Three-dimensional structure of axonal mitochondria reflects the age of drosophila.

This study aimed to reconstruct a three-dimensional map of axonal mitochondria using Fiji and Neurolucida software, and to observe directly the morphology and distribution of mitochondria in axons of motor neurons in dorsal longitudinal flight muscles of drosophila aged 5 days and 20 days, using electron microscopy. Results indicated that there was no difference in the total area and volume of mitochondria between 5-day-old drosophila and 20-day-old drosophila in all sections, but the ratio of mitochondrial total areas to axon total areas, as well as mitochondrial density of 20-day-old drosophila, was lower than that of 5-day-old drosophila. The number of mitochondria, whose volume was less than 1 000 000 μm(3), and between 1 000 000 μm(3) and 10 000 000 μm(3), was higher in 20-day-old drosophila than that in 5-day-old drosophila. The number of mitochondria with a volume between 1 000 000 μm(3) and 100 000 000 μm(3) was apparently higher than those with a volume less than 1 000 000 μm(3) or larger than 100 000 000 μm(3). In addition, the number of mitochondria with a volume more than 100 000 000 μm(3) was small; however, the volume was nearly 70% of the total volume in both 5-day-old and 20-day-old drosophila. In contrast, the number of mitochondria with a volume between 1 000 000 μm(3) and 10 000 000 μm(3) was large, but the volume was less than 30% of the total volume. These experimental findings suggest that changes in mitochondrial morphology and number in motor neurons from the dorsal longitudinal muscle of drosophila are present during different ages.

Research Highlights

(1) The present study aimed to understand the changes in axonal mitochondria, the imbalance of mitochondrial fission and fusion in neurodegenerative diseases, alterations in mitochondrial morphology, number, volume and distribution, as well as neuronal apoptosis.

(2) Fiji and Neurolucida software was applied to reconstruct a three-dimensional map of axonal mitochondria and to observe the morphology and distribution of mitochondria in drosophila at different ages.

(3) We can infer a mitochondrial fission and fusion imbalance and a possible mechanism of age-related neurodegeneration from studying the morphology, number and size of mitochondria.

INTRODUCTION

Mitochondria are subcellular organelles bounded by a double membrane and contain their own DNA[1]. Mitochondrial function is very important for the nervous system[2], and is involved in regulating organelle dynamics, protein importation, and programmed cell death[3]. In recent years, many studies have suggested that mitochondria have a central role in age-related neurodegenerative diseases[456]. The important role of mitochondria in cellular energy production has long been accepted[7]. Mitochondria are dynamic organelles that migrate and undertake fission, fusion and branching[8]. Mitochondrial fission and fusion can facilitate metabolism, help to ensure biological function, and play important roles for mitochondrial morphology and function[9]. Proper distribution and efficient function of mitochondria is particularly important for neurons. Functional disturbance of mitochondria can cause age-related neurodegenerative disorders[10]. Fission and fusion can allow the distribution of mitochondria throughout long neuronal axons particularly motor neurons[11]. Changes to mitochondrial morphology and length during fission and fusion reflect different conditions, particularly in morbid states such as Alzheimer's disease[12]. At present, it remains to be proven whether an increase in mitochondrial size is related to functional efficiency. Therefore, we have introduced three-dimensional imaging techniques to assess mitochondrial shape, number, length and volume in drosophila axons.

Image notation has recently been developed mainly based on the digital imaging of cell morphology brought about by great advances in microcomputer hardware and software[13]. Current approaches to three-dimensional imaging at subcellular resolution usually use confocal microscopy and electron tomography[14]. Some approaches to research mitochondrial morphology are limited by low data sampling and classification of complex morphological phenotypes[15]. In order to better understand the structure of mitochondria in the axon, we planned to create three-dimensional models of axonal mitochondria, and analyze drosophila axonal mitochondrial shape, number, length and volume at different ages.

RESULTS

Quantitative analysis of experimental animals

Ten drosophila were randomly divided into two groups with five drosophila in each: 5-day-old group and 20-day-old group. All ten drosophila were involved in the final analysis.

Three-dimensional images of mitochondria in motor neurons in the dorsal longitudinal muscle in drosophila at different ages reconstructed by computer

We captured images of mitochondrial morphology (Figure 1) and also reconstructed the axon. Thus, the three-dimensional position of mitochondria relative to the axon was observed. It was possible to rotate all the figures at 360 degrees using the computer. Thus, we could observe all the structures from different angles. It was also possible to magnify or zoom out particular regions. Furthermore, the software allowed us to focus on the structure of the image and also to view the entire model (supplementary Video 1 online). These images may serve as the bases for advanced research, and may also contribute to improving our understanding of the relationship between axonal mitochondrial morphology and function. Using the three-dimensional imaging system, we revealed mitochondria of tubular morphology and many mitochondria with different lengths that were lined up as single organelles along neuronal axons (Figure 2).

