SEARCH

SEARCH BY CITATION

Keywords:

  • hippocampal pyramidal neuron;
  • live-cell imaging;
  • mouse;
  • organelle;
  • presynaptic terminal

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Proper distribution of axonal mitochondria is critical for multiple neuronal functions. To understand the underlying mechanisms for population behavior, quantitative characterisation of elemental dynamics on multiple time scales is required. Here we investigated the stability and transport of axonal mitochondria using live-cell imaging of cultured mouse hippocampal neurons. We first characterised the long-term stability of stationary mitochondria. At a given moment, about 10% of the mitochondria were in a state of transport and the remaining 90% were stationary. Among these stationary mitochondria, 40% of them remained in the same position over several days. The rest of the mitochondria transited to mobile state stochastically and this process could be detected and quantitatively analysed by time-lapse imaging with intervals of 30 min. The stability of axonal mitochondria increased from 2 to 3 weeks in culture, was decreased by tetrodotoxin treatment, and was higher near synapses. Stationary mitochondria should be generated by pause of moving mitochondria and subsequent stabilisation. Therefore, we next analysed pause events of moving mitochondria by repetitive imaging at 0.3 Hz. We found that the probability of transient pause increased with field stimulation, decreased with tetrodotoxin treatment, and was higher near synapses. Finally, by combining parameters obtained from time-lapse imaging with different time scales, we could estimate transition rates between different mitochondrial states. The analyses suggested specific developmental regulation in the probability of paused mitochondria to transit into stationary state. These findings indicate that multiple mitochondrial behaviors, especially those regulated by neuronal activity and synapse location, determine their distribution in the axon.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The elaborate structure of the neuron requires a regulatory mechanism to allocate a sufficient number of organelles to its subcellular compartments, such as the soma, neurites and synapses. Proper distribution of the mitochondria is critical for multiple neuronal functions including energy production, calcium homeostasis, apoptosis, synaptic transmission and plasticity (Chang & Reynolds, 2006; MacAskill & Kittler, 2010). Impaired mitochondrial distribution has been linked to neurodegenerative disorders (Chen & Chan, 2009). Recent studies have identified a number of signaling pathways and key molecules that regulate mitochondrial trafficking and retention in the axon (Goldstein et al., 2008; Sheng & Cai, 2012). However, the underlying mechanism for maintaining proper axonal mitochondrial distribution is largely unknown.

Mitochondrial distribution is thought to be correlated with a spatial pattern of metabolic demands. Axonal mitochondria are enriched at presynaptic sites, nodes of Ranvier and the axon initial segments (Hollenbeck & Saxton, 2005). The recycling of synaptic vesicles (SVs) requires energy derived from ATP hydrolysis (Harris et al., 2012) and mitochondria near the presynaptic sites are thought to help this process (Kang et al., 2008; Ma et al., 2009). However, roughly half of presynaptic release sites are without nearby mitochondria (Shepherd & Harris, 1998; Chang et al., 2006) and mitochondria may change their positions with time and may be recruited to a subset of presynaptic sites that undergo active vesicle recycling. Mitochondria are bidirectionally transported along the axonal cytoskeleton and anchored at specific positions. Therefore, the distribution processes should be dependent on multiple dynamic factors involving fractions of mitochondria in stationary or mobile state, transition rates between these two states, and the dynamic properties of mobile mitochondria (Fig. 1A and B).

image

Figure 1. Mitochondrial dynamics in the axon. (A) Axonal mitochondria are classified into two dynamic states, mobile and stationary. Mobile mitochondria show saltatory movement, including moving state and short pauses (temporary stops). Short and long pause (stationary state) are distinguished by the length of pause duration. (B) Under distribution processes, the transition rates among moving mitochondria, mitochondria in short pause and in stationary state should be regulated. (C, D) Schematic representation of the time-lapse imaging. We constructed kymographs and analysed the state transition of axonal mitochondria. We determined the rate of transition from stationary to mobile state [SS[RIGHTWARDS ARROW]M] and from mobile to stationary state [M[RIGHTWARDS ARROW]SS] by intermediate and low-frequency imaging (C). We also measured the rate of mitochondrial short pauses from time-lapse images at high frequency (D).

Download figure to PowerPoint

Axonal mitochondrial transport is regulated by the intracellular and mitochondrial matrix Ca2+ concentration (Wang & Schwarz, 2009; Chang et al., 2011). The number of moving axonal mitochondria is also regulated by neuronal activity (Chang et al., 2006). However, whether the stop and start of mitochondrial movement are regulated by local cellular conditions, especially those associated with high ATP consumption at synaptic sites, has not been investigated. How changes in the characteristics of mitochondrial transport are related to the rearrangement of mitochondrial distribution also remains unclear. Although the signaling pathways and molecules involved in mitochondrial docking have been investigated, how transitions between mobile and stationary state are regulated in response to changes in physiological conditions is unknown (Wagner et al., 2003; Chada & Hollenbeck, 2004; Kang et al., 2008; Chen et al., 2009).

In this study, we analysed the dynamics of axonal mitochondria in cultured hippocampal neurons using live-cell imaging. We demonstrated that both the turnover of stationary mitochondria and behavior of mobile mitochondria were regulated by proximity to synaptic sites, neuronal activity, and maturity of axons. These results indicate that mitochondrial distribution is regulated by multiple dynamic parameters in response to physiological demands.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Molecular constructs

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 neuronal cultures and gene expression

All animal experiments performed in this study were carefully reviewed and approved by the animal welfare ethics committee at the University of Tokyo. Experiments were carried out in accordance with the Guidelines laid down by the NIH in the USA regarding the care and use of animals for experimental procedures. Pregnant ICR mice (SLC, Shizuoka, Japan) were briefly anesthetised with ether, and then killed by cervical dislocation. The preparation of hippocampal cultures from 17-day-old embryonic mice has been described previously (Okabe et al., 1999). The transfection of hippocampal neurons was performed by a Ca2+-phosphate transfection method at 5–7 days in vitro (DIV; Jiang & Chen, 2006).

Immunocytochemistry

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).

FM1-43 staining

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).

Imaging

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).

