If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Anatomy and Neurobiology, University of California Irvine, Irvine, CaliforniaDepartment of Pediatrics, University of California Irvine, Irvine, CaliforniaDepartment of Neurology, University of California Irvine, Irvine, California
Early-life adversity (ELA) is associated with increased risk for mood disorders, including depression and substance use disorders. These disorders are characterized by impaired reward-related behaviors, suggesting compromised operations of reward-related brain circuits. However, the brain regions engaged by ELA that mediate these enduring consequences of ELA remain largely unknown. In an animal model of ELA, we identified aberrant reward-seeking behaviors, a discovery that provides a framework for assessing the underlying circuits.
Methods
Employing TRAP2 (targeted recombination in active populations) male and female mice, in which neurons activated within a defined time frame are permanently tagged, we compared ELA- and control-reared mice, assessing the quantity and distribution of ELA-related neuronal activation. After validating the TRAP2 results using native c-Fos labeling, we defined the molecular identity of this population of activated neurons.
Results
We uniquely demonstrated that the TRAP2 system is feasible and efficacious in neonatal mice. Surprisingly, the paraventricular nucleus of the thalamus was robustly and almost exclusively activated by ELA and was the only region distinguishing ELA from typical rearing. Remarkably, a large proportion of ELA-activated paraventricular nucleus of the thalamus neurons expressed CRF1, the receptor for the stress-related peptide, corticotropin-releasing hormone, but these neurons did not express corticotropin-releasing hormone itself.
Conclusions
The paraventricular nucleus of the thalamus, an important component of reward circuits that is known to encode remote, emotionally salient experiences to influence future motivated behaviors, encodes adverse experiences as remote as those occurring during the early postnatal period and is thus poised to contribute to the enduring deficits in reward-related behaviors consequent to ELA.
Early-life adversity (ELA) consisting of trauma, poverty, or tumultuous environment impacts the lives of more than 30% of children in the United States (
). In humans, ELA is associated with poor cognitive and emotional health and increased risk for mood disorders such as depression as well as increased risk for substance use disorders (
Childhood adversities and adult psychiatric disorders in the National Comorbidity Survey Replication I: Associations with first onset of DSM-IV disorders.
The International Society for Developmental Psychobiology Sackler Symposium: Early adversity and the maturation of emotion circuits—a cross-species analysis.
). It is crucial to understand the nature of these associations because ELA and its consequences, in contrast to genetic contributors to vulnerability to psychopathologies, may be amenable to prevention. In human studies, it is difficult to demonstrate causality between ELA and adverse adult outcomes and to establish the underlying mechanisms, necessitating use of experimental animal models. Using an animal model in which ELA is induced by an impoverished environment that provokes aberrant maternal care, we have identified later life disruptions of reward-related behaviors (
Anhedonia following early-life adversity involves aberrant interaction of reward and anxiety circuits and is reversed by partial silencing of amygdala corticotropin-releasing hormone gene.
Childhood adversities and adult psychiatric disorders in the National Comorbidity Survey Replication I: Associations with first onset of DSM-IV disorders.
). These findings provide an impetus to employ this animal model for advancing our understanding of how transient ELA enduringly disrupts the operations of reward circuits.
Here, we aimed to determine which regions and neuronal populations within the brain, and specifically within reward circuits, were activated by ELA and are thus candidates for mediating the behavioral consequences of ELA. To this end, we employed TRAP2 (targeted recombination in active populations) mice expressing iCre-ERT2 recombinase at the locus of the immediate early gene, c-Fos, to genetically label neurons that are activated during ELA with the red fluorescent reporter tdTomato (tdT) using Ai14, a knock-in allele of the Rosa26 locus (
). To date, and to our knowledge, this technique had not been used during the first week of life. Using this approach, we established the paraventricular nucleus of the thalamus (PVT) as a key region engaged by these experiences.
The PVT is a dorsal midline thalamic nucleus that is a crucial component of the limbic system and emotional processing network and is engaged by emotionally salient stimuli of either valence (
). The PVT utilizes information derived from remote emotionally salient experiences to gate the expression of approach and avoidance behaviors and to influence responses to stress (
A cholecystokinin-mediated pathway to the paraventricular thalamus is recruited in chronically stressed rats and regulates hypothalamic-pituitary-adrenal function.
Collateralization of projections from the paraventricular nucleus of the thalamus to the nucleus accumbens, bed nucleus of the stria terminalis, and central nucleus of the amygdala.
) and is thus poised to regulate behaviors related to stress and reward following salient early-life experiences. Indeed, our prior work using c-Fos expression had indicated that the PVT is activated by positive early-life experience in the form of augmented maternal care (
), yet whether this applies to experiences of the opposite valence and whether this activation may influence future behaviors remain unknown.
