Early life stressful experiences escalate aggressive behavior in adulthood via changes in transthyretin expression and function
Curation statements for this article:-
Curated by eLife
Evaluation Summary:
This work will be of interest to biological and clinical specialists interested in the fields of behavioural neuroscience, biological psychiatry, neuroendocrinology, and developmental psychology for its focus on the origins of adult aggressive behavior in early life stress. The authors used an unbiased transcriptomic analysis and identified the thyroid hormone system as a potential mediator of the enduring impact of early stress and aberrant aggressive behavior in adulthood.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #2 agreed to share their name with the authors.)
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
- Neuroscience (eLife)
Abstract
Escalated and inappropriate levels of aggressive behavior referred to as pathological in psychiatry can lead to violent outcomes with detrimental impact on health and society. Early life stressful experiences might increase the risk of developing pathological aggressive behavior in adulthood, though molecular mechanisms remain elusive. Here, we provide prefrontal cortex and hypothalamus specific transcriptome profiles of peripubertal stress (PPS) exposed Balb/c adult male mice exhibiting escalated aggression and adult female mice resilient to such aberrant behavioral responses. We identify transthyretin (TTR), a well known thyroid hormone transporter, as a key regulator of PPS induced escalated aggressive behavior in males. Brain-region-specific long-term changes in Ttr gene expression and thyroid hormone (TH) availability were evident in PPS induced escalated aggressive male mice, circulating TH being unaltered. Ttr promoter methylation marks were also altered being hypermethylated in hypothalamus and hypomethylated in prefrontal cortex corroborating with its expression pattern. Further, Ttr knockdown in hypothalamus resulted in escalated aggressive behavior in males without PPS and also reduced TH levels and expression of TH-responsive genes ( Nrgn , Trh, and Hr ). Escalated aggressive behavior along with reduced Ttr gene expression and TH levels in hypothalamus was also evident in next generation F1 male progenies. Our findings reveal that stressful experiences during puberty might trigger lasting escalated aggression by modulating TTR expression in brain. TTR can serve as a potential target in reversal of escalated aggression and related psychopathologies.
Article activity feed
-
-
Evaluation Summary:
This work will be of interest to biological and clinical specialists interested in the fields of behavioural neuroscience, biological psychiatry, neuroendocrinology, and developmental psychology for its focus on the origins of adult aggressive behavior in early life stress. The authors used an unbiased transcriptomic analysis and identified the thyroid hormone system as a potential mediator of the enduring impact of early stress and aberrant aggressive behavior in adulthood.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #2 agreed to share their name with the authors.)
-
Reviewer #1 (Public Review):
The manuscript focuses on an important question, how early life trauma causes aggression later in life. As aggression may ruin the life of both the aggressor and the victim and the life of their families, this question influences the life of a relatively large population. Uncovering the mechanisms of this behavior may provide options for treatment.
Based on transcriptome analysis, the authors suggest that epigenetic downregulation of TTR and the resultant hypothalamic decrease of thyroid hormone availability are responsible for the long lasting effects of early life trauma on the behavior. Using virus mediated gene knock down, the authors replicated the behavioral effects of the early life trauma demonstrating the involvement of decreased TTR expression in the development of aggression.
Strengths
The well …
Reviewer #1 (Public Review):
The manuscript focuses on an important question, how early life trauma causes aggression later in life. As aggression may ruin the life of both the aggressor and the victim and the life of their families, this question influences the life of a relatively large population. Uncovering the mechanisms of this behavior may provide options for treatment.
Based on transcriptome analysis, the authors suggest that epigenetic downregulation of TTR and the resultant hypothalamic decrease of thyroid hormone availability are responsible for the long lasting effects of early life trauma on the behavior. Using virus mediated gene knock down, the authors replicated the behavioral effects of the early life trauma demonstrating the involvement of decreased TTR expression in the development of aggression.
Strengths
The well defined experimental model and the selection of extreme phenotypes helps to identify the genes that are involved in the development of phenotype. The examination of females where the PPS does not cause aggression also helped to identify the important genes.
The suggested role of TTR in the development of aggression is proved by virus mediated gene knock down.
