 μm. For some images, multiple optical sections (3–7 sections and z-spacing of 1.0 μm) were collected, and these images were recombined using a maximum-brightness operation. The axons were identified morphologically and we selected imaging areas at least 100 μm away from the soma.
For time-lapse imaging, live cells were mounted in a chamber at 37 °C with a water bath and continuous flow of humidified 5% CO2 to maintain the osmolality and pH of the medium during prolonged time-lapse experiments. For time-lapse imaging with tetrodotoxin (TTX; Wako, Tokyo, Japan), the first frame was imaged at least 30 min after adding TTX to the medium (final concentration, 1 μm). For time-lapse imaging at intervals of 1 day, the duration of single imaging sessions was restricted within 30 min. For an analysis of transport properties, mCherry-OMP was imaged at intervals of 3 s and APP-mCherry was imaged at intervals of 1 s. Because the average velocities measured in these conditions were similar to previously reported values, images taken at these intervals can faithfully report the transport characteristics of mitochondria and APP-containing vesicles (Fig. 5) (Kaether et al., 2000; MacAskill & Kittler, 2010).
For time-lapse imaging with electrical field stimulation, neurons in Tyrode's solution (119 mm NaCl, 2.5 mm KCl, 2 mm CaCl2, 2 mm MgCl2, 25 mm HEPES and 30 mm glucose, pH 7.4) with 10 μm 6-cyano-7-nitroquinoxaline-2,3-dione (Tocris, Ellisville, MO, USA) and 50 μm D(-)-2-amino-5-phosphonovaleric acid (Tocris) or in low-Ca2+ Tyrode's solution (119 mm NaCl, 2.5 mm KCl, 0.1 mm CaCl2, 4 mm MgCl2, 25 mm HEPES and 30 mm glucose, pH 7.4) with 10 μm 6-cyano-7-nitroquinoxaline-2,3-dione and 50 μm D(-)-2-amino-5-phosphonovaleric acid were placed on a heated stage (set at 37 °C) with a home-prepared acrylic box to prevent temperature fluctuation. Electrical field stimulation (1 ms duration, 400 stimuli, 40 Hz) was applied by two parallel platinum wires (between wires approximately 6 mm and approximately 1 mm distance from cells; Sigma-Aldrich, Tokyo, Japan) that were mounted in a plastic lid (Gärtner & Staiger, 2002). The mCherry-OMP dynamics were imaged at intervals of 3 s for 50 min with 3 min interval electrical stimulation of 40 Hz for 10 s. After time-lapse imaging, changes of G-CaMP6 fluorescence intensity elucidated by the same electrical stimulation were measured at approximately 3 Hz at the same axonal region. For time-lapse imaging in low-Ca2+ Tyrode's solution, the G-CaMP6 measurements were performed both before and after replacing the Tyrode's solution with normal Ca2+ concentration.
We set the excitation laser power to be minimal but sufficient to obtain images with enough dynamic range. During imaging periods, reduction of mitochondrial mobility and impairment of mitochondrial morphology or distribution were not observed (De Vos & Sheetz, 2007).
Image analysis and quantification were performed by using ImageJ (NIH, Bethesda, MD, USA) and custom-written software (Visual Studio; Microsoft, Seattle, WA, USA). For all images, the average background pixel intensity of the individual image was subtracted before image processing.
In mCherry-OMP, EGFP-VAMP2, FM1-43(Δ) and APP-mCherry images, puncta were identified as local fluorescence increases, which were > 0.15 μm2 and three times higher fluorescence intensities than the background fluorescence of nearby axonal regions without obvious fluorescence clusters. The fluorescence intensities of EGFP-VAMP2 and FM1-43(Δ) puncta were measured as the sums of pixel intensities in 8 × 8 pixel (approximately 1.2 × 1.2 μm2) rectangular regions manually centered on individual puncta after the subtraction of background fluorescence of nearby axonal regions. To combine separate sets of experiments, puncta fluorescence intensities were normalised by an average fluorescence intensity of all puncta in the same axonal region. When mCherry-OMP puncta overlapped with EGFP-VAMP2 puncta by at least one pixel, we defined mitochondria localised near presynaptic sites.
Images taken at intervals of 30 min and 1 day were aligned by using ImageJ plugin Stackreg (Thévenaz et al., 1998). Even if the mitochondrial morphology changed, mitochondria were defined as stationary when their images between consecutive frames mostly overlapped. A disappearance rate of stationary mitochondria can be written as
where P(t) is a position survival rate (the fraction of mitochondria that remained at their initial positions; Fig. 1C) at day t (or at t min for time-lapse imaging for 3 h), τ is a time constant and A is an offset that indicates a rate of stable mitochondria on time scales of several days. From this equation we obtain the following
where P(1) = 100 − mobile fraction.
