Plasticity of Social Behavior in Drosophila

Transcription

1 Neuroscience and Behavioral Physiology, Vol. 32, No. 4, 2002 Plasticity of Social Behavior in Drosophila N. G. Kamyshev, G. P. Smirnova, E. A. Kamysheva, O. N. Nikiforov, I. V. Parafenyuk, and V. V. Ponomarenko * Translated from Rossiiskii Fiziologicheskii Zhurnal imeni I. M. Sechenova, Vol. 86, No. 11, pp , November, Original article submitted April 28, This article presents results obtained from studies of the plasticity of changes in social behavior in Drosophila (interactions between individuals in groups) in conditions of homo- and heterogeneous environments. This is the first report of data illustrating self-starting acquisition by female Drosophila of a classical conditioned reflex to contextual factors signaling possible threats from other individuals and blocking the initiation of activity. A previously described operant conditioned reflex also helped flies avoid aggression from other individuals and make more efficient use of food resources by decreasing the initially high level of activity. Classical conditioning had the effect that the fly did not need to repeat acquisition of the conditioned reflex each time: when placed into an analogous situation, the fly s activity automatically decreased as a result of exposure to the conditioned stimulus, i.e., contextual factors. KEY WORDS: Drosophila melanogaster, learning, classical conditioned reflex, social behavior. The interaction of two or more organisms, usually of the same species, includes a mutual exchange of stimuli which regulates the triggering, maintenance, and termination of the corresponding behavioral acts, and this is termed social (community) behavior [11]. This definition includes sexual behavior. Here, however, the discussion concerns only those aspects of the interaction between individuals which are not associated with sexual behavior. When Drosophila individuals are placed in a group situation, different authors have observed strengthening [8 10, 14], suppression [12], and the absence of any changes in activity [13]. The contradictions in these data are evidence for the possible involvement of previously unknown motivations, unconditioned reflexes, and plastic changes in behavior in controlling movement activity. Our studies have demonstrated [1] that the direction of changes in activity when a group situation is created depends on whether the initial level of activity of the individuals is high or low. Initially high levels of activity decrease, while low levels increase. One of the factors * Deceased. Laboratory for Comparative Behavioral Genetics, I. P. Pavlov Institute of Physiology, 6 Makarov Bank, St. Petersburg, Russia. determining differences in the level of movement activity in individuals is the size of the experimental chamber. We interpreted increases in activity seen in flies when space is restricted as a Pavlovian freedom reflex [5]. The unconditioned reflex reaction whose appearance leads to increased activity in the group situation is escape from approaching individuals. This reaction leads to increases in the numbers of excursions made by the flies in a group. The unconditioned reaction whose appearance leads to decreased activity in the group situation is interruption of excursions in response to encountering another individual. This reaction leads to decreases in the duration of excursions made by flies in a group. Both unconditioned reactions led to a tendency of flies to avoid coming into contact with each other [9, 15] and exchanging blows [9]. The interaction of these two reactions fixes activity and the number of contacts dependent on activity at a relatively high level. There is, however, a factor responsible for time-related decreases both in the activity of flies within a group and in the frequency of encounters. From the very first seconds, the probability of the transition from activity to encounter is higher, while the probability of the transition from inactivity to encounter is lower than the probability of random movement. This creates the conditions for operant learning, whereby activity on the part of the flies leads to blows from other individuals and inactivity allows this to be /02/ $ Plenum Publishing Corporation