Figure 1

Electron micrographs of sections of dorsal longitudinal flight muscle (DLM) axons in drosophila at ages of 5 and 20 days (× 5 000).

Black arrows indicates axons, white arrows indicates mitochondria. PSI: Peripherally synapsing interneuron.

Figure 2

Fiji and Neurolucida software constructed three-dimensional images of an axonal mitochondrion (arrows) of dosal longitudinal muscle of drosophila at ages of 5 and 10 days.

Parameters of mitochondria in motor neurons of dorsal longitudinal muscle of drosophila at different ages

There was no significant difference in mitochondrial total areas between 5-day-old drosophila and 20-day-old drosophila in all sections (P > 0.05; Figure 3A). However, the ratios of mitochondrial total areas to axon total areas of 20-day-old drosophila were lower than that of 5-day-old drosophila (P < 0.01; Figure 3B). After examining the difference of mean areas, we further investigated whether the volume of mitochondria showed differences between 5-day-old drosophila and 20-day-old drosophila. We calculated the total volume of mitochondria in axons, and then examined the difference between two groups. The results revealed that there was no statistically significant difference regarding the total volume of mitochondria (P > 0.05; Figure 3C). In contrast, the mitochondrial density of 20-day-old drosophila was smaller than that of 5-day-old drosophila (P < 0.01; Figure 3D).

Figure 3

Comparison of axonal mitochondrial parameters between 5-day-old drosophila and 20-day-old drosophila.

(A) Mitochondrial total areas; (B) ratio of mitochondria to axon; (C) total volume of mitochondria; (D) mitochondrial density. There were five drosophila at each time point. aP < 0.01, vs. 5-day-old drosophila (t-test).

Ratio of mitochondrial volume in motor neurons from the dorsal longitudinal muscle of drosophila at different ages

In mitochondria whose volume was less than 1 000 000 μm3, the number of mitochondria was smaller in 5-day-old drosophila than that in 20-day-old drosophila (Figure 4A). In mitochondria whose volume was between 1 000 000 μm3 and 10 000 000 μm3, the number of mitochondria in 20-day-old drosophila was larger than that in 5-day-old drosophila. The number of mitochondria with a volume between 1 000 000 μm3 and 100 000 000 μm3 in the two groups of drosophila was apparently larger than those with a volume less than 1 000 000 μm3 or larger than 100 000 000 μm3. We also studied the ratio of mitochondrial volume. Results found that the number of mitochondria whose volume was more than 100 000 000 μm3 was smaller; however, the volume was nearly 70% of the total volume in both 5-day-old and 20-day-old drosophila. In contrast, the number of mitochondria whose volume was between 1 000 000 μm3 and 10 000 000 μm3 was larger, but the volume was less than 30% of the total volume (Figure 4B).

Figure 4

Comparison of mitochondrial volume ratio in drosophila at different ages.

(A) The number of mitochondria with a volume between 1 000 000 μm3 and 100 000 000 μm3 was apparently more than those with a volume less than 1 000 000 μm3 or larger than 100 000 000 μm3.

(B) The volume was 70% of the total volume in mitochondria with volumes of more than 100 000 000 μm3 in both 5-day-old and 20-day-old drosophila.

There were five drosophila at each time point. The mean differences were tested using the t-test.

According to the volume, all mitochondria were divided into four groups, I: < 1 000 000 μm3; II: ≥ 1 000 000 μm3 or < 10 000 000 μm3; III: ≥ 10 000 000 μm3, or < 100 000 000 μm3; IV: ≥ 10 000 000 μm3.

DISCUSSION

Recently, studies have shown a resurgence of interest in mitochondrial biology and in the links between mitochondrial dysfunction and disease[16]. There is evidence that mitochondrial dysfunction is an early event in neurodegeneration. Recent molecular, cellular, animal models, and postmortem brain studies have revealed that mitochondrial abnormalities are key factors that cause synaptic damage and cognitive decline in Alzheimer's disease. Mitochondrial structural changes caused by amyloid beta result in increased mitochondrial fragmentation, decreased mitochondrial fusion, mitochondrial dysfunction, and synaptic damage[17]. In the present study, we first used the software of Fiji and Neurolucida to reconstruct a three-dimensional map of axonal mitochondria. Based on this method, we now know the distribution of mitochondria in axons. We can also calculate the number, area means and the volume of mitochondria. Based on the results from our data analysis, we can determine mitochondrial function at different ages and in different diseases, which could contribute to the study of the relationship between mitochondrial morphology and function.

