Evidence of conditioned behavior in amoebae
Considering that the amoebae that had shown a specific behavior were needed to perform further experiments, the cells were collected opening the sliding lateral pieces with the tip of a micropipette.
Set of videos intended only for didactical purposes. They are merely descriptive, to make easier the understanding of our experimental procedure and the reproducibility of our studies. Note that steps 5 to 8 are performed directly under the microscope, and we have not filmed them under the microscope for better visualization.
In summary, the experimental chamber consisted in a sliding glass structure. The sliding lateral pieces could be displaced in the longitudinal direction. This way, when the sliding pieces were closed an inner laminar flux was available in the chamber and, when they were open, the placement and collecting of the cells were possible easily. Movies showing the main experimental procedures have been deposited in figshare (10.6084/m9.figshare.8868326).
A. proteus and M. leningradensis may display some physiological variations depending on culture conditions. Before the experiments, the cells were starved for 24 h in Chalkley’s Simplified Medium (the same medium that was used in the experiments), in the absence of external stimuli. Once starved, only the cells that were strongly attached to the substrate, actively moving through it and showing a little amount of thin pseudopodia were used in the experiments.
The cells were washed in clean Chalkley’s medium and placed in the middle of the glass set-up (experimental chamber), under the central piece of cover glass and left to rest until all of them appeared to be firmly attached to the bottom of the modified glass slide. Next, the two 4 cm long cover glasses (sliding lateral pieces) were placed on the sides of the glass structure, protruding outside of the middle platform of the block and over the lateral wells (Fig. 1). After that, each well was filled using 75 ml of Chalkley’s medium, in such a way that the glass protrusion over each well is in contact with the liquid’s surface. Finally, as the Chalkley’s medium slowly filled up the experimental chamber, both lateral cover glasses had to be gently pushed towards each other until they touched the middle cover glass, completely covering the whole structure and forming a laminar flux that connected both lateral wells.
The experiments were always made with small groups of cells. For instance, in Amoeba proteus along the induction process, we analyzed a total of 180 cells that were studied in 32 different times (experimental replicates) analyzing them in groups of 4–10 cells each (number of cells per replicate). Scenario 1 was repeated 32 times. Scenarios 2 and 3 were performed 27 and 9 times, respectively.
The induction process was usually performed using around 7 cells per experiment, sometimes as few as 4 or 5 and other times as many as 9 or 10, the average being 6–8 cells. The number of cells analyzed in the scenarios depended on how many cells appeared to be conditioned in the first step, so that the number of cells per experiment is lower each time, for instance, in scenario 1, the number of cells was usually between 2 and 4. Finally, in scenarios 2 and 3 the experiments were performed using fewer amounts of cells per experiment, usually 1–3 which were the cells that migrated towards the cathode during the conditioning process.
Compared to Amoeba proteus, the Metamoeba leningradensis showed a more varied array of behaviors and shapes. These cells were also more difficult to handle, as they were more prone to strongly stick to the micropipette tips, while usually showing a weaker attachment to the glass chambers.
An electric field was applied to the first electrophoresis block, which was then conducted to the second by the two agar bridges. Direct measurements taken with a multimeter in the second block (where the cells were placed) showed that the strength of the electric current oscillated between 58.5 and 60 V (334–342 mV/mm) while the intensity values varied between 0.09 and 0.13 mA.
After 30 min of exposure, during which the cellular migration movement were recorded, the power supply was turned off and the agar bridges removed.
All the experiments where the only stimulus was an electric field were performed in an electrophoresis block that had never been in contact with any chemotactic substance.
Groups of 4 to 10 amoeba were placed in the experimental chamber. Once all the cells were attached to the glass surface, the laminar flux was stablished by gently closing the structure using the sliding cover glasses. Next, the peptide, nFMLP was introduced in the left well of the electrophoresis block. After about 1 min, the power supply was turned on and the electric field established. The process lasted for 30 min; after that, the power supply was turned off and the cells that moved towards the anode removed from the experimental chamber and placed in a Petri dish with clean Chalkley’s medium for future experiments (Scenario 1, 2 or 3).
When the cells were subjected to both stimuli at the same time (induction process), a new population response arose. This population behavior, in Amoeba proteus, showed that about half the amoebae cells migrated towards the anode (where the nFMLP peptide was placed), and approximately the other half of the cellular population migrated to the cathode (Fig. 4). On the other hand, only 39% of the Metamoeba leningradensis moved towards the anode-peptide, in the induction process (Figure S1). Accordingly, the cellular migration response under two simultaneous stimuli is notoriously different from that observed when the stimuli were separated (Figs. 3b, c and 8a), and therefore it cannot be concluded that the chemotactic gradient has a stronger influence on cell migration than the electric field.
In order to homogenize the cellular responses as much as possible, we put all the amoebae under starving conditions for at least 24 h before performing any experiments.
Once the laminar flux was created, we added 750 µl of 2 × 10−4 M nFMLP (#F3506, Sigma-Aldrich) peptide solution to the medium (75 ml) in the positive pole well of the second electrophoresis block; therefore, the peptide solution was diluted to a final concentration of 2 × 10−6M. In all our experiments, we used the same concentration of nFMLP. In order to homogenize the solution and accelerate the creation of a chemotactic gradient in the experimental chamber we carefully mixed the content of the left well until the amoebae appeared to start moving towards it. Finally, the cells behavior was recorded for 30 min.