2 402 Kamyshev, Smirnova, Kamysheva, Nikiforov, Parafenyuk, and Ponomarenko 70 1 live and 9 dead lished data and summarizes results from studies of plastic changes accompanying social behavior of Drosophila in homo- and heterogeneous environments A P C1 C2 C3 C4 B P C1 C2 C3 C4 B B Fig. 1. The effects on the distribution of three-day Drosophila flies of the presence of nine cold-killed flies (A) on the nutrient medium and the absence of after-effects fro this procedure after removal of dead flies (B). The horizontal axes show vertical zones of the flask: P = plastic tube; C1 C4 = four zones of identical height within the flask; B = base (surface of the medium); the vertical axes show the numbers of individuals (%). The upper curve shows the percentage of individuals in each zone, which is made up of two components the percentage of immobile individuals (shaded) and the percentage of mobile individuals (unshaded zones). Each set was observed two times over a period of 30 min (one observation per fly per minute): 30 min after placing live individuals with dead flies and 30 min after removal of the dead flies. Set one consisted of 110 individuals; set 2 consisted of 102 individuals. METHODS All experiments were performed using virgin females of wild-type Canton S flies, aged three days. Flies were kept on raisin-yeast medium at 25 C and a 12:12 light regime. Experimental individuals were collected immediately after hatching without any kind of immobilization. Before experiments, flies were kept either singly or in groups of 10 individuals in glass flasks (diameter 25 mm, height 100 mm) on fresh nutrient medium without live yeast. Movement activity was recorded in flasks of the same type with or without medium, usually from 12:00 to 16:00, using a modification of the method described by Luchnikova [4]. Vessels were placed on a background of a white screen illuminated at an intensity of 150 Lx and recording was started 30 min later. The unit of observation consisted of 10 individuals kept either in a single flask or in individual flasks. The number of moving individuals in each observation unit were recorded at various time points. The index of activity (IA) for each time period and each observation unit was calculated as k m i IA = i= 1 100%, h k where h is the number of individuals in the observation unit; k is the number of observations made at the given time point; m i is number of moving individuals during the ith observation, and i = 1... k. Activity indexes calculated by this method were subjected to further statistical analysis using parametric statistical methods (see captions to Tables and Figures) without prior transformation of the data. avoided [1]. As a result, the frequency of excursions decreases over time and the duration of inactive periods increases. The acquisition of an operant conditioned reflex by flies is not accompanied by any kind of after-effect when individuals are isolated from the group [1]. These experiments were performed in conditions of a homogeneous environment flies were placed in experimental chambers made of a single material. After-effects, however, were seen if flies were kept in a heterogeneous environment before being placed in groups and activity recordings on individuals taken from groups were made in the heterogeneous environment standard flasks for breeding Drosophila, with a plastic tube containing nutrient medium on the floor [3]. This article presents previously unpub- RESULTS AND DISCUSSION In conditions of a heterogeneous environment, two activity-suppressing factors could be added to the mechanisms controlling the level of movement activity in groups. The first appeared only when the group of flies was located directly on the nutrient medium. Inhibition of activity among individuals on the medium occurred even when a group of dead individuals was placed there (Fig. 1, A; Table 1). This activity-suppressing component did not persist after individuals were transferred to another flask containing medium with no other flies (Fig. 1, B). Thus, all this unconditioned reflex activity needed was for the flies simply to perceive the presence of other individuals on the food resource. This response reflects competitive interactions for