Previous results suggest a possible role for defects in mitochondrial trafficking, fission and fusion in neurodegenerative diseases such as Parkinson's disease and Alzheimer's disease[18]. For example, in one study, researchers used immunocytochemistry and morphometry to determine whether mitochondrial abnormalities are associated with vulnerable neurons in Alzheimer's disease. Their findings showed major abnormalities in mitochondrial dynamics restricted to vulnerable neurons, suggesting a relationship between mitochondria and oxidative damage in Alzheimer's disease[19]. In addition, defects in axonal transport are early pathogenic events in a number of human neurodegenerative diseases. Age-related damage to mitochondria may amplify any primary defects to axonal transport[20]. In the present study, a deeper understanding of the prominent dynamic nature of mitochondria, characterized by a balance of fission and fusion, has aided studies on abnormal mitochondrial dynamics in neurodegenerative diseases[21]. Blocking fission can prevent apoptosis and neuronal cell death. Because mitochondrial dynamics play an important role in neurons, the ability to visualize changes in mitochondrial morphology is critical to gain a better understanding of disease mechanisms. In this study, we described a method to assess mitochondrial morphology and dynamics using three-dimensional imaging. We performed the set-up, preparation and fixation of samples and assessment of mitochondrial number- volume and density in axons. This study was made even more comprehensive with the use of three-dimensional models, which allowed us to rotate, zoom in or zoom out three-dimensional images. Thus, one can easily examine the structures from different angles. This model contributed to a new vision to investigate axonal mitochondrial morphology and quantitative analysis. Results of axonal mitochondrial shape and volume could be used in analyzing the degree of the neurodegenerative disease. Abnormal mitochondrial shape and volume are classically characterized by progressive dementia particularly in Alzheimer's disease. Furthermore, the three-dimensional models developed in this study can be used in creating virtual simulations and volumetric studies. This study may also contribute to a new vision for mitochondrial studies.

Our study involved the reconstruction of three- dimensional structures from axonal mitochondria using Fiji and Neurolucida software following electron microscopy. Our study also monitored the number, area, and volume of axonal mitochondria to further investigate the relationship between mitochondrial structure and function in neurodegenerative diseases. In summary, we established and validated a way for mitochondrial morphological and functional analysis. We further infer a specific and general relationship between mitochondrial morphology and age-related neurodegenerative diseases.

MATERIALS AND METHODS

Design

A randomized, controlled, animal experiment.

Time and setting

This study was performed at the Institute of Neuroscience and State Key Laboratory of Neuroscience, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai, China from October 2008 to March 2011.

Materials

A total of 10 drosophila were provided by the Drosophila Bloomington Stock Center, University of Indiana, USA. Drosophila stocks were cultured with standard medium at 24°C. After eclosion, the adult drosophila were cultured on standard medium and entrained into a 12-hour light/dark cycle at 28°C. W1118 female flies were used. The giant fiber system of drosophila mediates escape behavior through a small number of neurons to innervate the dorsal longitudinal flight muscles[12]. The P [Gal4]- A307 was used to drive the expression in the giant fiber system and other components of the nervous system. Drosophila were maintained in vials containing a mixture of agar, yeast, white cornmeal, molasses and water. Our experimental study used 5-day-old and 20-day-old drosophila to observe mitochondrial morphous and distribution.

Methods

Observations of axons in motor neurons using light and electron microscopy

Drosophila were anesthetized using carbon monoxide and were decapitated. The head, legs, wings and abdomen were removed[22]. The thoracic segment was fixed with 4% (w/v) paraformaldehyde[23] and 2.5% (v/v) glutaraldehyde for 48 hours at 4°C, and then washed in PBS for 10 minutes and transferred to 1% (w/v) osmium tetroxide for 2 hours[24], followed by a series of graded alcohols and embedded in Epon812[25]. Sections were polymerized for 48 hours at 60°C. The resin-embedded specimen was cut into 1-μm-thick sections[26]. These sections were stained with toluidine blue[27]. After staining, we investigated the structure of the axon of motor neurons innervating the dorsal longitudinal muscle[28] under the light microscope (Leica Microsystems Inc., Buffalo Grove, IL, USA).

Next, we used the ultramicrotomy to cut sections at 70 nm thickness[29]. The distance between sections was 70 nm. Thus, we obtained 14-μm-long axon sections. Ultrathin sections were collected on formvar-coated copper grids, were stained with uranyl acetate and lead citrate[30] and were viewed with the JEOL JEM-1230 transmission electron microscope (Leica Microsystems Inc.). After scanning of each slice, images at 5 000-fold magnification were saved.

Three-dimensional models and mitochondrial parameters

The software of Fiji and Neurolucida (MBF Bioscience, Williston, WT, USA) was used to reconstruct the three-dimensional coordinate models from serial sections[3132]. First, we imported images of 100 pieces of 70-nm-thick sequential sections into the software of Fuji. Second, we improved the local contrast of the pictures using CLAHE. Third, we aligned the 100 images and saved them as 1 stack of TIF images. We inputted the stack in the software of Neurolucida, and drew the contour of every mitochondrion and axon. Finally, the software of Neurolucida automatically formed the three-dimensional structure of axonal mitochondrion and simultaneously calculated the number-volume and surface areas of the mitochondria.

Statistical analysis

Data were analyzed with SPSS 16.0 software (SPSS, Chicago, IL, USA). The mean differences were tested using t-test and a P < 0.05 level was considered statistically significant.