The generation of an nFMLP peptide gradient was evaluated by the measurement of its concentration in the middle of the experimental chamber. To this end, 4μM fluorescein-tagged peptide (#F1314, Invitrogen), was loaded in the left side of the set-up (Fig. 1). Next, the central glass piece of the experimental chamber was slightly displaced and a small opening, the size of the tip of a 50–200 μL micropipette, was left between the sliding cover glasses and the central glass piece. This little separation allowed us to get samples of 60 μL from the middle part of the laminar chamber flux at 0, 2, 5, 10, 15, 20 and 30 min following the establishment of the laminar flux. Peptide concentration was calculated extrapolating the values from a standard curve with known concentration of the fluorescein-tagged peptide (Fig. 2). All measurements were duplicated and the experiment was repeated three times. Fluorescence was measured in 96 well glass bottom black plates (P96-1.5H-N, In Vitro Scientific) employing a SynergyHTX plate reader (Biotek) at Excitation/Emission wavelengths of 460/528 following standard laboratory techniques as described by Green and Sambrook36.
The motility of the cells was recorded using a digital camera attached to a SM-2T stereomicroscope. Images were acquired every 10 seconds, over a period of at least 30 min (180 frames). If a video was longer, only the first 30 min were quantified, except for the ones used for Fig. 7a. Since automated tracking software is often inaccurate37, we performed manual tracking using the TrackMate software in ImageJ (http://fiji.sc/TrackMate)38, as suggested in by Hilsenbeck et al.37 Each track corresponds to an individual amoeba.
In order to quantify and compare the directionality of cell migration towards the anode or the cathode, we computed the cosines of the angles of displacement of each amoeba17. More precisely, we calculated the cosine of the angle formed between the start and final positions of each cell. Consequently, we were able to analyze quantitatively if an amoeba moved towards the cathode (positive values of the cosine), or towards the anode (negative values). In addition, this study suggested the degree of directionality, since values closer to 1 (or to −1 in the case of the anode) indicated a very high preference towards that pole. Next, to estimate the significance of our results, we studied first if the distribution of cosines of angles came from a normal distribution, by applying the Kolmogorov-Smirnov test for single samples. Since the normality was rejected, the groups of cosines were compared in pairs by a non-parametric test, the Wilcoxon rank-sum test, and therefore, the results were depicted as median/IQ instead of as mean ± SD. Besides the p-value, we have reported the Z-statistic of the Wilcoxon rank-sum test39.
Note that the signs of the cosines from the second and third scenarios were changed to perform the respective tests with the galvanotaxis without previous induction (Fig. 3b) because the polarity of the electric field was inverted.
Researchers involved in the quantitative analysis of the cellular trajectories were never aware of what scenario each trajectory belonged to. Only when all the trajectories were quantified and processed, the researchers in charge of recording the amoeba’s movements informed the rest of the team about which trajectories belonged to each experiment or control.
Further information on research design is available in the Nature Research Reporting Summary linked to this article.
We would like to thank Dr. Andrew Goodkov from the Institute of Cytology (Russian Academy of Science) St. Petersburg, Russia, for valuable advices related to Amoeba organisms, Laura Pérez Gómez and Luis Rojo García for their assistance designing Fig. 1 and the AutoCAD 3D model, A-M Pérez Biedermann for her valuable contribution in our study, José González Romero and José Miguel Pérez Pérez from the Institute of Parasitology and Biomedicine “Lopez-Neyra” for their technical assistance. In addition, we thank María Calleja-Felipe for her valuable help in the peptide gradient experiments. This work was supported by a grant of the University of Basque Country (UPV/EHU), GIU17/066, the Basque Government grant IT974-16, and by the UPV/EHU and Basque Center of Applied Mathematics, grant US18/21”.
C.B., M.F. and J.C.-P.: performed the experiments; C.B.: performed the digitalization of trajectories; I.M. and L.M.: performed the quantitative studies; S.K. and M.M.: performed the gradient analysis; M.D.B. main funding and laboratory facilities; J.I.L.: supplementary movies; L.M., M.D.B., A.P.S., G.P.Y., J.I.L. and I.M.: designed the research mapping; all authors wrote the manuscript and agreed with its submission; I.M.D.F.: conceived, designed and directed the investigation.
All original videos obtained in the experiments can be found in figshare, with 10.6084/m9.figshare.8241284 (https://figshare.com/s/c59323fabced0c533fae). In addition, movies showing the main experimental procedures can be found in figshare with the 10.6084/m9.figshare.8868326 (https://figshare.com/articles/Set-Up_Video_Files/8868326).
All original videos obtained in the experiments can be found in figshare, with 10.6084/m9.figshare.8241284 (https://figshare.com/s/c59323fabced0c533fae). In addition, movies showing the main experimental procedures can be found in figshare with the 10.6084/m9.figshare.8868326 (https://figshare.com/articles/Set-Up_Video_Files/8868326).