Adolescent social isolation creates a latent vulnerability in maternal care with intergenerational social consequences, rescued by experienced mothers

Reviewer #1 (Public review):

Summary:

In the manuscript by Francis-Oliveira et al., the authors investigated whether mild adolescent social isolation in female mice produces a latent vulnerability that emerges during the postpartum period as impaired maternal caregiving. They further tested the hypothesis that maternal deficits alter offspring social development. Based on their findings, the authors propose that adolescent psychosocial adversity disrupts maternal behavior and that resulting alterations in offspring social function are mediated through dysfunction of the midcingulate cortex (mCg) to prelimbic cortex (PrL) pathway. They further suggest that exposure to experienced parous females during the postpartum period can rescue maternal behavior and normalize offspring outcomes through restoration of activity in this circuit.

To test these hypotheses, the authors exposed female mice to mild social isolation during late adolescence and subsequently bred those females. Maternal behaviors were assessed postpartum, and offspring were evaluated in adulthood using assays of sociability, social novelty recognition, social odor recognition, anxiety-like behavior, locomotion, and non-social memory. The authors also measured corticosterone levels in control and stress-reared offspring. To test circuit-specific effects, the authors employed chemogenetic activation and inhibition of the mCg→PrL pathway using DREADDs and performed electrophysiological recordings from identified projection neurons. Finally, stressed dams were co-housed with experienced parous females during the postpartum period to determine whether maternal and offspring phenotypes could be rescued, as well as the social deficits previously observed in offspring.

The authors found that adolescent isolation selectively impaired pup-directed maternal behaviors, including nursing, licking, nest building, and pup retrieval, while leaving self-directed behaviors intact. Adult offspring of stressed dams exhibited deficits in sociability, social novelty recognition, and social odor discrimination, but showed no impairments in locomotor activity, anxiety-like behavior, or novel object recognition. Chemogenetic activation of the mCg→PrL pathway restored social behavior in stressed offspring, whereas inhibition of the pathway induced social impairments in controls. Co-housing stressed dams with experienced parous females restored maternal caregiving, normalized offspring social behavior, and rescued reduced firing of mCg→PrL neurons observed in offspring of stressed dams. Collectively, these findings support the authors' model that adolescent psychosocial adversity disrupts maternal caregiving and contributes to offspring social deficits through dysfunction of the mCg→PrL circuit.

Strengths:

The study includes multiple levels of assessment, including behavioral measures, behavioral intervention, the use of DREADDs for circuit manipulation to both test effects of activation versus inhibition on behavioral outcomes as well as physiology experiments. The multilevel approach is a strength.

Weaknesses:

(1) Interpretation of the parous co-housing experiment:

The principal limitation of the study is that the communal housing paradigm does not distinguish rescue of maternal behavior in the stressed dam from direct caregiving provided by the experienced parous female. The authors interpret the rescue experiment as evidence that social support and/or social learning from experienced mothers restores maternal behavior in stressed dams, which in turn normalizes social behavior in offspring. However, pups were continuously housed with both the stressed dam and the parous female from P0-P7, and the parous female had unrestricted access to the pups throughout the intervention period. The parous female was removed only briefly during maternal behavior testing. This design raises an important alternative interpretation. The experienced parous female may have directly provided substantial maternal care to the pups, supplementing or compensating for deficits in the stressed dam. The rescue of offspring phenotypes may reflect care received from the parous female rather than improved caregiving by the stressed dam.

Were caregiving behaviors of the stressed dam and parous female quantified separately during the co-housing period? What proportion of licking, nursing, retrieval, and nest maintenance was performed by each animal? Can the authors exclude the possibility that direct maternal care from the parous female, rather than social learning or social support, accounted for the rescue of offspring outcomes? Without such controls, the central claim that restoration of maternal behavior in the stressed dam mediates normalization of offspring phenotypes is not fully supported.

(2) Specificity of the behavioral phenotype and rescue:

The manuscript repeatedly frames the findings as restoration of offspring outcomes and intergenerational vulnerability. However, the behavioral phenotype appears highly selective and restricted primarily to social behaviors. Offspring exhibited impairments in sociability, social novelty recognition, and social odor discrimination, but showed normal locomotion, anxiety-related behavior, and non-social memory. Thus, the authors should more explicitly acknowledge that maternal adversity produced a domain-specific social phenotype rather than broad behavioral dysfunction. Interestingly, the rescue studies only evaluated a subset of the affected behaviors, making it unclear whether co-housing with parous females restored broader aspects of offspring neural or behavioral function or selectively improved specific social behaviors.

Why was social olfactory recognition not included in the rescue experiments? Why were additional behavioral measures not reported following circuit activation or parous co-housing (e.g., anxiety-like behavior and novel object test) - were these improved in controls by enriched early parenting? Did manipulation of the mCg→PrL pathway or co-housing with parous females influence anxiety-like or depressive-like behaviors despite the absence of baseline group differences?