Weaknesses
However, the authors clearly demonstrated that both the TTR knock down and the early life trauma result in a decrease of hypothalamic thyroid hormone availability, they did not examine whether this local hypothalamic hypothyroidism is involved in the development of aggression. This question is important as in humans, hypothyroidism is not associated with aggression, rather increased T3 level was found in association with aggression. Therefore, it is possible that the decreased TTR expression causes the aggressive phenotype independently from its effect on the hypothalamic thyroid hormone availability. This could be tested by examining whether local hypothalamic T3 administration can reverse the aggressive phenotype of the used mouse models.
There is a discrepancy in the data. Despite of the large increase of hypothalamic TRH expression, the circulating thyroid hormone levels are not influenced. There are many TRH neuron populations in the hypothalamus and only a small portion of the hypothalamic TRH neurons are involved in the regulation of the circulating thyroid hormone levels. Therefore, it would be necessary to perform in situ hybridization to determine which TRH neuron population is regulated in the experimental model. Because of the unchanged circulating thyroid hormone levels, it is unlikely that the TRH expression is increased in the hypophysiotropic TRH neurons of the PVN. The in situ hybridization data could help to understand which cell populations of the hypothalamus could be involved in the development of aggression. For example, there is a TRH neuron population in the lateral hypothalamic attack area (PMID: 15908131) that could be involved in this behavior.
The authors measured serum total T4 and T3 levels. This could be misleading as the thyroid hormone binding capacity of blood may highly influence these data. Thus, measurement of free thyroid hormone levels would be far more informative.
The quality of the images illustrating immunocytochemistry is very weak. -
Reviewer #2 (Public Review):
In this manuscript, Rawat et al. have tackled a long-lasting problem of developmental origins of excessive aggressive behaviours in mammals, which remained to be heavily understudied even in our days of the emerged postgenome era. The authors were able to produce aggravated aggressive behaviours in male mice pre-exposed to repeated juvenile stress during the peripubertal period, screening also for molecular alterations related to thyroid function in their brains. They identified several molecular changes, including shifts in the levels of TTR protein and within-brain levels of thyroid hormones, in the brains of the animals that differed in the levels of excessive aggressive behaviours. These behaviour-related molecular changes were suggested to be associated with thyroid hormones bioavailability in the …
Reviewer #2 (Public Review):
In this manuscript, Rawat et al. have tackled a long-lasting problem of developmental origins of excessive aggressive behaviours in mammals, which remained to be heavily understudied even in our days of the emerged postgenome era. The authors were able to produce aggravated aggressive behaviours in male mice pre-exposed to repeated juvenile stress during the peripubertal period, screening also for molecular alterations related to thyroid function in their brains. They identified several molecular changes, including shifts in the levels of TTR protein and within-brain levels of thyroid hormones, in the brains of the animals that differed in the levels of excessive aggressive behaviours. These behaviour-related molecular changes were suggested to be associated with thyroid hormones bioavailability in the tissues of brain limbic regions linked with aggressive behaviours manifestations.
While the general topic of developmental origins of excessive aggressive behaviours highlighted in this study is of high importance, the initial motivation of this investigation provided in the introduction is not linearized well and remains fuzzy in parts. In particular, it is hard to delineate between general aggressiveness and excessive aggressiveness, while reading the "Introduction" section of the manuscript.
Comparing contrast phenotypes, which differed in the levels of aggression, is one of the main strengths of the experimental schedules designed by Rawat et al. for this work. The suggested experimental schedules with contrast phenotypes make possible dissecting such a complex trait as abnormal aggressive behaviour, in a style of solution for the Gordian-Knot-like problems. Nevertheless, it should be remembered that experimental schedules with contrasted, "supernormal" groups (for example, see Chen et al. 2005 https://doi.org/10.1016/j.mehy.2005.04.037) might identify aggression-related molecular changes specific only for the mice line studied (Balb/c mice in the case of Rawat et al.).
