Elsevier

Nano Communication Networks

Volume 2, Issues 2–3, June–September 2011, Pages 99-105
Nano Communication Networks

An associative learning experiment using the plasmodium of Physarum polycephalum

https://doi.org/10.1016/j.nancom.2011.05.002Get rights and content

Abstract

The plasmodium of Physarum polycephalum is a unicellular and multinuclear giant amoeba that shows adaptive behaviors. To test the presence of memory and learning ability in the plasmodium, we performed an associative learning experiment using the unicellular organism. The plasmodium in this experiment seemed to acquire a reversed thermotactic property, a new preference for the lower temperature. The result implied a possibility of unicellular learning, though in a preliminary way. We also discuss a possible mechanism of learning by the organism.

Introduction

In this paper, we discuss the possible presence of memory and learning ability in the plasmodium of Physarum polycephalum. Physarum plasmodium is a unicellular and multinuclear giant amoeba, which is visible to the naked eye (Fig. 1). The plasmodium is formed by aggregation and fusion of a massive number of uninucleate amoebae at one stage in the life cycle of Physarum polycephalum. The cell body of the plasmodium roughly consists of two parts; a sheet-like locomotive front and a tubular rear part (Fig. 1). The amorphous sheet rigorously crawls on plane surfaces searching for food sources, at the same time the tubes connect all the parts of the cell body so as not to separate the unicellular body. The plasmodium shows some adaptive behaviors in its search for food sources and quest for survival, and recent studies are revealing the organism’s computational, memory and learning abilities. The plasmodium is thus attracting a lot of attention not only from a biological standpoint but in a variety of scientific fields such as physics, information science, systems science and cognitive science as well.

The recent studies on the computational ability of the plasmodium revealed that the organism or the experimental system using it can find a solution in a maze [16], [17], [12], optimize a network between multiple nodes [15], [13], [23], and solve some other graph theoretical problems [22], [24]. The solutions for these problems are given by the arrangement of tubes. For example, the maze-solving by the amoeba is performed by selection of the shorter tubes connecting two food sources at the entrance and the exit of the maze, after full search of the maze space [16]. For these tubular dynamics, some models have already been proposed [26], [27], [29], [28], [5], [6], [18]. Furthermore, by utilizing the locomotive and tactic activities of the sheets, the plasmodium can be used in some logical operations [1], [2].

Additionally, the plasmodium is not merely a wet central performance unit, but also seems to have a reservoir for memories. Saigusa et al. showed that the plasmodium can be synchronized with the intervals of rhythmic stimuli, and is able to keep the record for more than 10 h [21]. Furthermore, we demonstrated in our previous study that the cell motility of the plasmodium depends on its locomotive history [25]. Though the two studies above did not fully demonstrate the presence of memory in the organism, the possibility was strongly implied.

In this study, we performed an associative learning experiment to test the presence of memory and learning ability in the plasmodium. Associative learning is a type of learning concerning the relationship between two different stimuli. The famous Pavlovian conditioning experiment gives a good example of associative learning (Fig. 2). In this example, the conditioned dog that has performed associative learning shows the same response to one stimulus as to the other stimulus. This demonstrates that the dog has memory and learning ability. In a similar way, we performed an associative learning experiment using the plasmodium, to test the presence of memory and learning ability in the organism.

We performed the experiment utilizing the thermotactic property of the plasmodium. In the temperature range around 10–30 °C, the plasmodium shows thermotaxis to the higher temperature. In the setup illustrated in Fig. 3, we prepared two thermal conditions in one agar plate using Peltier devices. The thermotaxis of the plasmodium is so strong and the organism without learning always proceeds to the higher temperature region. However, the food sources in this setup are located only in the lower temperature region. Though the plasmodium has the thermotaxis to the higher temperature, after the full search for the higher temperature region, it moves to the lower temperature region and is fed for the first time there. The plasmodium is forced to feed itself only in the lower temperature, and here the lower temperature is a conditioned stimulus for this experiment, and the food source is an unconditioned one. In such a way we tested the associative learning ability of the plasmodium.

Section snippets

Culture of the plasmodia

We cultivated the plasmodia using the method of Camp [3]. Briefly, glass Petri dishes were tightly arranged in a plastic box. Wet paper towels were laid out on the glass dishes, and the plasmodia were cultured on the towels. The space below the towels was filled with tap water to keep the moisture. The temperature was kept at 23 °C, and oatmeal was fed daily.

The setup for the associative learning experiment

Two Peltier devices (combination of SPE-UC-100 and TKG-311-230, SAKAGUCHI E.H VOC CORP., Japan), elements which have abilities of

Results

Using the setup illustrated in Fig. 3, we performed an associative learning experiment. The plasmodium set in the experimental condition for a week gradually acquired a new tendency to move to the lower temperature, that is, the thermotactic property of the plasmodium was reversed by the experiment (Fig. 5). In Fig. 5, the plasmodium in day 1–3 clearly showed the unconditioned thermotaxis to the higher temperature region, however, in day 4 it showed a tendency to move to the lower temperature

Discussion

In this study, we tried an associative learning experiment using the plasmodium of Physarum polycephalum. In the experiment, low temperature stimulus was used as a conditioned stimulus, and food sources were used as an unconditioned stimulus. The plasmodium was exposed to these stimuli, and as a result of the 1 week exposure, the plasmodium acquired a new tendency to move to the lower temperature, altering the original thermotaxis. In such a way we indicated the possibility of associative

Acknowledgment

This work was partly supported by Grant-in-Aid for JSPS fellows, 19-8731.