3 Plasticity of Social Behavior in Drosophila 403 TABLE 1. Effects of the Presence of Nine Cold-Killed Flies on the Distribution of Three-Day Drosophila Individuals (A) and the Absence of After-Effects after Removal of Dead Flies (B) Zone of flask First period (A) Second period (B) Set 1 (separate) Set 2 (1 live + 9 dead) Set 1 (separate) Set 2 (separate) Immobile individuals, % P 3.6 ± ± ± ± 2.13 C ± 1.59 * 4.3 ± ± ± 1.46 C ± ± ± ± 1.51 C3 6.7 ± ± ± ± 1.65 C4 2.1 ± 0.30 * 5.3 ± ± ± 1.07 All 32.2 ± ± ± ± 2.63 B 8.2 ± 1.56 * 60.3 ± ± ± 0.49 Total 43.9 ± 2.82 * 89.8 ± ± ± 2.73 Mobile individuals, % P 6.7 ± 0.64 * 1.0 ± ± ± 0.70 C ± 1.55 * 3.7 ± ± ± 1.57 C2 8.2 ± 0.60 * 1.6 ± ± ± 0.58 C3 6.6 ± 0.46 * 1.4 ± ± ± 0.56 C4 8.6 ± 0.61 * 1.8 ± ± ± 0.64 All 47.9 ± 2.46 * 8.5 ± ± ± 2.33 B 1.5 ± 0.22 * 0.8 ± ± ± 0.25 Total 56.1 ± 2.82 * 10.3 ± ± ± 2.73 Total of mobile and immobile individuals, % P 10.3 ± 1.38 * 3.4 ± ± ± 2.12 C ± 1.77 * 8.0 ± ± ± 1.91 C ± 1.39 * 10.3 ± ± ± 1.55 C ± ± ± ± 1.60 C ± 0.68 * 7.1 ± ± ± 1.20 All 80.0 ± 1.85 * 35.6 ± ± ± 2.07 B 9.7 ± 1.56 * 61.1 ± ± ± 0.57 Total Notes. This Table presents the data illustrated in Fig. 1. For details see caption to Fig. 1. All = total percentage of individuals in all flask zones. Asterisks between two values in neighboring columns identify significant differences between them (Welch t test for sets with different dispersions, p < 0.05). food among females, based on protecting the resource without overt manifestations of aggression [6], which has not previously been described in Drosophila. The second activity-suppressing component, which was seen in flies in groups in conditions of the heterogeneous environment, resulted from the flies acquisition of a classical conditioned reflex to contextual factors, associated predominantly with the formation of an accumulation of individuals in particular sites within the flask [3]. The results of the flies acquisition of the operant conditioned reflex is the formation of passive accumulations of individuals in the homogeneous environment: in apparently random order, flies avoid activity and stay close to each other [1]. The mechanism of staying close together is that attempts to initialize activity in individuals in accumulations are blocked by anticipatory threatening actions on the part of the neighboring flies. These include preening [7] the legs are rubbed together and application of preventative blows to individuals approaching too close [1]. These stimuli, which restrain the activity of the flies, are the unconditioned stimuli for the acquisition of a classical conditioned reflex to contextual factors. The group situation in culture flasks differs from the situation in homogeneous experimental chambers in that the flies accumulate mainly

4 404 Kamyshev, Smirnova, Kamysheva, Nikiforov, Parafenyuk, and Ponomarenko (set 3) A Isolated individuals (set 3) P C1 C2 C3 C4 B P C1 C2 C3 C4 B P C1 C2 C3 C4 B B Fig. 2. Distributions of three-day Drosophila flies in flasks containing medium. One-hour observations on each of three sets were performed twice; individuals in set 2 were observed initially in groups (A) and then 30 min after isolation from the groups (B). Each set consisted of 240 individuals (data from six repeat experiments were combined). The percentages of individuals in each zone were calculated for each ten flies using the results of ten sequential observations over 1 h. For further details see caption to Fig. 1. on the medium (Fig. 2, A; Table 2) and on the tube (Fig. 3, A; Table 3). This creates the conditions for association of the unconditioned stimuli coming from other individuals with conditioned stimuli distinguishing the place at which accumulations form from other places in the flask. When a fly lands alone in a new flask and finds itself randomly in a place in which it has previously experienced the action of an unconditioned stimulus, its attempts to initiate an excursion are blocked by the action of the conditioned stimulus (Fig. 2, B; Fig. 3, B; Tables 2, 3). In various experiments, not all of whose results are presented here, we showed that 1) significant after-effects of being in a group persist in adult individuals isolated from the group for 1 3 h; 2) subsequent retention is prevented by the daily activity peak; 3) flies only need to be in a group for 15 min in order to establish a modification of movement behavior in individuals isolated from a group which is as stable as that seen after being in the group for three days (Fig. 4); 4) during early ontogenesis, on the background of a low level of movement activity, the after-effects of being kept in a group persist in imagos for up to 20 h; 5) the after-effects of being kept in a group are seen only after flies have interacted with a group of live individuals; the presence on the medium of dead individuals suppresses the flies activity, but without after-effects (Fig. 1); 6) interaction of immature individuals in a group accompanied by long-term after-effects occurs virtually in the absence of movements in the flask, which suggests a role for fine movements in the mechanisms of interaction (Fig. 3); 7) the after-effects of keeping flies in groups in flasks appear when individuals isolated from a group are moved into flasks but do not occur when flies are isolated in other experimental chambers with or without medium; 8) after isolation, flies stay only in those places in the flask where they were located previously within accumulations in immature individuals, this place is the tube (Fig. 3, B), while in mature flies this is the nutrient medium (Fig. 2, B) or (in flasks without medium) the tube. Operant and classical learning mutually supplement each other as mechanisms for controlling the level of activity and decreasing aggression in flies in groups. Acquisition of the operant conditioned reflex helps flies avoid aggression from other individuals and achieve more efficient uti-