The provided description of employed animal housing procedures, experimental schedules and behavioural/molecular assays applied is too scarсe and non-consecutive in parts, compromising both general comprehensibility of the data reported and possible reinterpretations of the data by readers. For example, the scarcity in the descriptions of animal housing procedures provided in the "Materials and Methods" section of the manuscript does not allow identifying whether reported stress-induced effects had been produced purely by the specific experimental schedule employed by the authors, or were a combined product of the applied schedule and specific housing conditions related to possible post-weaning social isolation or social crowding events. The sudden emergence of two poorly-defined female groups within experimental schedules in the Results section also is hard to interpret somehow without detailed reading of the previous work (Ref 7.) Whether these female mice were non-aggressive spontaneously, or after some stress treatments, also remained undefined in the text.
The manifestations of excessive aggressive behaviours in a number of adult male mice tested, which were pre-exposed to the repeated stress during the juvenile period, evidence that the employed stress procedure had worked properly. Nevertheless, with no estimations of anxiety-/depression-related behaviours and corticosteroid/stress resilience levels in the animals, it is hard to compare the stress strength produced by the specific stress procedure employed in this study with other, more common paradigms of peripubertal stress reported by other groups (For example, see Takahashi et al. 2021, https://doi.org/10.1007/7854_2021_243) .
While the behavioural testing reported by the authors was limited only by the Resident intruder test traditional for aggressive behaviour studies in mice, molecular part of this work is more divert, with transcriptomic data followed by qPCR, immunoblot and immunohistochemistry validations of expression changes, as well as by biochemical estimations of thyroid hormones' concentrations and siRNA-based manipulations of TTR mRNA levels.
The transcriptomic estimations of mRNA levels might became one of the strengths for this manuscript, in case of (a) sufficient reporting on the pipeline bioinformatics procedures employed, freeware parameters applied for Tophat, Cufflinks, other programs; and (b) individual transcript levels for each biological replicate reported. In addition, the raw transcriptomic data have been remained underanalyzed, with several statements done on the DEGs in the "Results" section (for example, on TTR mRNA levels) contradicting the aggregate mRNA levels reported for experimental group on the GEO. Avoiding factorial analysis for individual transcript level estimates, neglecting interpretations of identified DEGs with Allen Brain Transcriptomic Atlas, and missing cell deconvolution analysis also nullifies possible strengths of the RNA-Seq data reported in this interesting study.
The series of assays employed by the authors to study thyroid-related molecular changes, including the experiment with direct suppression of TTR levels in the hypothalamus by siRNA, is another of main strengths inherent to this manuscript. While the within-text descriptions of stress-induced level changes for TTR and related transcripts is fuzzy in parts, the visual representation of transcript and protein levels' shifts, as well as shifts in thyroid hormones' levels mitigates the puzzles produced for these results in the text. Nevertheless, the interpretation of transcript levels' changes with false precision is not optimal.
At first glance, the provided line of observations on stress-induced changes in the levels of transcript presumed to be associated with the changes in thyroid availability in the limbic regions analysed is looked convincing. Based on these findings, the authors claimed that "Early life trauma leads to escalated aggressive behaviour and its inheritance by impairing thyroid hormone availability in brain". But after a detailed reading on the topic, the causation claim made on a role of thyroid hormones in the subsequent manifestation of excessive aggressive behaviour in stressed male mice appeared to be too preliminary.
In particular, not only thyroid hormones but also retinoids were suggested to be associated with manifestations of excessive aggressive behaviours in humans, as it was mentioned anecdotally in the number of classic medical grossbooks by clinical researchers. While the molecular basis of excessive aggressive behaviours linked either with thyroid hormones or retinoids remains heavily understudied in the postgenome era, several lines of evidence suggest a causal role of retinoids in aggression manifestations as well as in cognitive and mood disorders (For example see Mawson 2009 PNPP https://doi.org/10.1016/j.pnpbp.2008.10.019). And it should be stressed that the TTR carrier protein studied by the authors in this work is responsible not only for the transport of thyroid hormones, but for the retinol exchange also (Buxbaum et al. 2014 Ns https://doi.org/10.1016/j.neuroscience.2014.06.019 ). Since the authors did not estimate the levels of retinoids and retinoid-related molecules in the brain of their mice, it remains obscure whether vitamin A, or thyroid hormones, or both of these are linked to excessive aggressive phenotype identified.
