Recovery from social isolation requires dopamine in males, but not the autism-related gene nlg3 in either sex

Abstract

Social isolation causes profound changes in social behaviour in a variety of species. However, the genetic and molecular mechanisms modulating behavioural responses to social isolation and social recovery remain to be elucidated. Here, we quantified the behavioural response of vinegar flies to social isolation using two distinct protocols (social space preference and sociability, the spontaneous tendencies to form groups). We found that social isolation increased social space and reduced sociability. These effects of social isolation were reversible and could be reduced after 3 days of group housing. Flies with a loss of function of neuroligin3 (orthologue of autism-related neuroligin genes) with known increased social space in a socially enriched environment were still able to recover from social isolation. We also show that dopamine (DA) is needed for a response to social isolation and recovery in males but not in females. Furthermore, only in males, DA levels are reduced after isolation and are not recovered after group housing. Finally, in socially enriched flies mutant for neuroligin3, DA levels are reduced in males, but not in females. We propose a model to explain how DA and neuroligin3 are involved in the behavioural response to social isolation and its recovery in a dynamic and sex-specific manner.

Keywords: dopamine; neuroligin; recovery; social behaviour; social isolation; social space.

Figure 1.

Schematic of the experimental paradigm and parameters measured in this study. Flies were group housed in mixed-sex bottles for 2 days after eclosing to adulthood, prior to being single housed or remained group housed. Flies remained single or group housed for up to 7 days (reaching the age of 9-days-old). The length of isolation is always 7 days unless indicated in the figure. After 7 days of being group or single housed, flies were tested in either the social space assay, sociability assay or used to determine DA levels. Separate populations of flies followed the same experimental design but were group housed for up to 3 days (to test recovery, reaching the age of 12-days-old) before behaviour or molecular experiments took place. All flies in each independent repeat were naive to the tests performed, and no flies were used in multiple behavioural or molecular experiments. Created using BioRender.

Figure 2.

Isolation for 7 days leads to increased social space and decreased sociability, but flies recover after 3 days of group housing. (a,b,d,e) Social space presented in terms of mean number of flies within four body lengths ± s.e.m. (c,f) Effect of 2, 4 or 7 days of isolation in males (a) and females (b) social space. (a) Male flies had fewer flies within four body lengths after 2, 4 and 7 days of isolation (two-way ANOVA and Sidak post hoc test—F _1,46 = 22.01, p < 0.0001). (b) Females had fewer flies within four body lengths after isolation and that number decreased further with increasing days of isolation (two-way ANOVA—effect of isolation: F _1,47 = 6.653, p = 0.0131; effect of days of isolation: F _2,47 = 3.687, p = 0.0367). (c) Aggregation index in males and females after 7 days of isolation. Both males (Welch’s t‐test: t _12.08 = 4.039, p = 0.0016) and females (Welch’s t‐test: t _13.37 = 2.578, p = 0.0225) had a lower aggregation index after 7 days of isolation. (d,e) Males and females after 2 (d) and 3 (e) days of group housing (recovery) following 7 days of isolation. (d) Males showed no difference in the number of flies within four body lengths after 2 days of group housing compared with males continuously group housed. However, females had fewer flies within four body lengths compared with females group housed continuously (Welch’s t‐test: t _15.13 = 2.626, p = 0.0189). (e) Both males and females were not different in the number of flies within four body lengths comparing group-housed and recovery flies (one-way ANOVA: F _3,28 = 0.9434, p = 0.4329). (f) Sociability presented as the mean aggregation index ± s.e.m. of males and females group housed and recovery. Males and females did not differ in aggregation index between group housed and flies isolated 7 days, with 3 days of recovery. n = 7–12 for all treatments. On the graph, the results of the post hoc tests are shown only when p < 0.05. The results of all the comparisons, significant or not, can be found in tables 1–3. Grey shade: control group housed (GH); coloured shades: treatments; blue shades: single housed (SH) and recovery from social isolation (REC) males; red shades: SH and REC females.