Characteristics of mitochondrial dynamics

We classified the axonal mitochondria into two dynamic states, stationary and mobile (Fig. 1A). We defined mitochondria that remained for ≧ 30 min at the same axonal region as stationary states (SS), and others as mobile states. Mobile mitochondria showed saltatory movement, including moving periods (M) and short pauses (SP) (temporary stops). The definition of a short pause is given in the following section.

Image analysis

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

  • display math(1)

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

  • display math(2)

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 = 30 min). However, the mitochondria population that was in stationary state at = 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

  • display math(3)

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

  • display math(4)

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 [=(FF0)/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).

Statistical analysis

Data are presented as means ± SE. Statistical significance was determined by performing an unpaired t-test for comparing two samples, Z-test for examining the distribution bias of short-pause position preference and Pearson's chi-square test for assessing a difference between paired observations on two variables. All statistical analysis was performed using Origin (Light Stone, Tokyo, Japan).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Experimental design

Quantitative imaging analyses of mitochondrial dynamics and its relation to presynaptic sites need reliable fluorescence-based markers of these two structures. To visualise axonal mitochondria in cultured hippocampal neurons, we expressed the C-terminal transmembrane region of mitochondrial outer membrane protein of 25 kDa tagged with mCherry (mCherry-OMP; Nemoto & De Camilli, 1999; Song et al., 2009). Neurons expressing mCherry-OMP were stained by anti-cytochrome c, a mitochondrial marker, and their co-localisation was confirmed (Fig. 2A). An average length of axonal mCherry-OMP was 1.7 ± 0.1 μm at 19–21 DIV (eight cells, n = 127), which was consistent with the mitochondrial length of rat pyramidal neurons (Shepherd & Harris, 1998). We concluded that mCherry-OMP can be used for a mitochondrial marker in cultured hippocampal neurons. To visualise the positions of presynaptic structures, VAMP2, an abundant SV protein (Takamori et al., 2006), tagged with EGFP (EGFP-VAMP2) was expressed in cultured hippocampal neurons. EGFP-VAMP2 puncta showed reasonable co-localisation with functional presynaptic sites revealed by the uptake of styryl dye FM1-43 (Fig. 2B). The fluorescence intensities of EGFP-VAMP2 puncta and the extent of FM1-43 uptake correlated well [12–13 DIV (2 weeks), n = 118 puncta from three cells, = 0.94; 19–23 DIV (3 weeks), n = 140 puncta from three cells, = 0.85; Fig. 2C]. The extent of FM1-43 uptake did not differ between control neurons and neurons transfected with EGFP-VAMP2 [EGFP-VAMP2 positive neurons/all neurons = 1.00 ± 0.05 (at 12–13 DIV, 241 puncta) and 0.99 ± 0.04 (at 19–23 DIV, 263 puncta)]. These results suggest that EGFP-VAMP2 can be used as a marker of presynaptic sites and also that their fluorescence intensity can be used as an estimate of the presynaptic total SV pool size.

image

Figure 2. Imaging of axonal mitochondria and presynaptic sites with fluorescent protein-tagged OMP and VAMP2. (A) An axonal segment from a cultured hippocampal neuron expressing EGFP and mCherry-OMP immunostained with anti-cytochrome c antibody at 18 DIV. (B) An axonal segment from a cultured hippocampal neuron expressing EGFP-VAMP2 with high-K+-induced uptake of styryl dye FM1-43 at 19 DIV. Scale bar, 10 μm. (C) Correlation of EGFP-VAMP2 puncta fluorescence with the extent of FM1-43 uptake at the same positions. a.u., arbitrary unit.

Download figure to PowerPoint

After the establishment of reliable markers for both axonal mitochondria and presynaptic sites, we designed live imaging analyses with different sampling frequencies and total imaging duration. The final goal of this study was to provide a comprehensive description of mitochondrial behavior in the axon. Individual mitochondria in the axon changed their state with time (Fig. 1A). Moving mitochondria showed frequent pauses, but most pauses were transient and paused mitochondria restarted within seconds to minutes. A small fraction of mitochondria remained stationary for a prolonged period (over hours and days) and this transition from mobile to stationary state was important in the generation of a large population of stationary mitochondria in the axon. Therefore, the imaging experiments should provide data sufficient to determine the transition rates among moving mitochondria ([M]) and mitochondria in short pause ([SP]) and stationary state ([SS]) (Fig. 1B). An ideal imaging experiment monitors the entire process of state transitions of individual mitochondria with high sampling frequencies and long imaging durations. However, this is not practical with currently available fluorescence probes and the sensitivity of image detection devices because of photobleaching and phototoxicity. Instead, we first determined the rate of transition from stationary to mobile states by intermediate and low-frequency imaging (experimental design in Fig. 1C, actual data presented in Figs 3 and 4). Next, we measured the rate of mitochondria pauses from time-lapse images at high frequency (experimental design in Fig. 1D, actual data presented in Figs 5-7). Finally, these quantitative measures were combined and the rate of transitions from short pause to stationary states was estimated (Fig. 8).

image

Figure 3. The stability of axonal mitochondria was regulated by neuronal maturation, proximity to synaptic sites and neuronal activity. Cultured hippocampal neurons expressing EGFP-VAMP2 (A1-C1) and mCherry-OMP (A2-C2) were imaged at intervals of 30 min for 3 h at 13 DIV (A), 19 DIV (B) and 20 DIV with TTX treatment (C). Arrowheads mark the appearance (red) and disappearance (white) of axonal mitochondria during 3 h. Scale bars, 10 μm. (D–F) Percentage of remaining mitochondria after initial identification at t = 0 min [position survival rates; P(t)]. Mitochondria were classified by their proximity to presynaptic sites (synaptic or non-synaptic). Samples were analysed at 2 weeks (D) [12–13 DIV, 3482 (1405 for synaptic) mitochondria from n = 8 experiments], 3 weeks (E) [19–20 DIV, 4052 (2565 for synaptic) mitochondria from n = 7 experiments] and 3 weeks with TTX (F) [20 DIV, 3297 (2010 for synaptic) mitochondria from n = 7 experiments]. (G) Schematic representation of the analytical methods for position survival rate. (H) Amount changes of position survival rate from = 30 min to 180 min [Δ(P(30)−P(180))]. (I) Estimated fraction of mobile mitochondria at t = 0 min (mobile fraction). Error bars represent SEMs. **< 0.01, ***< 0.001, unpaired t-test.

