The present protocol describes the whole procedure of analysis for axonal transport. In particular, the calculation of transport velocity, not including the pausing, and the visualization method using open-access software “KYMOMAKER” are shown here.
Neuronal cells are highly polarized cells that stereotypically harbor several dendrites and an axon. The length of an axon necessitates efficient bidirectional transport by motor proteins. Various reports have suggested that defects in axonal transport are associated with neurodegenerative diseases. Also, the mechanism of the coordination of multiple motor proteins has been an attractive topic. Since the axon has uni-directional microtubules, it is easier to determine which motor proteins are involved in the movement. Therefore, understanding the mechanisms underlying the transport of axonal cargo is crucial for uncovering the molecular mechanism of neurodegenerative diseases and the regulation of motor proteins. Here, we introduce the entire process of axonal transport analysis, including the culturing of mouse primary cortical neurons, transfection of plasmids encoding cargo proteins, and directional and velocity analyses without the effect of pauses. Furthermore, the open-access software “KYMOMAKER” is introduced, which enables the generation of a kymograph to highlight transport traces according to their direction and allow easier visualization of axonal transport.
Kinesin family members and cytoplasmic dynein are motor proteins that move along the microtubules in cells to transport their cargo1. Most kinesins move toward the plus end, whereas dynein moves toward the minus end of a microtubule. The functions and mechanisms of cargo transport in the neuronal axon have been investigated extensively. Because of their length, axons require stable long-distance transport to maintain the health of neurons. Defects in the transport of mitochondria, autophagosomes, and vesicles containing amyloid-β protein precursor (APP) have been reported as causes of neurodegenerative diseases2,3. Numerous in vitro investigations have revealed the mechanisms underlying coordinated transport by motor proteins, and various studies using purified motor proteins and microtubules have uncovered how motor molecules move along microtubules4,5. Typically, multiple motors are involved in a single cargo6. However, there are some models of how the opposite motors determine the direction of cargo transport. One is the "association/dissociation model"; in this, only one-directional motors are associated with the cargo during the one-directional transport. In the second, the "coordination model", both motors are attached to the same cargo, and only one side of the motor is activated. In the third "tug-of-war model", the force balance between kinesins and dyneins determines the direction of transport7,8,9. In addition, several reports have suggested that the balance and number of activated motor proteins influence the velocity of cargo transport in vitro8,10.
An unanswered question is how these activities of kinesins or dyneins are regulated in living cells. Previous reports have shown acetylation of microtubules in axons, and neurotrophic factors enhance axonal transport11,12. Also, various types of cargo and cargo adapters function as motor activators in corresponding ways13,14. Many are associated with the membrane of transport vesicles, and their functions are regulated by signals such as post-translational modifications15. Therefore, observing transport direction and velocity in living cells offers valuable insight into the molecular regulation of cargo transport in in vitro experiments. The observation of axonal transport allows distinguishing between kinesin- and dynein-based transport. Because axons harbor plus-end unidirectional microtubules16, cargo is transported anterogradely (i.e., soma to axon terminal) by kinesins and retrogradely (i.e., axon terminal to soma) by dyneins.
In the present study, a method to observe and analyze axonal transport in primary cultured neurons is described. As an example, the procedure for observing the axonal transport of membrane proteins-APP, calsyntenin-1/alcadein α (Alcα), and calsyntenin-3/alcadein β (Alcβ)-containing vesicles-is described. It is known that the anterograde transport of APP-containing vesicles is considerably faster than that of Alcα-containing vesicles, although both are transported by kinesin-117,18,19. In previous reports, several methods have been used to measure velocity. The most variable step is the handling of pauses during transport. In live cells, transport is sometimes hindered by obstacles along the microtubules; however, motors can bypass a region with or without a pause20. The calculation of velocity over a longer observation time may be affected by the pause, which can result in a slower speed estimation. Here, a method is described using movement over a segmented (200 ms) period to exclude the effect of the physical pause of motors. Finally, an open-access (Windows only) software program called "KYMOMAKER"21 is introduced. Kymographs are used widely to visualize vesicular transport and are useful for visualizing the transport direction of each cargo without requiring a movie. The software generates kymographs from time-lapse movies by applying a one-dimensional Watershed algorithm multiple times during the rotation of images. This enables the resulting kymograph to show fine structures efficiently and easily. Furthermore, KYMOMAKER automatically detects and highlights trails according to their direction and enables the creation of easy-to-understand diagrams.
The experiments were approved by the Animal Studies Committee of Hokkaido University, following the ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines. Female C57BL/6J mice (pregnant, 15.5 days) were used for the present study.
1. Preparation of mouse primary cortical cultured neurons
2. Transfection of plasmids to primary cultured neuron using the calcium phosphate method
NOTE: The mentioned volumes are for a 0.8 cm2 well in an 8-well glass-bottom chamber (see Table of Materials).