Figure 3.

Mating status does not change social space in response to isolation and group housed with a single fly is enough to avoid a larger social space. (a,b) Number of flies within four body lengths in males (a) and females (b) group and single housed while mated or virgin. (a) Male flies had less flies within four body lengths after isolation (two-way ANOVA—effect of isolation: F _1,52 = 34.69, p < 0.0001) regardless of mating status. Males had less flies within four body lengths when virgin compared with mated flies (two-way ANOVA—effect of mating status: F _1,52 = 4.223, p = 0.0449). (b) Female flies had less flies within four body lengths after isolation (two-way ANOVA—effect of isolation: F _1,52 = 9.924, p = 0.0027) regardless of mating status. Females had less flies within four body lengths when virgin compared with mated flies (two-way ANOVA—effect of mating status: F _1,52 = 5.269, p = 0.0258). Grey shades: control group housed (GH); coloured shades: treatments; blue shades: single-housed (SH) males; red shades: SH females. (c,d) Number of flies within four body lengths in males (c) and females (d) reared single housed or group housed with 2, 6, 16 or a random number of flies. Single-housed flies had less flies within four body lengths than group-housed flies that were reared with any number of individuals in males ((c) one-way ANOVA with Holm-Sidak post hoc: F_{4 ,40} = 30.87, p < 0.0001) and females ((d) one-way ANOVA with Holm–Sidak post hoc: F _4,40 = 31.35, p < 0.0001). Grey shades: random number of flies; coloured shades: varying number of flies, from 16 to one (SH: single housed); blue shades: males; red shades: females. On the graph, the results of the post hoc tests are shown only when p < 0.05. The results of all the comparisons, significant or not, can be found in tables 1–3. n = 9–13 for all treatments. Bars: mean ± s.e.m.

Figure 4.

nlg3^Def1 flies have decreased social space after a recovery from isolation. (a, b) Number of flies within four body lengths in group- and single-housed males (a) and females (b) group housed. (a) Single-housed male flies had decreased flies within four body lengths compared with group-housed flies in both CS and nlg3^Def1 (two-way ANOVA—effect of social experience: F _1,32 = 46.48, p < 0.0001), and nlg3^Def1 flies had less flies within four body lengths in both group-housed and single-housed (two-way ANOVA—effect of genotype: F _1,32 = 10.10, p = 0.0033). (b) CS females had less flies within four body lengths in single-housed compared with group-housed flies; however, nlg3^Def1 flies were not different between group- and single-housed flies (two-way ANOVA—effect of social experience: F _1,32 = 27.97, p < 0.0001; effect of genotype: F _1,32 = 13.71, p = 0.0008; interaction of social experience and genotype: F _1,32 = 5.370, p = 0.0270). Grey shades: control group housed (GH); coloured shades: treatments; blue shades: single housed (SH); males; red shades: SH females. (c,d) Number of flies within four body lengths in group and recovered males (c) and females (d). (c) The number of flies within four body lengths for CS males was not different in group housed versus recovery flies; however, nlg3^Def1 flies had an increased number of flies within four body lengths in recovery compared with group-housed flies (two-way ANOVA—effect of social experience: F _1,37 = 17.72, p = 0.0002; effect of genotype: F _1,37 = 34.94, p < 0.0001; interaction of social experience and genotype: F_1,37 = 13.62, p = 0.0007). (d) The number of flies within four body lengths for CS females was not different in group-housed versus recovery flies; however, nlg3^Def1 females had an increased number of flies within four body lengths in recovery compared with group-housed flies (two-way ANOVA—effect of social experience: F _1,34 = 0.3197, p = 0.5755; effect of genotype: F _1,34 = 6.178, p = 0.0180; interaction of social experience and genotype: F _1,34 = 10.60, p = 0.0026). Grey shades: group housed (GH); coloured shades: recovery from social isolation (REC); blue shades: males; red shades: females. On the graph, the results of the post hoc tests are shown only when p < 0.05. The results of all the comparisons, significant or not, can be found in tables 1–3. n = 7–12 for all treatments. GH: group housed, SH: single housed, REC: recovery.