Methods and Materials
Animals
Fos2A-iCreER (Jax #030323) and Ai14 (Jax #007914) mice were obtained from The Jackson Laboratory or bred in house. All mice were housed in a temperature-controlled, quiet, and uncrowded facility on a 12-hour light/dark schedule (lights on at 6:30 am, lights off at 6:30 pm) except for Fos2A-iCreER+/+ litters. These litters were maintained on a 12-hour reverse light cycle (lights on at midnight, lights off at noon) to enable perfusion during the early active period of the mice; we have previously found no difference in any behavioral test between normally housed and reverse light cycle–housed mice when tested at the same point of their circadian cycle. Mice were provided with ad libitum access to water and a global soy protein-free extruded diet (2020X Teklad; Envigo). Fos2A-iCreER mice bred with the Ai14 reporter mice were employed for TRAP2 studies; Fos2A-iCreER+/+ were employed for the endogenous c-Fos validation studies. All experiments were performed in accordance with National Institutes of Health guidelines and were approved by the University of California Irvine Institutional Animal Care and Use Committee.
Limited Bedding and Nesting Model of ELA
Fos2A-iCreER dams bred with Ai14 males were singly housed on embryonic day 17 and monitored for birth of pups every 12 hours. On the morning of postnatal day (P) 2, Fos2A-iCreER+/−::Ai14+/− litters and Fos2A-iCreER+/+ litters were culled to a maximum of 8 pups, including both sexes, and the ELA paradigm was initiated as previously described (
). Control dams and pups were placed in cages with a standard amount of corn cob bedding (400 mL) and cotton nestlet material (one 5 × 5 cm square). ELA dams and pups were provided with one half cotton nestlet placed on a 2.5-cm-tall, fine-gauge plastic-coated aluminum mesh platform above sparse corn cob bedding on the cage floor. Accumulation of ammonia was avoided by placing cages in a room with robust ventilation. Both rearing groups were left completely undisturbed until the morning of P6, at which point pups were briefly removed from the cage, placed on a warming pad, and injected subcutaneously with 150 mg/kg tamoxifen (catalog no. T5648; MilliporeSigma) dissolved in corn oil (catalog no. C8267; MilliporeSigma) (Figure S1). Pups were then returned to their cage and left undisturbed until the morning of P10. Dams and pups were then transferred to standard cages where maternal behaviors rapidly normalize and the ELA pups’ stress dissipates. P6 was chosen as the time point for tamoxifen injections because it is roughly at the midpoint of the ELA period and should allow for optimal tagging of neurons activated by ELA rearing.
Brain Processing and Analyses
On the morning of P14 (for Fos2A-iCreER+/−::Ai14+/− litters), or at approximately 2 pm of P10 (for Fos2A-iCreER+/+ litters), the dam was removed from the home cage, and the cage was placed on a warming pad. Pups were euthanized with sodium pentobarbital and transcardially perfused with ice cold phosphate-buffered saline (PBS) (pH = 7.4) followed by 4% paraformaldehyde in 0.1M sodium phosphate buffer (pH = 7.4). Perfused brains were postfixed in 4% paraformaldehyde in 0.1M PBS (pH = 7.4) for 4 to 6 hours before cryoprotection in 25% sucrose in PBS. Brains were then frozen and coronally sectioned at a thickness of 30 μm (1:5 series) using a Leica CM1900 cryostat (Leica Microsystems). Fos2A-iCreER+/−::Ai14+/− sections were mounted on gelatin-coated slides and coverslipped with VECTASHIELD containing DAPI (catalog no. H-1200; Vector Laboratories). P14 was chosen as the sacrifice time point for Fos2A-iCreER+/−::Ai14+/− litters because optimal expression/accumulation of the tdT reporter is not achieved until at least 1 week following the tamoxifen injection.
Immunohistochemistry
Avidin-biotin complex–amplified, diaminobenzidine (DAB) reactions were used to visualize c-Fos and CRF1 receptor (CRFR1) on free-floating sections. Sections were first washed in PBS containing 0.3% Triton (PBST) (3 × 5 minutes) followed by quenching of endogenous peroxidase activity by incubation in 0.3% H2O2 for 20 minutes. Sections were blocked in 5% normal donkey serum or normal goat serum (NGS) in PBST for 1 hour. Sections were incubated with 1:40,000 rabbit anti-c-Fos (catalog no. ABE457; MilliporeSigma) for 3 days at 4 °C or 1:2000 goat anti-CRFR1 (catalog no. EB08035; Everest Biotech) for 16 hours at 4 °C. Following 3 × 5 minute washes in PBST, sections were incubated with 1:500 biotinylated goat anti-rabbit antibody (catalog no. BA-1000-1.5; Vector Laboratories) or 1:500 biotinylated donkey anti-goat antibody (catalog no. 705-065-147; Jackson ImmunoResearch) in 2% normal donkey serum or NGS for 2 hours. Sections were washed in PBST (3 × 5 minutes) and then incubated in 1% avidin-biotin complex solution (VECTASTAIN; Vector Laboratories) and washed again in PBST (3 × 5 minutes). The immunoreaction product was visualized using solution containing 0.005% H2O2 and 0.05% DAB. Sections were mounted onto gelatin-coated slides or co-labeled for tdT expression.
