Adolescent social isolation impairs pup-directed maternal behavior without affecting self-directed behavior at 1-week postpartum.

(A-B) Stressed dams showed reduced nursing (A) and pup-licking (B) behaviors. (C) Combined pup-directed care (nursing + licking) was significantly decreased in stressed dams. (D-E) Self-directed behaviors, including climbing/digging (D) and self-grooming (E), were unchanged. (F) Combined self-care behavior (climbing/digging + self-grooming) did not differ between groups. (G) Stressed dams built poorer-quality nests, reflected by lower nest scores. (H) Pup retrieval latency was significantly prolonged in stressed dams, indicating diminished maternal responsiveness. Data are shown as means ± SEM, N = 11-12 per group. *p < 0.05. Statistical tests: Student’s t-test (A, C), Welch’s t-test (B), and Mann-Whitney U (D-H) were applied as appropriate. Detailed statistical analyses are provided in Supplementary Table 1.

Offspring reared by stressed dams show selective social behavior deficits.

(A) In the three-chamber social interaction test (SIT), stressed offspring showed reduced sociability compared with unstressed offspring. (B) Stressed offspring also displayed impaired social novelty recognition in the SIT. (C) In the social olfactory recognition test (SORT), stressed offspring exhibited diminished investigation toward both familiar and novel social odors. (D) Performance in the novel object recognition test (NORT) was unaffected by maternal adversity. (E) In the open field test (ORT), time spent in the center did not differ between groups. (F and G) Anxiety-related behaviors in the elevated plus maze (EPM) and light-dark box (LDB) showed no significant differences between groups. Together, these results indicate that impaired maternal behavior in stressed dams selectively disrupts offspring social behavior while sparing cognitive and anxiety-related domains. Data are shown as means ± SEM, N = 18-32 per group. *p < 0.05. Statistical tests: Welch’s t-test (A, B), two-way mixed ANOVA (C), Mann-Whitney U (D, G), and Student’s t-test (E, F), and were applied as appropriate. Detailed statistics are provided in Supplementary Table 1.

Chemogenetic modulation of mCg→PrL pathway rescues or induces social deficits.

(A) Schematic illustrating virus delivery of AAV-hSyn-hM3D(Gq)-mCherry into the mCg and placement of guide cannulas into the PrL. (B and C) Activation of M3Dq-expressiong mCg→PrL neurons via CNO in stressed offspring rescued sociability and social novelty recognition deficits. (D) Schematic illustrating virus delivery of AAV-hSyn-hM4D(Gi)-mCherry into the mCg and placement of guide cannulas into the PrL. (E and F) Inhibition of the same circuit in unstressed offspring using M4Di receptors induced social deficits. Data are shown as means ± SEM, N = 6 per group. *p < 0.05. Statistical tests: Two-way ANOVA (B-C) and Student’s t-test (E-F) were applied. Detailed statistics are provided in Supplementary Table 1. Panel A and D created with BioRender.com

Co-housing with parous females rescues impaired maternal caregiving in stressed dams.

(A) Experimental design of co-housing with parous females. (B-D) Pup-directed behaviors were increased in stressed dams co-housed with parous, but not virgin, females. (E-G) Self-care behaviors were not affected by co-housing conditions. (H and I) Nest quality and pup retrieval performance were improved following co-housing with parous females. Data are shown as means ± SEM, N = 12 per group. *p < 0.05. Statistical tests: All non-normal data were analyzed using pairwise planned comparisons in Mann-Whitney U tests. Detailed statistics are provided in Supplementary Table 1. Panel A created with BioRender.com.

Co-housing with parous mice rescues social deficits in stressed offspring.

(A) Sociability index was decreased in stressed offspring but rescued by co-housing with parous females from birth to weaning. (B) Social novelty recognition followed a similar pattern and was rescued by parous co-housing. Data are shown as means ± SEM. Sample size for sociability index is N = 7-18 per group. *p < 0.05. Statistical tests: Kruskal-Wallis test (A) and Welch’s one-way ANOVA (B) were applied. Detailed statistics are provided in Supplementary Table 1.

The mCg→PrL pathway is predominantly excitatory and shows reduced firing in stressed offspring that is restored by co-housing.