Download figure to PowerPoint

image

Figure 4. Long-term imaging of axonal mitochondria. (A) Cultured hippocampal neurons expressing mCherry-OMP and EGFP-VAMP2 were imaged with intervals of 1 day for 4 days from 19 DIV. Scale bar, 10 μm. (B) Position survival rates of the time-lapse imaging for 4 days [day 1 = 18–19 DIV, 1267 (713 for synaptic) mitochondria from n = 8 experiments]. (C) Average fluorescence intensities of EGFP-VAMP2 puncta that existed for 4 days as a function of the total or maximum consecutive days in which mitochondria localised near the given EGFP-VAMP2 puncta. Error bars represent SEMs. a.u., arbitrary unit.

Download figure to PowerPoint

image

Figure 5. Neuronal maturation and activity affected axonal mitochondrial transport in a cargo-specific manner. Cultured hippocampal neurons expressing mCherry-OMP and EGFP-VAMP2 were imaged at intervals of 3 s for 20–30 min (2 weeks, n = 38 Antero, n = 29 Retro from 11 cells; 3 weeks, n = 22 Antero, n = 19 Retro from eight cells; 2 weeks with TTX, n = 44 Antero, n = 58 Retro from 12 cells; 3 weeks with TTX, n = 48 Antero, n = 43 Retro from 10 cells). (A) Average velocities of mitochondria at different maturation stages with or without TTX treatment. An average velocity is defined as an average of inter-frame velocities that are higher than 0.1 μm/s. (B) Short-pause rates (number of short pauses per unit length of transport) of mitochondria at different maturation stages with or without TTX treatment. A short pause is defined as an event with inter-frame velocities < 0.1 μm/s and duration of more than 3 s. Cultured hippocampal neurons expressing APP-mCherry and EGFP-VAMP2 were imaged at intervals of 1 s for 10 min [2 weeks (12–13 DIV), n = 53 Antero, n = 32 Retro from seven cells; 3 weeks (19–20 DIV), n = 76 Antero, n = 48 Retro from eight cells; 3 weeks (19–20 DIV) with TTX treatment, n = 78 Antero, n = 49 Retro from eight cells]. (C) Average velocities of APP-containing vesicles at different maturation stages with or without TTX treatment. An average velocity is defined as an average of inter-frame velocities that are higher than 0.25 μm/s. (D) Short-pause rates of APP-containing vesicles at different maturation stages with or without TTX treatment. A short pause is defined as an event with inter-frame velocities < 0.25 μm/s and duration of more than 1 s. The definition of short-pause rates was as in the case of mitochondria. Error bars represent SEMs. *< 0.05, **< 0.01, ***< 0.001; unpaired t-test.

Download figure to PowerPoint

image

Figure 6. Anterogradely moving mitochondria paused temporarily near presynaptic sites. (A) Kymographs of a neuron expressing mCherry-OMP and EGFP-VAMP2 at 20 DIV (left panel). To eliminate background and noise, time-averaged images of mCherry-OMP were subtracted from individual images and Gaussian filters were applied. The inter-frame velocity of transporting mitochondria changed with time (right panel). Asterisks indicate short-pause position. Scale bars, 10 μm and 15 s. (B) Short-pause position preference (deviation of short-pause events near presynaptic positions from random events; see 'Materials and methods'). Higher values indicate more preferential occurrence of mitochondrial pauses near presynaptic sites. The short-pause position preference of mitochondria and APP-containing vesicles was calculated at 3 weeks in vitro. APP-containing vesicles were used for a cargo control and stationary mitochondria localised away from presynaptic sites [stationary mitochondria (synapse (-))] were used for a positional control. *< 0.05, ***< 0.001, Z-test.

Download figure to PowerPoint

image

Figure 7. Axonal mitochondrial transport was regulated by neuronal activity on time scales of seconds. Cultured hippocampal neurons expressing mCherry-OMP and G-CaMP6 were imaged in Tyrode's solution with D(-)-2-amino-5-phosphonovaleric acid (APV) and 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) at intervals of 3 s for 50 min (19–20 DIV, seven cells). (A) Electrical field stimulation at 40 Hz for 10 s was applied every 3 min. Induction of neuronal activities was confirmed retrospectively by elevation of G-CaMP6 fluorescence intensities quantified as ΔF/F0. Temporal deviations of ΔF/F0 were averaged during stimulation periods (time-averaged ΔF/F0). (B) Time-lapse images of axonal mitochondria by mCherry-OMP fluorescence. Red lines indicate time points of stimulus initiation. To eliminate background and noise, images were processed as in Fig. 6A (also C, H and I). (C) Control mitochondrial images obtained at time points between electrical stimulations of the same region. (D) Velocities averaged over four frames before vs. after electrical stimulation. (E) Average velocities of mitochondria before and after electrical stimulations in both transport directions. (F) Time-averaged ΔF/F0 was normalised by the maximum value in the same axonal region. Normalised time-averaged ΔF/F0 data (average of all pixel values that the mitochondrial centroid passed) were plotted against subtracted values of average velocities (‘after stimulation’ – ‘before stimulation’). Cultured hippocampal neurons expressing mCherry-OMP and G-CaMP6 were imaged in low-Ca2+ Tyrode's solution with APV and CNQX with intervals of 3 s for 50 min (20 DIV, seven cells). (G) An obvious increase of G-CaMP6 fluorescence intensity was not observed with low-Ca2+ Tyrode's solution. Efficient firing of neurons evoked by electrical stimulation was confirmed retrospectively by stimulating identical neurons in Tyrode's solution with normal Ca2+ concentration. Time-lapse images of axonal regions presented in G taken during the time period with stimulations (H) and without stimulations (I). Red lines indicate time points of stimulus initiation. (J) Velocities averaged over four frames before vs. after electrical stimulation in low-Ca2+ Tyrode's solution. (K) Average velocities of mitochondria before and after electrical stimulation in both transport directions in low-Ca2+ Tyrode's solution. Scale bars, 10 μm. Error bars represent SEMs. **< 0.01, ***< 0.001, unpaired t-test. a.u., arbitrary unit.