For co-labeling to visualize the c-Fos reporter, tdT, benzidine dihydrochloride reactions were used following DAB staining. Sections were quenched in a solution of 50% methanol and 0.2% H2O2 in PBST for 5 minutes followed by 100% methanol containing 0.2% H2O2 for 20 minutes. Sections were washed in PBST (3 × 5 minutes) then blocked in 2% NGS for 30 minutes. Sections were then incubated with 1:10,000 rabbit anti-RFP (catalog no. 600-401-379; Rockland Immunochemicals) for 3 days at 4 °C. Following primary antibody incubation, sections were washed in PBST (3 × 5 minutes) and incubated in 1:500 biotinylated goat anti-rabbit antibody in 2% NGS for 2 hours. Sections were then washed in PBST (3 × 5 minutes) and incubated in 1% avidin-biotin complex solution, followed by additional washes in PBST (3 × 5 minutes). Sections were washed in 1× acidic buffer (3 × 5 minutes) (catalog no. 003850; Bioenno Lifesciences), then incubated in a buffer containing 0.025% sodium nitroprusside and 0.01% to 0.02% benzidine dihydrochloride for 5 to 10 minutes. The granular blue deposits were visualized by immersing the sections in fresh incubation solution containing 0.003% H2O2 for 3 minutes. Sections were washed in 1× acidic buffer (3 × 5 minutes) and mounted onto gelatin-coated slides. All mounted sections were dehydrated and coverslipped with Permount mounting medium (catalog no. SP15-500; Fisher Scientific).
For fluorescent labeling of corticotropin-releasing factor (CRF), the tyramide signal amplification technique was used (
). Free-floating sections were blocked in 5% NGS in PBST for 1 hour and then incubated in rabbit anti–corticotropin-releasing hormone (CRH) antiserum (1:20,000) (gifted by Dr. W. Vale, Salk Institute, La Jolla, California) for 14 days at 4 °C. Following washing in PBST (3 × 5 minutes), sections were incubated in horseradish peroxidase–conjugated anti-rabbit IgG (1:1000) (catalog no. NEF812001EA; PerkinElmer) for 1.5 hours. Fluorescein tyramide conjugate was diluted in an amplification buffer (1:150) (catalog no. NEL701A001KT; PerkinElmer) and applied to sections in the dark for 5 to 6 minutes, followed by washing and mounting on gelatin-coated slides. Sections were coverslipped with VECTASHIELD containing DAPI.
Image Acquisition
Images of sections processed with DAB and benzidine dihydrochloride were collected using a Nikon Eclipse E400 light microscope (Nikon) with 10× and 20× objective lenses. Confocal images were collected using an LSM-510 confocal microscope (Zeiss) with an apochromatic 10×, 20×, or 63× objective. Virtual z sections of 1 μm were taken at 0.2- to 0.5-μm intervals. Image frame was digitized at 12 bit using a 1024 × 1024 pixel frame size.
Analyses and Statistical Considerations
tdT+, CRFR1+, CRH+, and c-Fos+ neuron numbers were counted manually in Fiji (
). All quantifications and analyses were performed using stereological principles including systematic unbiased sampling and without knowledge of group assignment. Statistical analyses were carried out using GraphPad Prism (GraphPad Software). Differences between control and ELA groups of both sexes were assessed using two-way analysis of variance (ANOVA). To examine significance of cell number differences throughout the anteroposterior axis of the PVT, we used two-way mixed ANOVA with the Sidak multiple comparisons post hoc test.
Results
TRAP2 System Is Feasible and Effective in Neonatal Mice
We used TRAP2;Ai14 mice expressing iCre-ERT2 recombinase in an activity-dependent manner to initiate expression of an iCre-dependent tdT reporter. To initiate recombination and reporter expression, tamoxifen was administered to P6 pups reared in standard or ELA cages, and reporter expression was assessed 1 week later to allow for optimal accumulation of the reporter. Modest, consistent tdT expression was detected in cell bodies in several brain regions in ELA mice, including the hypothalamic paraventricular and suprachiasmatic nuclei (Figure 1). The most robust and striking reporter expression, indicative of neuronal activation during the P6 to P8 time period, was identified in the PVT.