Tomohiro Shirakawa received his Ph.D. in 2007 from Kobe University, Hyogo, Japan from the Department of Earth and Planetary Systems Science. He is presently working as a Research Associate in Department of Computer Science, School of Electrical and Computer Engineering, National Defense Academy of Japan. Prior to the National Defense Academy, he worked for the Department of Computational Intelligence and Systems Science, Tokyo Institute of Technology, for three years as a Postdoctoral Fellow of

References (29)

  • M. Aono et al.

    Spontaneous deadlock breaking on amoeba-based neurocomputer

    Biosystems

    (2007)
  • M. Aono et al.

    Amoeba-based neurocomputing with chaotic dynamics

    Commun. ACM

    (2007)
  • W.G. Camp

    A method of cultivating myxomycete plasmodia

    Bull. Torrey Bot. Club

    (1936)
  • C.T. Fernando et al.

    Molecular circuits for associative learning in single-celled organisms

    J. R. Soc. Interface

    (2009)
  • Cited by (51)

    • Magnitude-sensitivity: rethinking decision-making

      2022, Trends in Cognitive Sciences
    • Living systems are smarter bots: Slime mold semiosis versus AI symbol manipulation

      2021, BioSystems
      Citation Excerpt :

      Moreover, as P. polycephalum's use of bioelectric oscillators may be compared to the ionic signaling in neurons (Ridgway and Durham, 1976; Levin et al., 2006; Boussard et al., 2021) it may be considered an appropriate basic model of animal cognition. In a classical conditioning experiment, Shirakawa, et al. (2011) trained P. polycephalum to unlearn its association of warmth with food and learn a new association of cold with food. In the experiment, a low-temperature stimulus was used as a conditioned stimulus and a food source was used as an unconditioned stimulus.

    • Learning in single cell organisms

      2021, Biochemical and Biophysical Research Communications
      Citation Excerpt :

      These examples represent to date the most convincing cases of associative learning in single cell organisms but could only be partially reproduced recently [192]. In slime mold associative learning was investigated only once [193] and the results, as mentioned by the authors themselves, could find an alternative explanation. In the experiment the food was paired with low temperature, which usually hinders migration.

    • Slime mould: The fundamental mechanisms of biological cognition

      2018, BioSystems
      Citation Excerpt :

      It is easy to find several studies on minimal examples of cognition among prokaryotes (Müller et al., 2001) but the amoeboid organisms of class Myxogastria (Myxomycetes) are the best example of the power of such cognitive studies, because they show us the connection bridge between unicellular (Physarum) to pluricellular (Dictyostelium) living systems. Thus, Physarum polycephalum shows to be an extremely intelligent system which offers unique ways for the understanding of the emergence of complex behaviours and cognitive strategies (Dussutour et al., 2010; Shirakawa et al., 2011; Mori and Koaze, 2013). The mapping of such minimal cognitive elements in slime mould also makes the analysis of two different but fundamental ideas possible: first, to identify the computational nature of basic cognitive processes, which in our case offer a good example of a reliable Kolmogorov–Uspensky biomachine; this biocomputational approach offers a naturalistic way to explain the intricacies of informational processing from scratch to the supervenience of proto-consciousness.

    View all citing articles on Scopus

    Tomohiro Shirakawa received his Ph.D. in 2007 from Kobe University, Hyogo, Japan from the Department of Earth and Planetary Systems Science. He is presently working as a Research Associate in Department of Computer Science, School of Electrical and Computer Engineering, National Defense Academy of Japan. Prior to the National Defense Academy, he worked for the Department of Computational Intelligence and Systems Science, Tokyo Institute of Technology, for three years as a Postdoctoral Fellow of the Japan Society for the Promotion of Science. His research interests include cell biology, living systems theory and bio-computing.

    Yukio-Pegio Gunji received his Ph.D. in 1987 from Tohoku University, Miyagi, Japan from the Department of Earth Science. He is presently working as a Professor in the Department of Earth and Planetary Science, Graduate School of Science, Kobe University. His research interests include living systems theory, bio-computing and time theory.

    Yoshihiro Miyake received his Ph.D. in 1989 from Tokyo University, Tokyo, Japan from the Department of Pharmaceutical Chemistry. He is presently working as an Associate Professor in the Department of Computational Intelligence and Systems Science, Interdisciplinary Graduate School of Science and Technology, Tokyo Institute of Technology. His research interests include communication science, cognitive science, human interface, communication robotics, nonlinear dynamical systems, biophysics and co-creation system.

    View full text