Download figure to PowerPoint

image

Figure 8. State-transition properties of axonal mitochondria. Using the same experimental data from Figs 3 and 5, axonal short-pause rates, axonal appearance rates and axonal disappearance rates were measured. Assuming that a small number of temporarily paused mitochondria are stabilised and transit to stationary state, we estimated the stabilisation rate ([SP[RIGHTWARDS ARROW]SS]) using the axonal short-pause rate ([M[LEFT RIGHT ARROW]SP], calculated from pauses of mitochondria at synapses without pre-existing stationary mitochondria) and axonal appearance rate ([M[RIGHTWARDS ARROW]SS]). Error bars represent SEMs.

Download figure to PowerPoint

Rate of transition to mobile state estimated by mitochondrial imaging over hours

To analyse the stability [rate of transitions from stationary to mobile states ([SS[RIGHTWARDS ARROW]M]); Fig. 1C] of axonal mitochondria on time scales of several hours, cultured hippocampal neurons expressing mCherry-OMP and EGFP-VAMP2 were imaged at intervals of 30 min for 3 h. Neurons at 12–13 DIV (2 weeks, 3482 mitochondria from n = 8 experiments) and 19–20 DIV (3 weeks, 4052 mitochondria from n = 7 experiments) were compared to examine the relationship between the maturity of neurons and stability of mitochondria (Fig. 3A and B). Fractions of synapses that contained mitochondria at t = 0 min were calculated (2 weeks, 43.2 ± 1.8%; 3 weeks, 56.9 ± 2.6%). Although the fraction was similar to previous studies (Shepherd & Harris, 1998; Chang et al., 2006), we found a developmental shift of mitochondrial distribution to synapses. The majority of axonal mitochondria were stationary for 3 h at both developmental stages, with a small number of appearance (red arrowheads) and disappearance (white arrowheads) events.

The mitochondrial population identified at = 0 min progressively changed their positions with time. The fraction of mitochondria that remained at their initial positions was calculated as a position survival rate P(t) (see 'Materials and methods'). To examine the relationship between the proximity to presynaptic sites and mitochondrial dynamics, P(t) was measured from mitochondria near presynaptic sites (synaptic) and also away from presynaptic sites (non-synaptic; Fig. 3D and E). Because mitochondria found at = 0 min included both stationary and mobile mitochondria, Δ(P(0) − P(180)) was not an appropriate estimate of mitochondria that started to move during the 180 min observation period. Instead, we used Δ(P(30) − P(180)) as an index of the transition from stationary to mobile state (Fig. 3G and H). Using this index, we found that synaptic mitochondria were less likely to restart translocation than non-synaptic mitochondria at both developmental stages (2 weeks, t14 = 4.32, < 0.001; 3 weeks, t12 = 3.57, = 0.004; unpaired t-test; Fig. 3H). Both synaptic and non-synaptic mitochondria were less likely to transit to mobile state at 3 weeks than at 2 weeks (all, t13 = 9.65, < 0.001; synaptic, t13 = 8.05, < 0.001; non-synaptic, t13 = 4.89, < 0.001; unpaired t-test; Fig. 3H). The treatment of neurons at 20 DIV with the sodium channel blocker TTX increased the transition probability to mobile state (3 weeks + TTX, 3297 mitochondria from n = 7 experiments; all, t12 = 4.72, < 0.001; unpaired t-test; Fig. 3C,E,F and H). This effect was present in both synaptic and non-synaptic mitochondria (synaptic, t12 = 3.95, = 0.002; non-synaptic, t12 = 3.88, = 0.002; unpaired t-test; Fig. 3H). These results suggest that neuronal maturation, proximity to synaptic sites and neuronal activity affect the stability of stationary mitochondria in the axon.

We estimated the fraction of mobile mitochondria at = 0 min [mobile fraction; calculated from P(t) at = 0, 30 and 60 min; see Eqn (3) in 'Materials and methods'] (Fig. 3G). The mobile fraction at 3 weeks was smaller than at 2 weeks (t13 = 4.98, < 0.001; unpaired t-test; Fig. 3I) and at 3 weeks with TTX (t12 = 3.82, = 0.002; unpaired t-test; Fig. 3I). These results suggest that the ratio of mobile to stationary mitochondria in the axon was dependent on neuronal maturation and activity.

Duration of stationary state estimated by daily imaging of mitochondrial position

In time-lapse imaging over 3 h, the majority of axonal mitochondria imaged at the initial time point remained stationary throughout the experiments (Fig. 3D–F), suggesting that the duration of stationary state is usually longer than several hours. In order to estimate the pause duration of the entire mitochondrial population, we performed imaging at 1 day intervals over 4 days using cultured hippocampal neurons expressing mCherry-OMP and EGFP-VAMP2 starting from 18 or 19 DIV [1267 (713 for synaptic) mitochondria from n = 8 experiments; Fig. 4A]. Neurons were without any obvious damage to the axonal, mitochondrial and synaptic morphology during observation. Although the mitochondrial distribution was rearranged, the density of mitochondria did not change (normalised by 4 day average: day 1, 99.0 ± 1.4%; day 2, 97.4 ± 5.9%; day 3, 101.0 ± 1.4%; day 4, 102.6 ± 1.4%, eight experiments). This further supported the absence of damage to the imaged neurons. With a longer imaging duration, the rearrangement of mitochondrial distribution increased (Fig. 4A). To quantify the long-term stability of axonal mitochondria, we measured P(t) in the same way as we did in the time-lapse imaging for 3 h (Fig. 4B). Synaptic mitochondria again showed higher stability than non-synaptic mitochondria. P(t) was fitted by the single exponential decay equation (Eqn (2) in 'Materials and methods'). By this curve fitting, we could obtain both the time constant for P(t) decrease and an offset value (Table 1). An offset indicates the size of a mitochondrial fraction immobile on time scales of several days.

Table 1. Estimated pause duration of stationary mitochondria in the axon
 ▵((P(30)-P(180)) (%)Mobile fraction (%)τ (days)A (%)
  1. Experimental data are from time-lapse imaging for 3 h. Position survival rates from time-lapse imaging for 4 days (Fig. 4B) were fitted by the single exponential decay equation (Eqn (2) in 'Materials and methods').