Figure 1PVT neurons are robustly and selectively activated during exposure to adverse rearing conditions. Low magnification images of brain sections from TRAP2 mice sacrificed a week after receiving tamoxifen on postnatal day 6 to enable c-Fos-dependent Cre expression for 24–48 hours. Strong activation of the PVT, with little reporter expression in the rest of the brain, is apparent. aPVT, anterior PVT; MPO, medial preoptic nucleus; pPVT, posterior PVT; PVN, paraventricular nucleus of the hypothalamus; PVT, posterior paraventricular nucleus of the thalamus; SCN, suprachiasmatic nucleus.
). Here, we sought to determine if PVT activation occurred during emotionally salient experiences in the neonatal/infancy period. To this end, we reared mice either in typical cages or in our ELA model. In this simulated poverty model, mouse pups are reared in cages with limited bedding and nesting materials, conditions that provoke aberrant maternal care behaviors and stress in the pups (
). We administered tamoxifen to both control and ELA pups on P6, which results in induction of reporter expression in all neurons activated during a 24- to 36-hour time window following tamoxifen administration (
There was very little neuronal activation, measured by number of neurons expressing tdT, throughout the brains of P6 to P8 male and female mice (Figure 1). However, a screen of serial coronal sections throughout the brain suggested that prominent c-Fos expression took place in the PVT in both control and ELA mice. Analyzing male mice, a two-way ANOVA with the Sidak post hoc test identified significantly greater overall PVT activation in ELA male compared with control male mice (p = .002) (Figure 2A, B, E). A two-way mixed model ANOVA with coordinate as a repeated factor identified a main effect of rearing on reporter expression (p < .001), and the Sidak multiple comparisons post hoc test indicated a significantly larger number of neurons that had undergone cre-mediated recombination (“TRAPed” neurons, tdT+) in ELA male compared with control male mice at several coordinates along the anteroposterior axis of the PVT (at −0.46 mm, −1.22 mm, −1.46 mm, and −1.7 mm from bregma), indicating that this differential activation was particularly prominent in the mid- to posterior PVT (Figure 2H). Comparison of overall PVT activation between control male and control female mice demonstrated a greater number of TRAPed PVT neurons in control females mice (p = .019), and this was particularly prominent in the posterior PVT (−1.94 mm from bregma; p = .007; two-way mixed model ANOVA with the Sidak post hoc test) (Figure 2E, F). Strikingly, and in contrast to male mice, ELA did not further augment the density of TRAPed cells in the PVT of female mice (Figure 2C, D, I).
Figure 2ELA induces greater PVT activation in male mice compared with female mice. (A) CTL and (B) ELA postnatal day 14 male TRAP2 mouse reporter (tdT) expression in the posterior PVT following tamoxifen administration on postnatal day 6. (C) CTL and (D) ELA postnatal day 14 female TRAP2 mouse tdT expression in the posterior PVT following tamoxifen administration on postnatal day 6. Scale bar = 50 μm. (E) Comparison of overall PVT activation in male and female CTL and ELA mice, normalized to activation in CTL male mice (n = 7–9 mice/group; two-way mixed model ANOVA with the Sidak post hoc test). (F) Quantification of tdT+ neurons across the AP axis of the PVT comparing CTL male and female mice. Two-way mixed model ANOVA with the Sidak post hoc test shows significantly more TRAPed neurons at −1.94 mm from bregma in female mice (n = 7–10 mice/group; p = .007). (G) Comparing the number of tdT+ PVT neurons in ELA male and female mice indicates no difference (n = 8–14 mice/group; p = 0.296; two-way mixed model ANOVA). (H) Two-way mixed model ANOVA with the Sidak post hoc test comparing CTL and ELA male mice indicates significantly more tdT+ PVT neurons at −0.46 mm, −1.22 mm, −1.46 mm, and −1.70 mm from bregma (n = 10–14 mice/group; two-way mixed model ANOVA with the Sidak post hoc test). (I) Comparing the number of tdT+ PVT neurons in CTL and ELA female mice indicates no difference (n = 7–8 mice/group; p = .472; two-way mixed model ANOVA). ∗p < .05; ∗∗p < .01; ∗∗∗p < .001. Data are presented as mean ± SEM values. A, anterior; ANOVA, analysis of variance; AP, anteroposterior; CTL, control; ELA, early-life adversity; P, posterior; PVT, paraventricular nucleus of the thalamus; tdT, tdTomato; TRAP, targeted recombination in active populations.
Reporter expression in TRAP2 mice is driven by the c-Fos promoter and aims to reflect c-Fos expression as a marker of neuronal activation. Because the use of the TRAP2 transgenic system during the first week of life has not yet been published, we determined the validity of this method during the neonatal period by visualizing native c-Fos expression in the PVT during control and ELA rearing conditions. Immunolabeling against c-Fos in the PVT of P10 mice was congruent with that of the TRAP2 reporter (Figure 3). As expected, significantly more c-Fos+ neurons were found in the PVT of ELA male compared with control male mice. In accord with the TRAP2 reporter expression, we found no difference in number of c-Fos+ neurons when comparing control female and ELA female mice. These data support the conclusion that the use of the TRAP2 system in early life produces reporter expression reflective of native c-Fos expression during this period.