(3) Strength of the DREADD-mediated causal claims:

The DREADD experiments implicate the mCg→PrL pathway in regulating social behavior; however, several aspects limit the strength of the causal conclusions. Sample sizes were relatively small (n = 6/group). In addition, the variance observed in the DREADD cohorts appears substantially reduced relative to that observed in the non-surgical cohorts. For example, vehicle-treated groups appear more clearly separated than would be expected based on the original behavioral data presented in Figure 2. Can the authors comment on potential reasons for this discrepancy?

Second, although activation and inhibition experiments support involvement of the mCg→PrL pathway in social behavior, the manipulations do not fully recapitulate the broader phenotype observed in stressed offspring. Thus, the data support a role for this pathway but may not justify the stronger conclusion that dysfunction of this circuit alone accounts for the entirety of the offspring phenotype.

The authors argue that altered maternal behavior is causal for social deficits in offspring. However, in the absence of a cross-fostering experiment, the study cannot fully exclude alternative explanations, including direct influences of caregiving by the co-housed parous female, gestational effects, altered maternal physiology during pregnancy, or germline-mediated influences. A cross-fostering design would substantially strengthen the causal interpretation of the findings.

Additionally, clarification of litter effects is important: For example, how many litters contributed to each experimental group? Was litter treated as a random effect in statistical analyses? Were multiple offspring from the same litter analyzed as independent observations? Because maternal behavior is manipulated at the litter level, litter rather than individual offspring may represent the appropriate experimental unit for many analyses.

(4) Integration of corticosterone findings into the mechanistic model:

The corticosterone findings appear somewhat disconnected from the central mechanistic narrative. The authors report elevated corticosterone levels in stressed offspring and suggest that HPA-axis dysregulation may contribute to the observed behavioral phenotype. However, the manuscript does not establish whether corticosterone plays a causal role in the social deficits or instead represents a parallel physiological consequence of altered maternal care.

Were corticosterone levels normalized by co-housing with parous females? Did DREADD-mediated activation of the mCg→PrL pathway normalize corticosterone levels? Could corticosterone manipulation alone drive aspects of the behavioral phenotype independent of circuit manipulation? Do corticosterone levels correlate with the severity of social behavioral impairments? It is difficult to determine whether corticosterone is mechanistically relevant or simply serves as an associated physiological marker. The authors should either more directly integrate the endocrine findings into their mechanistic framework or temper discussion suggesting a causal role for HPA-axis dysfunction.

In summary, this manuscript addresses an important and understudied question concerning how adolescent adversity influences maternal caregiving and offspring social development. The behavioral, circuit, and electrophysiological findings are generally coherent and support a role for the mCg→PrL pathway in mediating offspring social outcomes. However, the strongest mechanistic claim, that social support rescues offspring phenotypes by restoring maternal behavior in stressed dams, is weakened by the communal rearing design, which allows direct caregiving by parous females. Additional clarification regarding caregiver-specific behaviors, litter effects, the role of corticosterone, and the specificity of the DREADD-mediated phenocopy would strengthen the causal interpretation of the findings. Overall, the study is potentially impactful, but several conclusions currently extend beyond what is directly supported by the data.

Reviewer #2 (Public review):

Summary:

Studies in rodents have demonstrated that early life adversity (ELA) impacts many aspects of the exposed offspring's brain and behavior. Work in this field has traditionally focused on how
stress in very early life can impact cognitive and emotion-related behaviors in the ELA-exposed offspring. By contrast, this manuscript focuses on how stress in a slightly later adolescent period can produce latent and intergenerational effects by impacting maternal caregiving from female offspring, as well as social outcomes of the next generation of animals born to ELA-exposed females. Specifically, the manuscript describes that female mice exposed to ELA in the form of social isolation in late adolescence show reduced pup-directed maternal behaviors, while self-directed behaviors remain intact. Offspring reared by these dams in turn show deficits in social behavior, which are linked to reduced activity in an excitatory connection between the medial cingulate cortex (mCg) and prelimbic cortex (PrL). Further, the authors find that social behavior can be rescued by chemogenetic activation of the mCg-PrL pathway in the offspring of ELA-exposed/stressed mice or recapitulated in control mice by chemogenetic inhibition of this connection. Importantly, co-housing ELA-exposed/stressed dams with experienced parous females during the early postpartum period restores pup-directed maternal behaviors in these mice and normalizes offspring social outcomes as well as mCg-PrL activity.

Strengths:

Strengths of the manuscript include the focus on an important and novel question about intergenerational effects of adolescent ELA transmitted via subsequent maternal care, and the use of multiple techniques to link circuit function to behavior, including slice electrophysiology and chemogenetics. While the findings that maternal care can influence offspring behavior and that experienced females can instruct and improve maternal care of less experienced mice are not novel, they add support to this important area of literature.

Weaknesses:

Weaknesses of the paper include the lack of validation that the viral chemogenetic paradigm was appropriately targeted in the brain and impacted the excitability of mCg to PrL projections as anticipated, the use of inappropriate statistical tests that do not account for non-independence of pups from the same litter or cells measured from the same pup or categorical versus continuous data, and the lack of important descriptions of methods or experimental paradigms in several places that altogether make it difficult to judge the rigor of the findings in its current state.

If these weaknesses are addressed, these findings will provide important information about circuit mechanisms underlying intergenerational effects of adolescent stress on social behavior in next-generation offspring.