Experimental results
 All4.9 ± 0.73.0 ± 0.4  
 Synaptic3.0 ± 0.51.7 ± 0.4  
 Non-synaptic8.4 ± 1.45.3 ± 0.8  
Fitting results
 All4.5 ± 0.1 1.3 ± 0.135.7 ± 0.1
 Synaptic2.4 ± 0.2 2.4 ± 0.142.5 ± 1.3
 Non-synaptic7.6 ± 0.3 1.0 ± 0.117.6 ± 0.5

The time constants and offsets that we obtained by curve fitting should be consistent with the results from the time-lapse imaging for 3 h. We used the time constants and offsets to calculate estimated Δ(P(30) − P(180)) and compared them with the experimentally obtained Δ(P(30) − P(180)) (Table 1). All three estimated Δ(P(30) − P(180)) matched reasonably well with the actual data from time-lapse imaging for 3 h. Although statistically insignificant, there was a small tendency for the estimated Δ(P(30) − P(180)) to be smaller than the experimental data for all conditions. This may reflect the reappearance of mitochondria at the same position within a day (Fig. 4A, arrowheads), which causes underestimation of the P(t) decrease with time. We therefore concluded that 57% of synaptic mitochondria were considered to be ‘potentially mobile’ with an expected duration of prolonged pause of approximately 2.4 days. The remaining 42% of synaptic mitochondria were immobile on time scales of several days. The expected duration of stationary mitochondria that were localised near synaptic sites (approximately 2.4 days) was twofold longer than that of non-synaptic mitochondria (approximately 1.0 days in 78% of total non-synaptic mitochondria).

To determine whether the stability of synaptic mitochondria was related to the size of nearby synapses, the relationships between the fluorescence intensities of EGFP-VAMP2 puncta and mitochondrial localisation frequency near synaptic sites were examined (Fig. 4C). Only presynaptic sites that existed for 4 days were analysed and the total or maximum consecutive number of days in which mitochondria were co-localised was examined. Stationary mitochondria near presynapses with higher EGFP-VAMP2 fluorescence intensity showed higher stability. Because the EGFP-VAMP2 fluorescence intensity is likely to reflect the number of SVs present in presynaptic sites (Fig. 2C), the number of SVs may influence the stability of nearby stationary mitochondria.

Analysis of mitochondria in mobile state by high-frequency imaging

Our time-lapse imaging experiments with low (intervals of 1 day) and intermediate (intervals of 30 min) frequencies were useful for detecting transition between stationary and mobile states, but they did not provide information about the behavior of single mitochondria in mobile state. To analyse the switch between move and pause of mitochondria and their velocities, cultured hippocampal neurons expressing mCherry-OMP and EGFP-VAMP2 at 12–14 DIV (2 weeks) and 19–21 DIV (3 weeks) were imaged at intervals of 3 s for 20–30 min [2 weeks, n = 38 anterogradely moving mitochondria (Antero), n = 29 retrogradely moving mitochondria (Retro) from 11 cells; 3 weeks, n = 22 Antero, n = 19 Retro from eight cells; 2 weeks with TTX, n = 44 Antero, n = 58 Retro from 12 cells; 3 weeks with TTX, n = 48 Antero, n = 43 Retro from 10 cells; Figs 1D, and 5A and B]. Mitochondria were tracked as particles and inter-frame velocities were calculated. Mobile mitochondria showed saltatory movement, including moving periods and short pauses (temporary stops). Mobile mitochondria were defined to be in pause when an inter-frame velocity was below 0.1 μm/s. A short pause was defined as a pause duration of ≧ 3 s and reinitiation of transport during the observation period. An average velocity was defined as an average of inter-frame velocities after the exclusion of short-pause events (see 'Materials and methods').

The average velocities of mobile mitochondria were higher at 2 weeks than at 3 weeks (Antero, t58 = 3.33, = 0.002; Retro, t46 = 4.37, < 0.001; unpaired t-test; Fig. 5A), but this difference disappeared with TTX treatment (Antero, t90 = 0.36, = 0.72; Retro, t99 = 1.26, = 0.21; unpaired t-test; Fig. 5A). With TTX treatment, the average velocities at 3 weeks increased in both transport directions (Antero, t68 = 4.69, < 0.001; Retro, t60 = 5.65, < 0.001; unpaired t-test; Fig. 5A). Short-pause rates were defined as the number of short-pause events per transported length of individual mitochondria. Most of the pause events had short durations and detection of transition events from mobile to stationary state was practically impossible. The short-pause rate was decreased in the presence of TTX treatment at 3 weeks (Antero, t68 = 4.11, < 0.001; Retro, t60 = 4.37, < 0.001; unpaired t-test; Fig. 5B). The effect of TTX on average velocities (2 weeks, t85 = 3.02, = 0.003; unpaired t-test; Fig. 5A) and short-pause rates (2 weeks, t83 = 4.97, < 0.001; unpaired t-test; Fig. 5B) for retrogradely moving mitochondria was similar at 2 and 3 weeks. The TTX effects for anterogradely moving mitochondria showed similar tendencies at both 2 and 3 weeks, but were statistically significant only at 3 weeks (average velocity at 2 weeks, t80 = 1.52, = 0.13; short-pause rate at 2 weeks, t77 = 1.59, = 0.12; unpaired t-test; Fig. 5A and B).

Next, we asked whether the developmental changes and effects of TTX treatment on average velocities and short-pause rates were cargo specific. Membrane organelles positive for amyloid precursor protein (APP) are also known to be transported by kinesin-1, which mediates anterograde transport of axonal mitochondria (Kamal et al., 2000; Hirokawa et al., 2010). Therefore, we compared the behavior of mCherry-OMP-positive mitochondria with that of APP-mCherry-positive membrane organelles. Cultured hippocampal neurons expressing APP-mCherry and EGFP-VAMP2 were imaged at intervals of 1 s for 10 min [2 weeks (12–13 DIV), n = 53 Antero, n = 32 Retro from seven cells; 3 weeks (19–20 DIV), n = 76 Antero, n = 48 Retro from eight cells; 3 weeks (19–20 DIV) with TTX treatment, n = 78 Antero, n = 49 Retro from eight cells]. A short pause of APP-containing vesicles was defined as an event with inter-frame velocities < 0.25 μm/s and duration of more than 1 s, together with the occurrence of restart during observation periods. An average velocity was calculated using the same method as we used for mitochondria. Consistent with the previous work, APP-containing vesicles moved faster in the anterograde direction than in the retrograde direction (Fig. 5C) (Kaether et al., 2000). The transport of APP-containing vesicles showed properties that were different from those of mitochondrial transport. Both the average velocities and short-pause rates of APP-containing vesicles were similar at 2 and 3 weeks after plating (average velocity: Antero, t127 = 1.14, = 0.26; Retro, t78 = 1.34, = 0.19; short-pause rate: Antero, t127 = 0.79, = 0.43; Retro, t78 = 0.46, = 0.65; unpaired t-test; Fig. 5C and D). In addition, TTX did not affect the transport of APP-containing vesicles (average velocity: Antero, t152 = 0.66, = 0.51; Retro, t95 = 0.09, = 0.92; short-pause rate: Antero, t152 = 0.28, = 0.78; Retro, t95 = 0.34, = 0.73; unpaired t-test; Fig. 5C and D). These results indicate that the regulation of organelle transport by neuronal maturation and activity is cargo specific.