Figure 3Endogenous c-Fos expression in the PVT is congruent with and validates the TRAP2 method. (A) Representative images of c-Fos expression in the posterior PVT of postnatal day 10 CTL male mice, (B) ELA male mice, (C) CTL female mice, (D) and ELA female mice sacrificed immediately following exposure to normal or adverse rearing conditions; Scale bar = 50 μm. (E) Comparison of overall PVT c-Fos expression in male and female CTL and ELA mice, normalized to activation in CTL male mice (n = 5–6 mice/group; two-way mixed model analysis of variance with the Sidak post hoc test). (F) Quantification of c-Fos+ neurons across the AP axis of the PVT in postnatal day 10 CTL and ELA male mice (G) and female mice (n = 7–11 mice/group; two-way mixed model analysis of variance). ∗p < .05; ∗∗p < .01. Data are presented as mean ± SEM values. A, anterior; AP, anteroposterior; CTL, control; ELA, early-life adversity; P, posterior; PVT, paraventricular nucleus of the thalamus; 3rd, third ventricle.
Increased Neuron Activation by ELA in Males Is Selective to the PVT
Quantification of TRAPed neurons in control and ELA male mice was performed in several additional brain regions, including regions involved in reward and responses to stress, such as the nucleus accumbens, amygdala, ventral tegmental area, and paraventricular nucleus of the hypothalamus. This quantification largely revealed sparse reporter expression in both rearing groups, suggesting little activation during P6 to P8 in most regions outside of the PVT (Figure 4; Table 1). Importantly, in all areas analyzed besides the PVT, reporter expression did not differ significantly between ELA male and control male mice (Table 1). These findings indicate that the augmented number of TRAPed neurons in the PVT of ELA male compared with control male mice is specific to this nucleus.
Figure 4The numbers of TRAPed neurons differ between CTL and ELA male mice exclusively in the PVT. Representative images for multiple brain regions show no difference in tdTomato expression in male TRAP2 mice reared in CTL vs. ELA conditions. Few tdTomato+ neurons are found in most regions, including those important in reward- and stress-related behaviors. White arrow heads indicate tdTomato+ neurons in the NAc shell and core. Coordinates indicate the distance from bregma of each image. Scale bars = 100 μm. BLA, basolateral amygdala; CeA, central nucleus of the amygdala; CTL, control; ELA, early-life adversity; NAc, nucleus accumbens; PVN, paraventricular nucleus of the thalamus; SCN, suprachiasmatic nucleus; TRAP, targeted recombination in active populations; 3rd, third ventricle.
Table 1Region-Specific Quantification of tdTomato+ Neurons in Male Postnatal Day 14 TRAP2 Mice TRAPed During Normal Rearing or Adverse Rearing Conditions
Brain Region
Control TRAPed Neurons
ELA TRAPed Neurons
p Value
Paraventricular Thalamus
Anterior
79.31 [60.74, 97.88]
133.60 [116.80, 150.40]
.0001
Mid
39.33 [31.39, 47.27]
57.65 [46.73, 68.57]
.008
Posterior
80.31 [59.56, 101.06]
145.10 [109.70, 180.50]
.005
Nucleus Accumbens
Core
9.16 [6.67, 11.65]
7.76 [5.18, 10.34]
.397
Medial shell
1.73 [1.08, 2.38]
1.63 [0.90, 2.36]
.823
Amygdala
Basolateral
17.00 [13.01, 20.99]
13.61 [8.37, 18.85]
.278
Central
11.29 [5.11, 17.47]
8.90 [2.59, 15.21]
.556
Medial Prefrontal Cortex
Prelimbic
5.29 [4.26, 6.32]
6.88 [1.64, 12.12]
.521
Infralimbic
7.13 [3.49, 10.77]
6.14 [3.19, 9.09]
.631
Ventral tegmental area
4.80 [2.26, 7.34]
6.11 [2.26, 9.96]
.516
Hypothalamus
Suprachiasmatic
56.00 [46.56, 65.44]
65.57 [54.76, 76.38]
.137
Paraventricular
21.67 [13.75, 29.59]
25.58 [16.24, 34.92]
.489
Lateral
12.14 [5.11, 19.17]
10.43 [4.96, 15.90]
.646
Ventromedial
6.71 [2.65, 10.77]
3.43 [0.82, 6.04]
.121
Hippocampus
CA1
14.33 [8.64, 20.02]
11.75 [8.80, 14.70]
.301
CA3
10.17 [6.69, 13.65]
10.38 [6.56, 14.2]
.926
Dentate gyrus
20.50 [10.53, 30.47]
19.13 [2.20, 26.06]
.778
Values are presented as mean [95% confidence interval lower limit, 95% confidence interval upper limit].