(A) Left: Schematic illustrating virus delivery of rgAAV-hSyn-EGFP into the PrL. Right: Quantification of EGFP-labeled mCg neurons retrogradely labeled from PrL co-localized with CaMKII or GABA. Data are shown as means ± SEM, N = 3 per group. Each dot represents an average of three slices per animal. (B and C) Representative immunohistochemistry images of EGFP-labeled mCg neurons retrogradely labeled from PrL co-localized with CaMKII or GABA. Scale bars: 50 µm. (D) Schematic image and representative mCg image showing mCherry-labeled neurons following injection of a CaMKll promoter-driven retrograde virus into the PrL, confirming the excitatory nature of the mCg→PrL circuit. Scale bars: 100 µm. (E) Firing frequency of action potentials in layer V mCg→PrL neurons was reduced in stressed offspring, and rescued in stressed offspring co-housed with parous mice. Sample size for firing frequency is 10-13 cells per group, obtained from 4 different mice in each group. *p < 0.05. Statistical tests: Two-way mixed ANOVA (C) was used. Detailed statistics are provided in Supplementary Table 1. Panel A and D created with BioRender.com.

Early postpartum maternal care deficits are limited at delivery.

(A-C) Caring behavior toward pups shortly after delivery (postpartum day 1) was largely unchanged, except for pup licking. (D-F) Self-care behaviors, including climbing/digging and self-grooming, were not affected. (G) Nest scores were similar between unstressed and stressed dams. (H) Pup retrieval latency was increased in stressed dams, indicating reduced maternal responsiveness. Data are shown as means ± SEM, N = 11-12 per group. *p < 0.05. Statistical tests: Mann-Whitney U (A-D, F-H) and Welch’s t-tests (E) were applied as appropriate. Detailed statistics are provided in Supplementary Table 1.

Effects of impaired maternal caregiving on offspring growth and survival.

(A) Pup body weight gain over the first postnatal week was similar between groups. (B) Survival rates were reduced in stressed offspring compared to unstressed controls. (C) Number of dead pups per litter was not significantly different, although stressed litters tended to have more deaths. Data are shown as means ± SEM, N = 10-11 litters per group. *p < 0.05. Statistical tests: Mann-Whitney U and Student’s t Test (A, C) and Log-rank Mantel-Cox (B) were applied as appropriate. Detailed statistics are provided in Supplementary Table 1.

Validation of behavioral measures.

(A) Total distance moved in the open field test was similar between groups, indicating no differences in locomotor activity. (B) Discrimination index in the novel object recognition test during the training session was similar across groups. Data are shown as means ± SEM, N = 24-25 per group. *p < 0.05. Statistical tests: Student’s t-test (A) and Mann-Whitney U test (B) were applied as appropriate. Detailed statistics are provided in Supplementary Table 1.

Increased plasma corticosterone levels in stressed offspring.

Plasma corticosterone levels were elevated in adult offspring reared by stressed dams. Data are shown as means ± SEM, N = 19-22 per group. *p < 0.05. Statistical tests: Mann-Whitney U test was applied. Detailed statistics are provided in Supplementary Table 1.

Parous mice display robust maternal behavior.

(A-C) Pup-directed behaviors, including nursing, licking, and retrieval, were increased in parous females compared to virgins. (D-F) Self-care behaviors were similar between virgin and parous females. (G) Pup retrieval latency was reduced in parous females, reflecting enhanced maternal responsiveness. Data are shown as means ± SEM, N = 6 per group. *p < 0.05. Statistical tests: Mann-Whitney U test (A, C, E, G) and Student’s t-test (B, D, F) were applied as appropriate. Detailed statistics are provided in Supplementary Table 1.

Electrophysiological characterization of mCg→PrL neurons in stressed offspring.

(A) Rheobase of these neurons showed no change. (B) Resting membrane potential was similar between groups. (C-F) Amplitude and frequency of spontaneous excitatory and inhibitory currents were unchanged, indicating intact presynaptic input. Data are shown as means ± SEM, N = 10-12 cells per group. Each group included cells obtained from at least three animals. Statistical tests: Two-way mixed ANOVA (A) and Student’s/Welch’s t-tests (B-G) were applied. Detailed statistics are provided in Supplementary Table 1.

Statistical analyses for all results.

For each dataset, sample sizes, normality tests, statistical tests performed, p values, effect sizes, and post hoc comparisons are reported. For repeated-measures ANOVAs, selected within-group factors were included in post hoc analyses.