Short pause of anterogradely moving mitochondria near presynaptic sites

High-frequency time-lapse imaging revealed developmental regulation of mitochondrial transport in the axon (Fig. 5). In the presence of TTX, the short-pause rates of mobile mitochondria were reduced, suggesting the involvement of axonal excitability and associated events in the regulation of mitochondrial short pause. Many mitochondrial short pauses occurred near presynaptic sites [number of synaptic short pauses/number of all short pauses = 67 ± 6% (Antero) and 44 ± 5% (Retro); Fig. 6A]. However, even if mitochondrial short pause occurred randomly, short pauses near presynaptic sites could be observed by chance, due to the high density of presynaptic sites. To critically evaluate whether short pauses of mitochondria preferentially occur near presynaptic sites, experimental data were compared with values generated by a stochastic simulation. We defined the parameter ‘short-pause position preference’, which indicates how preferentially mitochondrial short pauses occurred near presynaptic positions compared with random events (Eqn (4) in 'Materials and methods'). Higher values of the short-pause position preference indicate that mitochondrial short pauses occurred more preferentially near presynaptic sites. APP-containing vesicles were used as a cargo control and stationary mitochondria localised away from presynaptic sites were used as a positional control. The short-pause position preferences for each condition at 3 weeks are summarised in Fig. 6B. Anterogradely moving mitochondria showed significantly high values of the short-pause position preference at synaptic sites (= 4.13, < 0.001; Z-test). Additionally, retrogradely moving APP-containing vesicles with TTX showed preferential short pause near synapses (= 2.24, = 0.03; Z-test).

In order to examine a relationship between short-pause events and synaptic properties, presynapses were grouped into those with higher total fluorescence intensities of EGFP-VAMP2 (possibly containing more SVs; Fig. 2C) and those with lower intensities (containing less SVs). Anterogradely moving mitochondria preferentially stopped temporarily near the positions of synapses with more SVs (inline image = 7.99, = 0.005; Pearson's chi-square test; Table 2), but this preference of anterogradely moving mitochondria was attenuated by TTX application (inline image = 1.85, = 0.17; Pearson's chi-square test; Table 2). However, retrogradely moving mitochondria showed a higher tendency towards temporal stop near synapses with more SVs in the presence of TTX (inline image = 10.92, = 0.001; Pearson's chi-square test; Table 2). These seemingly opposite tendencies may indicate that the regulation of mitochondrial preferential pause at larger synapses may differ between anterograde and retrograde transport.

Table 2. Relationships between short-pause frequencies near presynaptic sites and total SV pool size
 Normalised EGFP-VAMP2 (F) (a.u.)N (synapse)N (pass)N (short pause)Short-pause frequencyχ2 P
  1. Presynapses were grouped into those with higher total fluorescence intensities of EGFP-VAMP2 (possibly containing more SVs) and those with lower intensities (containing less SVs). Short-pause frequencies (number of paused mitochondria/number of all passed mitochondria) of two groups were compared. Pearson's chi-square test was used for statistical analysis. a.u., arbitrary unit.

OMP
 Antero≥ 13483540.657.990.005
< 154112500.45
 Retro≥ 13372400.560.980.33
< 14784400.48
OMP + TTX
 Antero≥ 151224640.291.850.17
< 181349820.22
 Retro≥ 146193800.4110.920.001
< 175282760.27

Acute regulation of mitochondrial transport by neuronal activity

Chronic TTX treatment decreased the short-pause rates of axonal mitochondria (Fig. 5B), suggesting that neuronal activity regulates the transport of axonal mitochondria. To gain further insight into the acute regulation of mitochondria transport by neuronal activity, axonal mitochondria were imaged under the application of electrical stimulation. Cultured hippocampal neurons expressing mCherry-OMP and G-CaMP6 (Ohkura et al., 2012) were imaged in Tyrode's solution with the N-methyl-d-aspartate receptor blocker D(-)-2-amino-5-phosphonovaleric acid and the AMPA receptor blocker 6-cyano-7-nitroquinoxaline-2,3-dione, which were added to prevent glutamate toxicity under electrical stimulation (Antero, n = 110 mitochondria; Retro, n = 120 mitochondria from seven cells; Fig. 7A–F). Live cells were placed on a heated stage and imaged at intervals of 3 s for 50 min. Electrical field stimulations of 40 Hz for 10 s were applied every 3 min. The induction of neural activities was confirmed by the elevation of G-CaMP6 fluorescence intensity quantified as ΔF/F0 (Fig. 7A). The mitochondrial dynamics in four frames before and after stimulation was compared (Fig. 7B). The mitochondrial dynamics at time points between electrical stimulations was analysed as a control (Fig. 7C). When average velocities before stimulation were < 0.1 μm/s, those mitochondria were excluded from the analysis. Although not all mitochondrial velocities were changed by electrical stimulations, average velocities were decreased by electrical stimulations in both transport directions (Antero, t86 = 2.98, = 0.004; Retro, t120 = 3.71, < 0.001; unpaired t-test; Fig. 7D and E). Short-pause frequencies (number of paused mitochondria/number of all passed mitochondria) were increased by electrical stimulation in both transport directions (Antero, inline image = 4.79, = 0.03; Retro, inline image = 8.45, = 0.004; Pearson's chi-square test; Table 3). These results clearly show that neuronal activity acutely regulates mitochondrial transport on the order of seconds.