ELA, early-life adversity; TRAP, targeted recombination in active populations.
Molecular/Phenotypic Characterization of ELA-TRAPed PVT Neurons
The PVT is a heterogeneous structure composed of numerous cell types that are defined by the expression of distinct neurotransmitters, neuropeptides, and receptors. Hence, we next sought to determine the molecular characteristics of PVT neurons engaged by early-life experiences, including ELA, to gain insight into the potential functional roles of this population. In addition, because ELA intrinsically involves stress, and neurons expressing the stress-related neuropeptide CRH are impacted by ELA in the hypothalamus (
Single-cell transcriptional changes in hypothalamic corticotropin-releasing factor–expressing neurons after early-life adversity inform enduring alterations in vulnerabilities to stress.
Anhedonia following early-life adversity involves aberrant interaction of reward and anxiety circuits and is reversed by partial silencing of amygdala corticotropin-releasing hormone gene.
), we focused on PVT neurons expressing CRH or its receptors. Immunostaining against CRH demonstrated a rich population of CRH-expressing neurons in the PVT, in accord with prior reports (
New viral-genetic mapping uncovers an enrichment of corticotropin-releasing hormone-expressing neuronal inputs to the nucleus accumbens from stress-related brain regions.
). However, no TRAPed PVT neurons expressed CRH (Figure 5A, B). In contrast, immunostaining against the type 1 receptor to CRH, CRFR1, demonstrated a major increase in the proportion of activated PVT neurons expressing CRFR1 in ELA mice compared with control mice. Specifically, in male mice, 40.8% of TRAPed PVT neurons expressed this receptor in ELA mice (95% CI [0.349, 0.468]), whereas 20.3% of TRAPed PVT neurons expressed this receptor in control mice (95% CI [0.151, 0.255]). In female mice, 50.5% of TRAPed PVT neurons expressed CRFR1 in ELA mice (95% CI [0.441, 0.569]), and 24.5% of TRAPed PVT neurons expressed CRFR1 in control mice (95% CI [0.185, 0.304]). This effect was particularly prominent in the anterior and mid-PVT in female mice and across the entire anteroposterior axis of the PVT in male mice (Figure 5E, F). Therefore, in both male and female mice, a significantly larger proportion of neurons engaged in early life expressed CRFR1 in ELA groups compared with control groups. Additionally, a two-way mixed model ANOVA comparing the proportion of TRAPed anterior PVT (aPVT) neurons that express CRFR1 in control mice and ELA male and female mice indicated a significant interaction between rearing condition and sex (p = .002) (Figure S2), with female mice experiencing a larger ELA-induced increase in proportion of TRAPed aPVT neurons expressing this receptor. Surveys of CRFR2 expression in PVT using both in situ hybridization (
) and immunohistochemistry revealed only low levels, so this receptor was not quantified.
Figure 5TRAPed PVT neurons express CRFR1, but not CRH. (A) Representative image from the pPVT of tdTomato (red) and CRH (green) expression in a postnatal day 14 TRAP2 male mouse demonstrating lack of overlap between CRH and the TRAP2 reporter. White box indicates the portion of the section shown at higher magnification in the next panel. Scale bar = 50 μm. (B) Higher magnification reveals numerous CRH+ puncta around TRAPed PVT neurons (white arrows), but no CRH expression within TRAPed neurons. Scale bar = 10 μm. (C) Representative image from the pPVT of tdTomato (blue) and CRFR1 (brown) expression in a postnatal day 14 TRAP2 mouse, showing robust overlap of CRFR1 and reporter expression. Black box indicates the portion of the section shown at higher magnification in the next panel. Scale bar = 50 μm. (D) Higher magnification reveals partial coexpression of tdTomato and CRFR1 (black arrowheads). Scale bar = 10 μm. (E) In male mice, across aPVT, mPVT, and pPVT, a significantly larger proportion of TRAPed PVT neurons express CRFR1 following ELA compared with CTL rearing (n = 6 mice/group; two-way mixed model analysis of variance with the Sidak post hoc test). (F) In female mice, across anterior and mid PVT, a significantly larger proportion of TRAPed PVT neurons express CRFR1 following ELA compared with CTL rearing (n = 5–6 mice/group; two-way mixed model analysis of variance with the Sidak post hoc test). ∗p < .05; ∗∗p < .01. Data are presented as mean ± SEM values. aPVT, anterior PVT; CRFR1, CRF1 receptor; CRH, corticotropin-releasing hormone; CTL, control; ELA, early-life adversity; mPVT, mid-PVT; pPVT, posterior PVT; PVT, posterior paraventricular nucleus of the thalamus; TRAP, targeted recombination in active populations.