3. Observation of axonal transport
4. Image processing
5. Drawing a kymograph and detecting traces using KYMOMAKER
6. Determining the velocity
Primary cultured neurons from the E15.5 mouse cortex were cultured in a glass-bottom dish as described. As an example, APP-EGFP, Alcα-EGFP, or Alcβ-EGFP were expressed in the primary cortical neurons. APP and Alcα are known to be transported in the axon by kinesin-12,17. APP associates with kinesin-1 via the adapter protein JIP1 (JNK-interacting protein 1), while Alcα directly binds via its double W-acidic motifs. Alcβ is a member of the Alcadein family and has only one W-acidic motif24. Each protein was transfected to Div. 5 neurons, and images were acquired every 200 ms using TIRF microscopy 14-16 h following transfection. Several moving vesicles in the axons were observed in neurons transfected with either APP-EGFP, Alcα-EGFP, or Alcβ-EGFP (Supplementary Movies 1–3, respectively).
Using KYMOMAKER, kymographs for Alcβ-EGFP were generated with and without the Rotational Watershed algorithm (Figure 1C,D, respectively). Furthermore, KYMOMAKER enabled the extraction of traces in anterograde and retrograde directions from the original and Watershed kymographs (Figure 1E,F). Major traces were detected in both kymographs. Compared with the original method, applying the Watershed algorithm enabled better detection of low-intensity traces (Figure 1F).
The velocities of the vesicles in each neuron were tracked and calculated as described. The velocity of APP-EGFP was 3.05 ± 1.03 µm/s, which is consistent with previous results18,19. Alcα-EGFP was transported at a speed of 1.89 ± 0.77 µm/s, which was considerably slower than APP, consistent with the reported velocity of kinesin-1. The transport speed of Alcβ-EGFP was 2.68 ± 1.07 µm/s, which fell between the velocities of Alcα and APP. Histograms were generated using a 0.2 µm/s bin width to show differences in the distribution. Furthermore, we calculated the relative ratio of transport direction using the kymographs generated with KYMOMAKER (Figure 2A,B). All three proteins were predominantly transported in the anterograde direction (Figure 2C).
Figure 1: Workflow of the detection method using KYMOMAKER. (A) Setting the main windows for KYMOMAKER. (B) The obtained movie was opened in KYMOMAKER, and the area that did not contain the axon was removed by "trimming". (C) By clicking on generate, the original kymograph was generated. (D) Clicking on Rotational Kymograph opened a new "Rotational Kymograph" window, and detection parameters could be changed in the top dialog box in (A). (E) Each anterograde or retrograde trace was detected from the kymograph in (C). The detection parameters can be changed in the bottom dialog box in (A). (F) Traces detected from (D). Scale bars = 5 µm. Please click here to view a larger version of this figure.
Figure 2: Calculation of velocity and direction for kinesin-1 transported vesicles. APP-EGFP, Alcα-EGFP, or Alcβ-EGFP were transfected to Div. 4-6 of mouse primary cortical cultured neurons, and the images of axons were acquired after 14-18 h of transfection. (A) Kymographs of axonal transport were generated using KYMOMAKER. Left: original kymograph; middle: anterograde traces; right: retrograde traces. (B) Histograms of the velocity of anterograde transport are shown (n = 583 for APP-EGFP, n = 388 for Alcβ-EGFP, n = 264 for Alcα-EGFP from two independent experiments.). Mean ± SD is shown on the histograms. (C) The percentages of traces that were transported in each direction are shown (each bar shows mean ± SD. n = 2 from independent experiments, each containing 6-10 cells). Scale bars = 5 µm. Please click here to view a larger version of this figure.
Supplementary Figure 1: Culture density of primary cultured neuron on glass bottom dish. Differential interference contrast (DIC) image of Div. 1 mouse primary cortical cultured neurons plated, as in step 1.23. Scale bar = 50 µm. Please click here to download this File.
Supplementary Table 1: The constructions of plasmids used in this protocol. Please click here to download this File.
Supplementary Movie 1: Axonal transport of APP-EGFP. The movie was taken every 200 ms for 30 s. Scale bar = 5 µm. Please click here to download this Movie.
Supplementary Movie 2: Axonal transport of Alcα-EGFP. The movie was taken every 200 ms for 30 s. Scale bar = 5 µm. Please click here to download this Movie.
Supplementary Movie 3: Axonal transport of Alcβ -EGFP. The movie was taken every 200 ms for 30 s. Scale bar = 5 µm. Please click here to download this Movie.
An analysis method for axonal transport is described, which includes the calculation of segmented velocity and the generation of kymographs. A critical step during the transfection step is maintaining the health of cultured neurons. The transfection method described by Jiang and Chen29 was followed with minor modifications. Gentle mixing of the DNA/CaCl2 solution and 2x HBS increased the efficiency of transfection by reducing the size of the precipitations. To prevent the precipitations from damaging cells, the cells were rinsed with an acidified medium. Quick rinses (two or three times) without incubation were sufficient to remove precipitation from the cells. Here, we used the cultured primary neuron; it should be worth mentioning that transport in mice sciatic nerves can be observed by confocal microscopy in vivo with surgery26, and the recent development of multiphoton microscopy has enabled the observation of axonal transport in mice brains27.