Table 3. Changes of short-pause frequencies induced by electrical stimulation in Tyrode's solution with normal Ca2+ concentration and in low-Ca2+ Tyrode's solution
 N (mito)N (short pause, pre)N (short pause, post)χ2 P
  1. Short-pause frequencies (number of paused mitochondria/number of all mitochondria) calculated from four image frames before and after stimulation were compared. N (mito) indicates analysed mitochondria and pre/post indicate before/after electrical stimulation, respectively. Pearson's chi-square test was used for statistical analysis.

Normal Ca2+
 Antero
Control5617170.001
Stimulation4412224.790.03
 Retro
Control5921271.260.26
Stimulation6120368.450.004
Low Ca2+
 Antero
Control6711130.200.65
Stimulation7110172.240.13
 Retro
Control4213110.230.63
Stimulation4516170.050.83

Mitochondrial transport is regulated by the intracellular and mitochondrial matrix Ca2+ concentration (Wang & Schwarz, 2009; Chang et al., 2011). To examine whether changes of mitochondrial transport elicited by electrical stimulation were related to intracellular Ca2+ elevation, the variation of average velocities was compared with normalised time-averaged ΔF/F0 (Fig. 7F). However, there were no obvious correlations. To confirm whether Ca2+ signaling is involved in changes of mitochondrial transport elicited by electrical stimulation, neurons were imaged in low-Ca2+ Tyrode's solution with D(-)-2-amino-5-phosphonovaleric acid and 6-cyano-7-nitroquinoxaline-2,3-dione (Antero, n = 138 mitochondria; Retro, n = 87 mitochondria from seven experiments; Fig. 7G–K). The efficient firing of neurons evoked by electrical stimulation was confirmed retrospectively by stimulating identical neurons in Tyrode's solution with normal Ca2+ concentration (Fig. 7G). In low-Ca2+ Tyrode's solution, the average velocities were not changed by electrical stimulation in both transport directions (Antero, t140 = 0.16, = 0.87; Retro, t88 = 0.44, = 0.66; unpaired t-test; Fig. 7H–K). Short-pause frequencies were also not changed in both transport directions (Antero, inline image = 2.24, = 0.13; Retro, inline image = 0.05, = 0.83; Pearson's chi-square test; Table 3). These results support the idea that Ca2+ signaling is important for the activity-dependent regulation of mitochondrial transport in the axon.

Quantitative analysis of state transitions

The goal of this study was to provide a comprehensive description of mitochondrial behavior in the axon (Fig. 1). We measured the rate of transition from stationary to mobile states ([SS[RIGHTWARDS ARROW]M]) (Figs 3 and 4). The rate of transition between short pauses and moving states ([M[LEFT RIGHT ARROW]SP]) is presented in Fig. 5. Due to a low rate of transitions to stationary states and long duration of stationary periods, imaging of the entire stabilisation process ([M[RIGHTWARDS ARROW]SP[RIGHTWARDS ARROW]SS]) was not practical. By assuming that a small number of mitochondria in [SP] are stabilised and transit to [SS], we estimated the stabilisation rates ([SP[RIGHTWARDS ARROW]SS]) from the short-pause rates ([M[LEFT RIGHT ARROW]SP]) and appearance rates of stationary mitochondria ([M[RIGHTWARDS ARROW]SS]). To combine these two separate experimental data, event frequencies should be normalised by the unit length of the axon (axonal short-pause rates, axonal appearance and disappearance rates; see 'Materials and methods'; Fig. 8). The axonal appearance and disappearance rates were measured from the same experimental data shown in Fig. 3 (Fig. 1C). The short-pause rate of individual mitochondria was suppressed by TTX treatment at 3 weeks (Fig. 5B). However, the axonal short-pause rate was not changed by TTX treatment because the number of mobile mitochondria was increased by TTX treatment (Figs 3I and 8). By using these normalised rates, we could calculate the stabilisation rates at different conditions ([SP[RIGHTWARDS ARROW]SS]; Fig. 8). The stabilisation rate near synapses ([SP[RIGHTWARDS ARROW]SS]synaptic) declined significantly from 2 to 3 weeks (1.01 vs. 0.53%) and was modulated by TTX treatment. Because stabilisation rates away from synapses ([SP[RIGHTWARDS ARROW]SS]non-synaptic) were less affected by culture periods and TTX treatment, regulation of the stabilisation rate near synapses is likely to be the parameter that is important for the control of mitochondrial replacement along the axon. Although the axonal appearance rate of mitochondria near synapses ([M[RIGHTWARDS ARROW]SS]synaptic) was more than twofold higher at 2 weeks, this increase was counterbalanced by the comparable rate of disappearance ([SS[RIGHTWARDS ARROW]M]synaptic). It is likely that there exists a mechanism that keeps the balance between [M[RIGHTWARDS ARROW]SS] and [SS[RIGHTWARDS ARROW]M], as these rates were maintained in parallel in all experimental conditions (Fig. 8). This regulation may be important to keep the density of both synaptic and non-synaptic mitochondria constant with time.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We report here the dynamic properties of axonal mitochondria using live-cell imaging with multiple sampling frequencies ranging from seconds to days. High-frequency image sampling is necessary to trace the accurate positions of mobile mitochondria, transported by motor proteins with their velocity of 0.1–1.4 μm/s (De Vos & Sheetz, 2007; MacAskill & Kittler, 2010). In turn, the probability of transitions between stationary and mobile states is low (a few events per hour within an image area; Fig. 8) and time-lapse imaging with longer durations is required. Here we performed time-lapse imaging with high (intervals of 3 s), intermediate (intervals of 30 min) and low (intervals of 1 day) frequencies. Our results demonstrated that mitochondrial dynamics on multiple time scales differ between developmental stages and are regulated by neuronal activity and proximity to synaptic sites.

