Figure 5.

The withdrawal behavior of the subordinate mice is determined by their internal status Figure S3

(A) Associative memory model. An individual's unique social cues evoke a memory of previous social experiences.

(B) Stranger assay. The tube test was applied to compare the behavior of two cagemates (individuals housed together) vs. two strangers (individuals housed in different cages).

(C) The win rate for paired strangers compared with the expected rate of 50% (dashed line) (rank-1 vs. rank-1: p = 0.188, N = 15 pairs; rank-1 vs. rank-4: p < 0.001, N = 30 pairs; rank-4 vs. rank-4: p = 0.991, N = 15 pairs, one-sample Wilcoxon signed-rank test).

(D) Assay time across four trials for two cagemates and two strangers (p = 0.0179, unpaired t test, N = 12 pairs of cagemates, N = 30 pairs of strangers).

(E) Specialized social cue(s) model. Dominant winners emit an instructive social signal(s) that promotes retreat by the losers.

(F) Sensory removal. Social hierarchy was evaluated using mice without normal sensory sensations. Olfactory cues were removed by intraperitoneal injection of dichlobenil. Tactile cues were blocked by injection of lidocaine into the whisker pads of the mice. Visual signals were eliminated by performing the experiments under infrared light.

(G) Assay time of all paired conditions across four trials for evaluation of the social hierarchy in mice that lacked a sensory input (dichlobenil vs. DMSO mice: p = 0.57, lidocaine vs. PBS mice: p = 0.581, infrared-light condition vs. white-light condition: p < 0.001, Mann-Whitney test, N = 60–72 pairs) (see also Figure S4).

(H) Win rate for control mice (DMSO + PBS) and mice without olfaction and somatosensation (dichlobenil + lidocaine) compared with the expected rate of 50% (dashed line) under infrared light (Intact: p = 0.877, Sensory deficits: p = 0.877, one-sample Wilcoxon signed-rank test, N = 10 pairs).

(I) Internal status model. Losers are less persistent and more likely to withdraw when obstructed.

(J) Blocker assay. Further forward motion of winners or losers was blocked by an immovable object placed midway in the tube.

(K) Assay time across four trials that higher-ranked mice and lower-ranked mice interacted with the blocking material (p < 0.001, Mann-Whitney test, N = 10 pairs).

(L) Assay time across four trials that higher-ranked mice and lower-ranked mice spent in the tube containing the blocking material (p < 0.001, Mann-Whitney test, N = 10 pairs). ∗ = p < 0.05, ∗∗∗ = p < 0.001. Mean ± S.E.M.

See also Figure S3.