The principal findings in this set of experiments are the following: 1) genetic tagging of neurons activated during the neonatal period in mice is feasible, with high sensitivity and fidelity; 2) the PVT is the major brain region activated by ELA; 3) sex is an important determinant of neuronal activation by early-life experiences; and 4) neurons expressing the CRH receptor, CRFR1, likely a target of CRH signaling, are preferentially activated by ELA in a sex-dependent manner and are poised to contribute to the mechanisms by which ELA contributes to alterations in adult behaviors.
Using the TRAP2 transgenic mouse, we identified here region-specific neuronal activation during the early postnatal period in the mouse. Reporter expression is highly congruent with native c-Fos. In contrast to native c-Fos expression, the TRAP2 system allows for labeling of neuronal activity over a much longer time period (up to approximately 36 hours) following tamoxifen administration (
). This characteristic is advantageous when visualizing neuronal activity during a chronic stimulus, such as ELA. The finding that activity-dependent genetic labeling of neurons in P6 mice is possible is novel and demonstrates that c-Fos can be robustly expressed in the brains of neonatal mice. This transgenic system can therefore be an effective and advantageous tool for the investigation of early life neuronal activation within the brain and its consequences later in life.
An important consideration when pursuing activity-dependent labeling is the particular gene locus governing Cre expression. Immediate early genes represent a well-described connection between neuronal activity and subsequent gene expression changes (
) and thus provide a useful strategy for targeting active cell populations for genetic access. While several immediate early genes are expressed in the brain, c-Fos is known to be expressed in the neonatal rodent brain, and this expression in the PVT is dependent on an ongoing stimulus (
). In contrast to other immediate early genes that act rapidly via direct influences on synapses and cellular function, c-Fos functions through more protracted pathways via regulation of downstream target genes (
). Additionally, the role of c-Fos in memory is well established, including a role in mediating responses following acquisition of contextual memories (
). While c-Fos is expressed following activation in both excitatory and inhibitory neurons and therefore cannot distinguish between the two, the vast majority of the PVT neurons are excitatory glutamatergic neurons (
). Therefore, the lack of distinction between excitatory and inhibitory neurons is not of consequence in this context. These characteristics highlight c-Fos as an excellent tool for understanding brain activity in early life.
Using TRAP2, we found that the PVT is prominently activated during exposure to ELA compared with normal rearing in male mice. By contrast, in female mice, there is little additional apparent neuronal activation in ELA versus control groups. In both sexes, the selectivity of PVT engagement by early-life experiences is striking, as other brain regions related to stress and reward contain few experience-engaged neurons, and this activation does not differ between control and ELA mice.
In adult rodents, the PVT is activated by stimuli of both positive valence (e.g., drugs of abuse) (
), but activation of the PVT by emotionally salient stimuli early in life is not well characterized. We have previously identified PVT activation in 9-day-old rat pups by sensory input from their mothers, a positive emotionally salient experience (
). Here, we show that the PVT is activated by ELA occurring during postnatal days 2 to 9 in mice. This is important, as adult studies demonstrate that activation of the PVT by stressful events influences responses to an additional stress later in life (
). Our demonstration that the PVT is selectively engaged by ELA renders this brain region a candidate for mediating the long-lasting consequences of ELA on future responses to stress throughout life.
). Activation of subsets of PVT neurons may mechanistically contribute to these consequences: the PVT is required for the retrieval of remote emotionally salient experiences (i.e., those that have occurred >24 hours ago) and their influence on the regulation of motivated behaviors in adult rodents (
Oxytocin receptor-expressing neurons in the paraventricular thalamus regulate feeding motivation through excitatory projections to the nucleus accumbens core.
The paraventricular thalamus provides a polysynaptic brake on limbic CRF neurons to sex-dependently blunt binge alcohol drinking and avoidance behavior in mice.
). Disruptions in similar reward-related behaviors characterize numerous mood disorders, including substance use disorders and depression, for which ELA is a risk factor (
Childhood adversities and adult psychiatric disorders in the National Comorbidity Survey Replication I: Associations with first onset of DSM-IV disorders.
). Thus, the robust and differential engagement of the PVT during ELA reported here suggests that the PVT may encode these early-life experiences and use this information later in life to impact reward-related behaviors. These hypotheses will be subjects of future studies.
We found that ELA exerts the largest increase in activation in specific subregions of the PVT. The specific topology of this differential activation sheds light on the potential connectivity and functions of ELA-engaged PVT neurons: the posterior PVT (pPVT) sends strong projections to regions including the ventromedial nucleus accumbens shell, central amygdala, basolateral amygdala, and bed nucleus of the stria terminalis, whereas the aPVT projects in a more diffuse manner to regions including the dorsomedial nucleus accumbens shell, suprachiasmatic nucleus, and ventral subiculum (
). In addition to this heterogeneity in projection patterns, different contributions to motivated behaviors are attributed to the aPVT versus pPVT. For example, inactivation of the aPVT, but not pPVT, decreases sucrose seeking when an expected sucrose reward is omitted (
A cholecystokinin-mediated pathway to the paraventricular thalamus is recruited in chronically stressed rats and regulates hypothalamic-pituitary-adrenal function.