To understand the dynamics of axonal mitochondrial distribution, mitochondrial properties in mobile and stationary states, and the transition process between them should be examined (Fig. 1). Our analyses revealed that the properties of stationary mitochondria are highly regulated by neuronal maturation and activity. The transition probability from stationary to mobile state of axonal mitochondria decreased with neuronal maturation and increased with TTX treatment at 3 weeks in vitro (Fig. 3). Previously, Chang et al. (2006) reported that residence times of axonal mitochondria were not changed by TTX treatment at 14–15 DIV. The effect of TTX may be dependent on neuronal maturation, as we observed that axonal mitochondria at 2 weeks showed a lower response to TTX than those at 3 weeks (Fig. 5A). In addition to neuronal maturation and activity, mitochondrial stability was regulated by proximity to synapses (Figs 3 and 4). The expected duration of mitochondrial pause near synaptic sites (approximately 2.4 days) was twofold longer than that of non-synaptic mitochondria (approximately 1.0 days). Furthermore, mitochondria near presynaptic sites with a higher number of SVs were more stable (Fig. 4C). SV recycling involves numerous ATP-consuming steps and may require stationary mitochondria (Vos et al., 2010; Harris et al., 2012; Sheng & Cai, 2012). This interpretation is consistent with the idea that mitochondria are preferentially localised and stabilised near positions with high energy demands (Hollenbeck & Saxton, 2005). The number of SVs at a bouton and the volume of the bouton show a good correlation (Shepherd & Harris, 1998). Therefore, there is a possibility that the effects of bouton size on mitochondrial dynamics might be simply related to steric constraints imposed by larger boutons, e.g. a higher probability of interaction between moving mitochondria and the cytoskeletal meshwork that anchors SVs.

Although synapses with high activity of SV recycling require stationary mitochondria, about half of presynaptic sites are without nearby mitochondria (40–60% in our culture) (Shepherd & Harris, 1998; Chang et al., 2006). How is ATP supplied to presynaptic sites without nearby mitochondria? We can speculate on two possible mechanisms. One is by diffusion from distant stationary mitochondria and the other is by mobile mitochondria passing the active presynaptic sites. Electrical field stimulation decreased the average velocity and increased short-pause frequencies in both transport directions within seconds (Fig. 7E and Table 3). This indicates that the mitochondrial transport machinery may have an ability to respond to physiological demands such as SV recycling and associated ATP hydrolysis.

The molecular mechanisms of mitochondrial transport have been intensively investigated (Goldstein et al., 2008; Sheng & Cai, 2012). Intracellular and mitochondrial matrix Ca2+ is a key regulator of mitochondrial transport (Wang & Schwarz, 2009; Zhang et al., 2010; Chang et al., 2011). In low-Ca2+ Tyrode's solution, electrical stimulation failed to induce the down-regulation of mitochondrial mobility (Fig. 7K and Table 3), suggesting the importance of Ca2+ signaling for the activity-dependent regulation of mitochondrial transport. However, both previous studies (Chada & Hollenbeck, 2004; Zhang et al., 2010) and the results presented here support the idea that the regulation of mitochondrial transport involves factors other than Ca2+ signaling. Because a subset of mitochondria did not respond to electrical stimulation, they may lack regulatory machinery sensitive to Ca2+ signaling (Fig. 7B and D). The absence of an obvious relationship between changes in mitochondrial transport by electrical stimulation and intracellular Ca2+ elevation (Fig. 7F) also supports the presence of a signaling system other than Ca2+. In addition to Ca2+ signaling, our data indicate that the presence of a presynaptic structure regulates the short-pause rate of anterogradely moving mitochondria (Fig. 6). This specificity cannot be explained by regulatory mechanisms independent of the cargo–motor complex, such as post-translational modifications of tubulin or obstacles on microtubule tracks (Verhey et al., 2011). Further identification of signaling molecules involved in functions of the cargo–motor complex is required.

To clarify the influence of neuronal activity on mitochondrial distribution, we estimated the transition rate from short pauses to stationary states near and away from synapses with or without TTX (stabilisation rate; Fig. 8). The stabilisation rates were up-regulated by TTX at 3 weeks in culture and this increase was prominent near synapses. This indicates that paused mitochondria are more likely to enter stationary state when neurons do not fire. In contrast, the short-pause rate of mitochondria was increased within seconds by field stimulation (Table 3), suggesting that moving mitochondria are more likely to stop in phase of spike bursts. These opposite influences of axonal firing on mitochondria may be coordinated in specific situations. For example, if neurons show burst-spiking activities with intervening resting periods, spike bursts can elicit short pauses of moving mitochondria and subsequent resting periods can stabilise them, leading to enhanced placement of mitochondria close to synapses. Hippocampal CA1 pyramidal neurons generate high-frequency bursts both in vivo and in vitro (Kandel & Spencer, 1961; Wong & Prince, 1978; Epsztein et al., 2011) and it may be possible to speculate that these bursts facilitate the synaptic localisation of mitochondria. Other mechanisms should be present in the developmental transition of mitochondrial distribution along axons and the biological significance of spike bursts in mitochondrial redistribution should be validated by further experiments.

In summary, our time-lapse imaging revealed axonal mitochondrial dynamics, which were spatiotemporally regulated by neuronal maturation, neuronal activity and synaptic positions. Proper distribution of mitochondria, which is important for neuronal development, functions and diseases, should be achieved by these multiple parameters and the underlying mechanisms should be clarified in future.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We thank Dr Sakurai (Juntendo University) for the hAPP695-Venus plasmid and Dr Nakai (Saitama University) for the G-CaMP6 plasmid. We thank for Ms Sato, Dr Ebihara and Dr Urushido for cell culture, Ms Sato and Ms Morimoto for plasmid construction, and Dr Ito-Ishida for helpful comments on this manuscript. This work was supported by Grants-in-Aid for Scientific Research (21220008 to S.O.) from the Ministry of Education, Culture, Sports, Science and Technology of Japan and a part of this study is the result of ‘Development of biomarker candidates for social behavior’ carried out under the Strategic Research Program for Brain Sciences by the Ministry of Education, Culture, Sports, Science and Technology of Japan (S.O.). K.O. was supported by the Graduate Program for Leaders in Life Innovation. The authors declare no conflict of interest.

Abbreviations
Antero

anterogradely moving mitochondria/APP-containing vesicles

APP amyloid precursor protein

DIV

days in vitro

EGFP enhanced green fluorescent protein

[M] moving periods/mobile state

OMP

C-terminal transmembrane region of mitochondrial outer membrane protein of 25 kDa

Retro

retrogradely moving mitochondria/APP-containing vesicles

[SP] short-pause [SS] stationary state

SV

synaptic vesicle

TTX

tetrodotoxin

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References