Figure 5.

DA is important for a response to social isolation and decreases after isolation in males but not females. Social space measured in terms of number of flies within four body lengths. Bars: mean ± s.e.m. (a,b) Social space in isolated male (a) and female (b) flies. (a) The social space in male TH>THmiR-G was not different in isolated versus group-housed flies; however, TH-GAL4/+ and UAS-THmiR-G/+ males had lower flies within four body lengths when in isolation (two-way ANOVA—effect of isolation: F _1,43 = 19.13, p < 0.0001; interaction between isolation and genotype: F _2,42 = 8.254, p = 0.0009). (b) In females, all three genotypes had a decrease in the number of flies within four body lengths after isolation and TH>THmiR G flies had the lowest number of flies within four body lengths compared with their genetic controls (two-way ANOVA—effect of isolation: F _1,37 = 21.91, p < 0.0001; effect of genotype: F_{2 ,37} = 4.009, p = 0.0265). (c,d) Social space in recovered male (c) and female (d) flies. (c) Male recovery flies in all genotypes were not different compared with group-housed males. (d) In females, all three genotypes were similar when comparing group-housed versus recovery treatments; however, TH>THmiR-G females had a less flies within four body lengths compared with the controls (two-way ANOVA—effect of genotype: F _2,35 = 5.124, p = 0.0112). (e) DA levels (pg head⁻¹) in males decreased in isolated compared with group-housed flies but do not differ in recovery flies compared with group housed in males or females (two-way ANOVA—effect of social experience: F _3,27 = 3.043, p = 0.0459, but the only significant difference after correcting for multiple comparison through a Dunnett test is indicated on the graph, p = 0.0132). Appropriate age comparisons were made (refer §2 for details). *p < 0.05. ****p < 0.0001. Bars: mean ± s.e.m. On the graph, the results of the post hoc tests are shown only when p < 0.05. The results of all the comparisons, significant or not, can be found in tables 1–3. n = 5–9 for all treatments in social space. n = 4 for DA quantification. GH: group housed, light grey shade; SH: single housed, dark grey shade; REC: recovery from social isolation, medium grey shade.

Figure 6.

nlg3^Def1 flies have less DA, and those levels are not altered by social experience. (a) DA levels (pg head⁻¹) in CS and nlg3^Def1 males (blue shade) and females (red shade). nlg3^Def1 flies had less DA when group-housed and isolated compared with group-housed CS, with females showing a lesser amount overall (two-way ANOVA—effect of genotype: F _1,7 = 17.21, p = 0.0043 and effect of sex: F _1,7 = 6.107, p = 0.0428). (b) DA levels (pg head⁻¹) in male (blue shade) and female (red shade) group-housed and recovery nlg3^Def1 flies are unchanged with social experience. Appropriate age comparisons were made (refer §2 for details). Bars: mean ± s.e.m. Grey shades: group housed (GH); coloured shades: social isolation (SH) or recovery from social isolation (REC), as indicated; blue shades: males; red shades: females. On the graph, the results of the post hoc tests are shown only when p < 0.05. The results of all the comparisons, significant or not, can be found in tables 1–3. n = 3–6 for all treatments. GH: group housed; SH: single housed; REC: recovery from social isolation.

Figure 7.

(a) Manipulations performed in this study. (b) Working model: nlg3 and DA are part of a pathway responsible for the modulation of social behaviour after isolation in males. nlg3 would be required downstream of DA, for proper DA signalling in response to the social environment, with some feedback regulation. Females do require nlg3 for a typical response to isolation, but other neurotransmitters or neuromodulators, responsive to social experience in females, may be interacting with nlg3. Blue outline: Drosophila brain outline. Created with BioRender.