In this report we defined a mobile fraction as a fraction of mitochondria in mobile state at the time point of initial observation. Simply, a mobile fraction can be estimated by subtracting the mitochondria lost in the second time frame from the initial population [100 − P(30)] (in time-lapse experiments with a total observation time of 3 h, the second image was taken at t = 30 min). However, the mitochondria population that was in stationary state at t = 0 min and started to move during the 30 min interval should be estimated and further subtracted. The fraction of mitochondria that started to move during the first interval should be similar to that during the second interval, which can be calculated from the actual experimental data (the second term in Eqn (3)). In summary, the mobile fraction can be calculated as follows
where P(t) is position survival rate at t min.
The properties of mobile mitochondria and APP-containing vesicles were analysed by the method introduced by De Vos & Sheetz (2007) with some modifications. To analyse the transport of mitochondria and APP-containing vesicles, axons were manually straightened by using ImageJ plugin (Kocsis et al., 1991). To present mobile mitochondria clearly, time-lapse images were averaged and this intensity-averaged template was subtracted from each image and then Gaussian filters were applied. Centroids of puncta were measured from time-lapse images, and inter-frame velocities were calculated. In order to determine the average velocity of mitochondria and APP-containing vesicles, it is necessary to define the time period of pause of objects and exclude these time points from the calculation of average velocities. We first defined the objects in a state of pause from the data of time-lapse imaging. Mitochondria and APP-containing vesicles that were obviously immobile during imaging periods (20 mitochondria in 10 experiments and 20 APP-containing vesicles in nine experiments) were manually selected and their inter-frame velocities were calculated. Most of these mitochondria (99.4%) showed inter-frame velocities of < 0.1 μm/s and most of these APP-containing vesicles (99.0%) showed inter-frame velocities of < 0.25 μm/s. Therefore, mitochondria and APP-containing vesicles were defined to be in pause when an inter-frame velocity was below 0.1 or 0.25 μm/s, respectively. From these considerations, we calculated the average velocities of mitochondria and APP-containing vesicles as averages of inter-frame velocities excluding the time points defined to be in pause. When reinitiation of moving occurred, the pause was defined as short pause (3 s ≤ pause duration < 30 min for mitochondria; 1 s ≤ pause duration < 10 min for APP-containing vesicles; the majority of pauses were less than a few minutes). In the other cases, when mitochondria were stationary through the imaging periods, mitochondria were defined to be in stationary state (long pause). Mitochondria and APP-containing vesicles that moved over 10 μm within an imaging period were used for the analysis of dynamic properties in mobile state.
To examine a positional specificity of mitochondrial short pauses, random short-pause positions were made by a stochastic simulation. For the simulation, the total short-pause number and moving distance of individual mitochondria, presynaptic distributions and sizes of moving mitochondria obtained experimentally were used. Distances between respective short-pause positions were set over 1 μm and calculations were repeated 500 000 times for each mitochondrion. The expected means and SDs of the short-pause number near presynaptic sites were calculated. The short-pause position preference is expressed as
where Nexp and Nsim are the average numbers of short pauses near presynaptic sites obtained from experiments (Nexp) and simulation (Nsim); SDsim is the SD of the expected average number of short pauses near presynaptic sites. Higher values of short-pause position preference indicate that mitochondrial short pause occurred preferentially near presynaptic sites. The short-pause position preference was not high in the specific axonal region and was not dependent on the short-pause rates or sizes of moving mitochondria. Short-pause position preference was not significantly changed by alternation of the scale of spatial resolution for the simulation. APP-containing vesicles were used for a cargo control and stationary mitochondria localised away from synaptic sites were used for a positional control.
In order to integrate the information about the properties of mobile mitochondria and probability of transition between stationary and mobile states, it is necessary to convert the parameters linked to individual mitochondria into parameters of events that take place per unit length of axons. For this purpose, we re-examined the number of short pauses, appearance (transition from mobile to stationary states) and disappearance (transition from stationary to mobile states) events of stationary mitochondria per 100 μm segment of the axon. Axonal short-pause rates were defined as the number of short-pause events per 100 μm length of the axon per 30 min of time-lapse imaging. Axonal appearance rates were defined as the number of mitochondria that appeared within 30 min and existed for at least the next 30 min. Axonal disappearance rates were defined as the number of mitochondria that were observed at 0 min and disappeared between the next 30 and 60 min.
The intracellular Ca2+ changes induced by electrical stimulation were estimated as ΔF/F0 [=(F−F0)/F0], where F was the G-CaMP6 fluorescence intensity at a given time point and F0 was the fluorescence signal at resting state measured from 10 frames before stimulation. The ΔF/F0 of 10 consecutive images were averaged. To combine separate sets of measurements, time-averaged ΔF/F0 during electrical stimulation were normalised by the maximum value in the same axonal region (normalised time-averaged ΔF/F0).