Kymographs are used widely to visualize the velocity and direction of axonal transport. There are various tools to make kymographs, including the Fiji plugin KymographBuilder. Some of them, like KymoAnalyzer or Kymolyzer, can calculate the parameters from tracks assigned semi-manually28,29. Although the KYMOMAKER software cannot modify the assignment of traces, it has the advantage of easy visualization of traces per direction21. It manages the trimming process, generates kymographs by adjusting detection parameters, and automatically extracts traces by directions. Using the Rotational Watershed method allows the detection of fine structures. Although most traces could be detected, it was difficult to determine whether two traces were independent or temporarily disconnected, especially where traces crossed. Further, KYMOMAKER is not used for calculation of the parameter of each trace. For anterograde and retrograde trace counts, it is recommended to use the "Kymograph" window, which masks the detected traces with red lines.
This calculation method has the advantage of excluding the effect of unpredictable pauses on live cells. Pauses can be caused by physical obstacles, such as other motors and cargo, and microtubule-associated proteins, such as tau20,30. The calculation of transport velocity was performed using several procedures. Typically, longer (more than a few seconds) periods are used to determine the transported distance, including the pause. However, using segmented (200 ms) periods, this protocol can identify and remove pauses from the calculation more efficiently. In some cases, the cargo velocity showed a two- or three-peak distribution ,such as the case in Alcβ, which may reflect the different types or statuses of involved motor proteins. Thus, generating a histogram is recommended to visualize the distribution of velocities, and this more-accurate calculation method can contribute to a clearer estimation of the peak velocities.
The authors have nothing to disclose.
This work was supported by KAKENHI (22K15270, Grant-in-Aid for Young Scientists) and the Akiyama Life Science Foundation for YS. We would like to express our gratitude to Dr. Masataka Kinjo and Dr. Akira Kitamura, Laboratory of Molecular Cell Dynamics, Faculty of Advanced Life Science, Hokkaido University for providing critical input and expertise that greatly assisted the research. The observation with TIRF microscopy was performed using the instrument installed at the Laboratory of Molecular Cell Dynamics, Faculty of Advanced Life Science, Hokkaido University. The Instrument is registered in the Open Facility system managed by the Global Facility Center, Creative Research Institution, Hokkaido University (AP-100138). We thank Dr. Seiichi Uchida, Human Interface Laboratory, Department of Advanced Information Technology, Faculty of Information Sciences and Electrical Engineering, Kyushu University, Fukuoka Japan, for the assembly with the application Kymomaker. Advanced Prevention and Research Laboratory for Dementia, Graduate School of Pharmaceutical Sciences, Hokkaido University is supported by Japan Medical Leaf co., Ltd.
5-Fluoro-2-deoxyuridine | Sigma-Aldrich | F0503 | |
Apo TIRF 100x/1.49 OIL | Nikon | ||
B-27 Supplement (50x), serum free | Thermo fischer scientific | 17504044 | |
Bovine serum albumin | Wako | 013-25773 | |
CalPhos Mammalian Transfection Kit | Takara | 631312 | |
Cell strainer 40 µm Nylon | Falcon | 352340 | |
CoolSNAP HQ | Photometrics | ||
Deoxyribonuclease I | Sigma-Aldrich | DN-25 | |
D-Glucose | Wako | 041-00595 | |
DMEM/Ham’s F-12 | Wako | 042-30555 | |
Dumont No. 7 forceps | Dumont | No.7 | |
Feather surgical blade | Feather | No.11 | |
Feather surgical blade handle | Feather | No. 3 | |
Gentamicin | Wako | 079-02973 | |
Gentamicin Sulfate | Wako | 075-04913 | |
GlutaMAX Supplement | Thermo fischer scientific | 35050061 | |
HEPES | DOJINDO | 342-01375 | |
Horse Serum, heat inactivated | Thermo fischer scientific | 26050088 | |
KCl | Wako | 163-03545 | |
KH2PO4 | Wako | 169-04245 | |
KYMOMAKER | http://www.pharm.hokudai.ac.jp/shinkei/Kymomaker.html | ||
L-15 Medium (Leibovitz) | Sigma-Aldrich | L5520 | |
MetaMorph version 6.2r1 | Metamorph | ||
Na2HPO4 | Wako | 197-02865 | |
NaCl | Wako | 197-01667 | |
NaHCO3 | Wako | 191-01305 | |
Neurobasal Medium | Thermo fischer scientific | 21103049 | |
Nikon ECLIPSE TE 2000-E | Nikon | ||
Nunc Lab-Tek 8 well Chambered Coverglass | Thermo fischer scientific | 155411 | |
Papain | Worthington | LS003126 | |
Penicillin-Streptomycin (10,000 U/mL) | Thermo Fisher Scientific | 15140122 | |
Poly-L-lysine hydrobromide | Sigma-Aldrich | P2636-500MG | |
Trizma base | Merck | T1194-10PAK | solved with water to make 0.1 M Tris-HCl (pH.8.5) |
.