5 Plasticity of Social Behavior in Drosophila 405 TABLE 2. Distribution of Three-Day Drosophila in Flasks Containing Medium (multiple comparisons of means) Zone of flask First period (A) Second period (B) Set 1 (separate) Set 2 (in groups) Set 3 (in groups) Set 1 (separate) Set 2 (isolated) Set 3 (in groups) Immobile individuals, % P * 21.4 C C C C All B 6.3 * * Total 55.4 * * 70.8 * 83.2 Mobile individuals, % P C * * C2 8.3 * * C3 6 * * C4 7.8 * * 1.9 All 43.3 * * 27.9 * 16.2 B Total 44.6 * * 29.2 * 16.8 Total of mobile and immobile individuals, % P C * * 21.3 C C * C * 2.9 All 88.3 * * 72.9 * 59.9 B 6.5 * * Total Notes. The data presented here are plotted in Fig. 2. Unifactorial dispersion analysis was performed for each period and each flask zone and multiple comparisons of means were made by the Tukey method for a 95% level of significance. Asterisks between two values in neighboring columns identify significant differences between them. lization of the food resource by decreasing initially high levels of activity. Classical learning has the result that flies do not need to acquire the operant reflex anew on each occasion: when they find themselves in a similar situation, the fly s activity automatically decreases because of the action of the conditioned stimulus, i.e., contextual factors. Thus, we were able to identify the following mechanisms controlling movement activity in flies in groups. 1. Unconditioned reflex reactions: escape from approaching individuals; stopping on encountering another individual; staying on the food resource in the presence of other individuals. 2. Conditioned reflexes: an operant activity-suppressing conditioned reflex which is induced by encounters with other individuals; a classical conditioned reflex to contextual factors signaling possible threats from other individuals during initiation of excursions;. Operant interaction between individuals in groups was suggested and utilized by us as a test for identifying mutants with disturbances to operant behavior [2]. A collection of P-inversion lines was used to select four lines with alterations in the acquisition of the operant conditioned reflex suppressing activity in the group situation. The results of this study supported the fact that the inability to undergo

6 406 Kamyshev, Smirnova, Kamysheva, Nikiforov, Parafenyuk, and Ponomarenko (set 3) A Isolated individuals (set 3) P C1 C2 C3 C4 B P C1 C2 C3 C4 B P C1 C2 C3 C4 B B Fig. 3. Distributions of Drosophila flies in flasks containing medium at ages 4 (A) and 9 (B) h. Individuals in set 2 were observed initially in groups (A) and then 3 h after isolation from groups (B). Each set consisted of 160 individuals. For further details see caption to Fig Fig. 4. Durations of retention of the effects of previously being kept in groups for different periods on the index of activity of Drosophila flies isolated from groups (continuous tests in flasks containing medium). The abscissa shows time after isolation from groups, h; the ordinate shows the index of activity. Experiments: 1) Three days before experiments, kept individually, tested individually (control); 2) 3 days before experiments in groups, tested in groups (control); 3) 3 days before experiments in groups, tested individually; 4) individually, 24 h in groups, tested individually; 5) individually, 3 h in groups, tested individually; 6) individually, 1 h in groups, tested individually; 7) individually, 15 min in groups, tested individually. Each experiment used 110 individuals. Dispersion analysis and multiple comparisons of means as described by Tukey with a 95% level of significance showed that experiments 3 7 showed no differences from each other but were significantly different from controls during the first 3 h of isolating individuals from groups.