With these talking-points, the proposed claim on the causation link between the impaired thyroid hormones' availability in the brain and the development of excessive aggressive behaviour in the adulthood is not directly supported by the results presented. Regardless of all abovementioned unsupported claims, the evidences collected by Rawat et al. on associations of repeated juvenile stress with manifestations of excessive aggressive behaviour, changes in the levels of specific transcripts and within-brain thyroid levels availability, are of importance for the general community.
Finally, it should be stressed that the studies on the possible roles of thyroid hormones and retinoids on origins of excessive aggressive behaviours often are hard to conduct due to the provocativity of the problem of pathological aggressiveness and uneasiness of ethical approval for such investigations. The work by Rawat et al. has overcome these initial obstacles already, providing a number of novel clues on the developmental origins of aggressive behaviours, highlighting the novel routes to be researched by the scientific society on this not-easy topic.
-
Reviewer #3 (Public Review):
Early life trauma is a risk factor for adult aberrant aggressive behavior but this important public health issue remains under examined in the neurosciences. This study seeks to fill the gap with a mouse model of adolescent trauma that involves a combination of fearful and anxiety-provoking experiences and assessment on gene expression in brain region controlling aggression, the hypothalamus, and another controlling executive function, the prefrontal cortex. Mice are categorized for aggressive phenotype as being extreme or moderate, with the extreme being compared to controls for transcriptomic analyses of the hypothalamus and PFC. Females did not show increased adult aggression in the resident-intruder paradigm following adolescent fear and anxiety. Pathway analysis implicated the thyroid hormone pathway in …
Reviewer #3 (Public Review):
Early life trauma is a risk factor for adult aberrant aggressive behavior but this important public health issue remains under examined in the neurosciences. This study seeks to fill the gap with a mouse model of adolescent trauma that involves a combination of fearful and anxiety-provoking experiences and assessment on gene expression in brain region controlling aggression, the hypothalamus, and another controlling executive function, the prefrontal cortex. Mice are categorized for aggressive phenotype as being extreme or moderate, with the extreme being compared to controls for transcriptomic analyses of the hypothalamus and PFC. Females did not show increased adult aggression in the resident-intruder paradigm following adolescent fear and anxiety. Pathway analysis implicated the thyroid hormone pathway in male hypothalamus with the thyroid receptor, Ttr, being the top candidate gene. This formed the basis of an in depth analyses of thyroid hormone pathway and discovery of reduced T3 following adolescent stress which was causally linked to adult aggression. This is a novel observation with potentially important implications.
The strengths of the study are the detailed behavioral analyses, inclusion of both sexes and down regulation of Ttr specifically in hypothalamus, reducing T3 and increasing aggression. The weaknesses are a lack of mechanistic explanations for how reduced T3 and T4 leads to pathological aggression in males, weakly supported claims of transgenerational inheritance, lack of consideration of other pathways and no explanation for the profound sex difference.
Specific Comments
- The KEGG analyses does implicate the thyroid hormone pathway but the more consistent changes seem to be in drug addiction pathways and estrogen signaling, leaving one to wonder if the emphasis on the TH pathway is truly warranted.
- Aggression in females under normal circumstances is not evoked by a male intruder unless the female has a litter. Thus, it is not that surprising that the peripubertal stress did not evoke aggression in virgin females. Rather, the more interesting question is whether maternal aggression would become aberrant after peripubertal stress.
- Regarding the trans-generational transmission of the PPS, since the germ cells were present in the animals that were subject to PPS and gave rise to the offspring that were then tested, this is not truly transgenerational as the germ cells were residing in the stressed body. The transmission needs to be to at least the F2 generation with no stress in the F1 for this to be considered transgenerational.
- Regarding the methylation status of the Ttr, confidence in this result requires consideration of other targets as well in order to understand whether the epigenetic modifications are specific to just Ttr or are more widespread.
- The statistical analysis rests on unpaired t-tests but in most experiments a 2-way ANOVA is warranted with treatment and brain region as factors.
- The word "trauma" in the context used here connotes an emotional interpretation of stressful or fearful events. We do not know if the mice are experiencing trauma, instead we know they are being subject to fearful and stress-inducing experiences. It is suggested that the word trauma be removed throughout and replaced with more precise terminology.
-
-
-