). Therefore, the distinct topological distribution of neuronal activation by ELA may suggest specific roles and projection targets that warrant further investigation.
Our finding that CRFR1-expressing PVT neurons are preferentially activated by ELA is intriguing. The receptor is well expressed during the first week of life in the rodent (
) and was abundantly expressed in the PVT of both male and female mice in our findings. Why and how might ELA augment activation of CRFR1 neurons? As shown in other brain regions, ELA might increase CRH expression and release in the PVT, activating receptor-expressing neurons. CRH release regulates CRFR1 expression in a biphasic manner (
Corticotropin-releasing hormone (CRH) downregulates the function of its receptor (CRF1) and induces CRF1 expression in hippocampal and cortical regions of the immature rat brain.
), such that it is conceivable that more cells in the PVT express CRFR1 above detection threshold in ELA versus control animals. The augmented activation of CRFR1 neurons by ELA is notable because CRFR1 plays important roles in responses to stress, including mediating changes in reward-related behaviors following stress (
Enhanced CRFR1-dependent regulation of a ventral tegmental area to prelimbic cortex projection establishes susceptibility to stress-induced cocaine seeking.
). Further analysis revealed a significant interaction between rearing condition and sex in the aPVT, whereby the effect of ELA on proportion of TRAPed aPVT neurons that express CRFR1 was significantly more pronounced in female compared with male mice (Figure S2). This, combined with the finding that overall PVT activation was greater in control female compared with control-reared male mice, suggests sex-dependent mechanisms of PVT activation in early life. Future studies will determine whether these CRFR1 neurons mediate the influence of ELA on adult reward behaviors.
In conclusion, the present studies describe a novel use of activity-dependent genetic tagging to demonstrate robust and selective activation of the PVT during ELA. Compared with typical rearing, PVT neurons activated during ELA are significantly more likely to express the receptor to CRH, CRFR1. Establishing the functional role of these ELA-engaged PVT neurons will be a crucial next step toward determining their role in disruptions of reward-related behaviors induced by ELA and will provide important information toward understanding the mechanisms underlying the consequences of ELA on mental health.
Acknowledgments and Disclosures
This work was supported by the National Institutes of Health (Grant Nos. F30 MH126615 [CLK], T32 DA050558 [CLK], T32 GM008620 [CLK], and P50 MH096889 [YC, MTB, TZB]).
We thank Qinxin Ding and Manlin Shao for technical assistance.
The authors report no biomedical financial interests or potential conflicts of interest.
Childhood adversities and adult psychiatric disorders in the National Comorbidity Survey Replication I: Associations with first onset of DSM-IV disorders.
The International Society for Developmental Psychobiology Sackler Symposium: Early adversity and the maturation of emotion circuits—a cross-species analysis.
Anhedonia following early-life adversity involves aberrant interaction of reward and anxiety circuits and is reversed by partial silencing of amygdala corticotropin-releasing hormone gene.
A cholecystokinin-mediated pathway to the paraventricular thalamus is recruited in chronically stressed rats and regulates hypothalamic-pituitary-adrenal function.
Collateralization of projections from the paraventricular nucleus of the thalamus to the nucleus accumbens, bed nucleus of the stria terminalis, and central nucleus of the amygdala.
Single-cell transcriptional changes in hypothalamic corticotropin-releasing factor–expressing neurons after early-life adversity inform enduring alterations in vulnerabilities to stress.
New viral-genetic mapping uncovers an enrichment of corticotropin-releasing hormone-expressing neuronal inputs to the nucleus accumbens from stress-related brain regions.
Oxytocin receptor-expressing neurons in the paraventricular thalamus regulate feeding motivation through excitatory projections to the nucleus accumbens core.
The paraventricular thalamus provides a polysynaptic brake on limbic CRF neurons to sex-dependently blunt binge alcohol drinking and avoidance behavior in mice.
Corticotropin-releasing hormone (CRH) downregulates the function of its receptor (CRF1) and induces CRF1 expression in hippocampal and cortical regions of the immature rat brain.
Enhanced CRFR1-dependent regulation of a ventral tegmental area to prelimbic cortex projection establishes susceptibility to stress-induced cocaine seeking.
00002-2&id=gr1.jpg){kind=link}
00002-2&id=gr2.jpg){kind=link}
00002-2&id=gr3.jpg){kind=link}
00002-2&id=gr4.jpg){kind=link}
00002-2&id=gr5.jpg){kind=link}