7 Plasticity of Social Behavior in Drosophila 407 TABLE 3. Distribution of Drosophila in Flasks Containing Medium in Early Imaginal Ontogenesis (multiple comparisons of means) Zone of flask First period (A) Second period (B) Set 1 (separate) Set 2 (in groups) Set 3 (in groups) Set 1 (separate) Set 2 (isolated) Set 3 (in groups) Immobile individuals, % P 46.9 * * C * 12.5 C * C C All 50 * B Total * Mobile individuals, % P C C C * C All * B Total * Total of mobile and immobile individuals, % P 46.9 * * C C * C * C All 51.9 * * B Total Notes. The data presented here are plotted in Fig. 3. Unifactorial dispersion analysis was performed for each period and each flask zone and multiple comparisons of means were made by the Tukey method for a 95% level of significance. Asterisks between two values in neighboring columns identify significant differences between them. operant learning can be detected from a complex of characteristic changes in movement behavior in flies in groups. The most informative characteristics in this regard were: the absence of any reproducible decrease in the number of contacts, increases in the total time spent at rest and in the duration of individual periods of rest. Classical training of Drosophila individuals in groups is not suitable for selecting mutants because of the laboriousness of recording, but could be used for additional characterization of previously identified mutants. This study was supported by the Federal Targeted Scientific-Technological Program Priorities in Genetics. REFERENCES 1. N. G. Kamyshev, E. A. Kamysheva, G. P. Smirnova, and I. V. Parafenyuk, Mutual learning by Drosophila individuals in a group situation by the trial-and-error method, Zh. Obshch. Biol., 55, No. 6, (1994). 2. N. G. Kamyshev, E. A. Kamysheva, and G. O. Ivanova, Autosomal mutations in Drosophila which increase operant learning ability, Ros. Fiziol. Zh. im. I. M. Sechenova, 81, No. 8, (1995). 3. L. I. Korochkin, L. G. Romanova, N. G. Kamyshev, and G. P. Smirnova, The effects of naloxone on brain development and behavior in Drosophila melanogaster, Dokl. Akad. Nauk SSSR, 316, No. 4, (191).

8 408 Kamyshev, Smirnova, Kamysheva, Nikiforov, Parafenyuk, and Ponomarenko 4. E. M. Luchnikova, Movement Activity in Insects as a Factor in Behavioral Resistance to Insecticides. Genetic Studies [in Russian], Leningrad State University, Leningrad (1964), Second edition, pp I. P. Pavlov, The freedom reflex (with Dr M. M. Gubergritz), in: Twenty Years of Experience in the Objective Study of Higher Nervous Activity (Behavior) of Animals Conditioned Reflexes [in Russian], Complete Collection of Works (1949), Vol. 3, pp R. R. Baker, Insect territoriality, Ann. Rev. Entomol., 28, (1983). 7. K. J. Connolly, The social facilitation of preening behaviour in Drosophila melanogaster, Anim. Behav., 16, (1968). 8. F. R. van Dijken, H. Stolwijk, and W. Scharloo, Locomotor activity in Drosophila melanogaster, Neth. J. Zool., 35, (1987). 9. A. Ewing, Attempts to select for spontaneous activity in Drosophila, Anim. Behav., 11, (1963). 10. A. Ewing, Genetics and activity in Drosophila melanogaster, Experientia, 23, (1967). 11. J. L. Fuller and M. E. Hahn, Issues in the genetics of social behavior, Behav. Genet., 6, No. 4, (1976). 12. D. A. Hay, Effects of genetic variation and culture conditions on the social behaviour of Drosophila melanogaster, Behav. Genet., 3, (1973). 13. L. Kovac, E. Peterajova, and J. Pogady, Drosophila melanogaster, a new subject in research on behaviour and in pharmacology, Aggressologie, 20D, (1979). 14. K. Sakai, T. Narise, Y. Hiraizumi, and S. Iyama, Studies on competition in plants and animals. IX. Experimental studies on migration in Drosophila melanogaster, Evolution, 12, (1958). 15. C. J. Sexton and H. D. Stalker, Spacing patterns of female Drosophila melanogaster, Anim. Behav., 9, (1961).