Corticotropin releasing hormone-expressing inputs to the thalamic paraventricular nucleus: pathways from mother to memory?

TL;DR

This study explores how stress-related CRH reaches brain cells involved in early life stress and memory formation.

Contribution

The paper identifies specific brain regions and pathways that supply CRH to the thalamic PVT, linking maternal sensory input to stress signaling.

Findings

CRH-expressing neurons are found within and around the PVT, including the paratenial nucleus.

CRH neurons in the parabrachial and Barrington nuclei project to the PVT.

These pathways may transmit maternal sensory signals to the PVT, as proposed by Seymour Levine.

Abstract

Thalamic paraventricular nucleus (PVT) neurons expressing the corticotropin releasing hormone receptor 1 (CRHR1) are preferentially activated during early life stress. However, it is unclear how the receptor ligand, the stress-related peptide CRH, reaches receptor-bearing cells. To address this question, we mapped local, proximal and distal sources of CRH, i.e., CRH expressing neurons within, adjacent to and projecting to PVT. The combined use of retrograde and anterograde viral-genetic tracing approaches, validated with immunohistochemistry, identified an array of CRH neurons within PVT, in the adjacent paratenial nucleus as well as projecting to PVT from the parabrachial and Barrington nuclei. The latter are poised to convey to the PVT sensory signals from maternal grooming, as envisioned by Seymour Levine in the 1950s.

Figures

Fig. 5Dual anterograde/retrograde tracing of PBN CRH+ neurons demonstrate the existence of the projection. (A) Schematic of administering a CRE-dependent anterograde AAV1 (green) into the parabrachial nucleus in combination with a CRE-independent retrograde AAV2 (red) into the pPVT or aPVT of a CRH-CRE mouse. (B) Injection site in the parabrachial nucleus, showing successful labeling of CRH-CRE expressing neurons with EGFP. (C–D) In the same mice, the aPVT ((C), n=4 males) or pPVT ((D), n= 4 males, 2 females) was injected with a CRE-independent retrograde AAV. (E–F) EGFP-labeled axons from the

Full text

Introduction

1

This work focuses on the mechanisms by which experiences, i.e., signals from the environment, experienced early in life reach key brain regions to influence emotions and behaviors later in life. Experiences early in life are critical for circuit maturation and influence behaviors such as the response to stress and reward later in life (Levine, 2005; Birnie and Baram, 2022). In humans, early life adversity (ELA), such as abuse, neglect, or poverty, is associated with higher risk for developing mental health disorders such as depression, anxiety and addiction (Hackman and Farah, 2009; Green et al., 2010; Silvers et al., 2017; Nelson et al., 2025). Rodent models provide evidence for a causal role of ELA. The widely adopted, created by our group, limited bedding and nesting (LBN) paradigm, simulates poverty and stresses the dam, disrupting the patterns of maternal care (Walker et al., 2017; Naninck et al., 2015; Wang et al., 2011; Peña et al., 2019). Mice and rats raised in this paradigm during postnatal days 2-9 display sex-specific changes in stress and reward behaviors in adulthood (Levis et al., 2021a; Birnie et al., 2023; Dixon et al., 2025). Male rodents display anhedonic behavior when presented with rewards (Levis et al., 2021a; Birnie et al., 2023), while females show enhanced motivation for rewards, a “pro-hedonic” phenotype (Levis et al., 2021a; Kooiker et al., 2024; Taniguchi et al., 2025). To understand the mechanisms involved in ELA-induced brain changes and behavior, we first need to identify the brain areas that are directly activated by ELA and might encode the ELA experience. To this end, we recently used activity-dependent genetic tagging in mice during ELA. Specifically, we used Targeted Recombination in Active Populations (TRAP) and identified the thalamic paraventricular nucleus (PVT) as one of several brain areas prominently activated during ELA (Kooiker et al., 2023). Importantly, the PVT was the only area that distinguished ELA from CTL rearing in terms of number of activated cells. The PVT is a key element of the reward and stress circuits and is believed to play an adjudicating role during motivational conflict (Do-monte et al., 2015; Millan et al., 2017; Keyes et al., 2020). The PVT receives stress-related inputs from the midbrain and hindbrain, and sends wide projections throughout the limbic system, thereby contributing to how rodents remember and respond to salient experiences (Bhatnagar and Dallman, 1998; Yeh et al., 2021; Engelke et al., 2021; Kirouac, 2025).

The PVT receives a variety of neuromodulators that continuously regulate its function (Gao et al., 2023; Kirouac, 2025). One key neuromodulator with a potentially crucial role in the PVT is the corticotropin releasing hormone (CRH), a neuropeptide that contributes to both stress and reward behaviors (Joëls and Baram, 2009; Berridge and Dunn, 1989; Lemos et al., 2012; Dedic et al., 2018; Birnie et al., 2023). As mentioned above, within PVT up to 60% of cells activated by ELA express the CRH receptor type 1 (CRHR1), suggesting that CRH release in the PVT, from either local or distant neurons, may contribute to the mechanisms by which ELA influences reward and stress functions. Indeed, the PVT is densely innervated by CRH-positive axons that could originate within or outside of the PVT (Kooiker et al., 2023). CRH-expressing neurons have been identified within the PVT itself and are critical for modulating motivational conflicts during decision making (Engelke et al., 2021). Furthermore, ELA disrupts CRH expression patterns, leading to significant deficits in memory and stress regulation (Ivy et al., 2010; Gunn et al., 2013; Chen and Baram, 2016). Understanding the origin of these CRH inputs is crucial to understanding how the PVT contributes to the long-lasting effects of ELA.

CRH inputs to the PVT have been characterized in rats and were found to originate primarily from the central amygdala, parabrachial nucleus and the Barrington nucleus, but much less is known about CRH inputs to the mouse PVT (Otake and Nakamura, 1995). Here we use a combination of anterograde and retrograde viral tracers and the Cre recombinase system to systematically identify, brain-wide, CRH-expressing afferent projections to the mouse PVT. Characterizing these areas and their connections to the PVT could provide valuable insight into the mechanisms through which ELA changes the brain.

Methods

2

The objective of this study was to characterize CRH-expressing afferent projections to the PVT. We thus targeted the PVT via both anterograde and retrograde tracing and analyzed for potential projections from 18 established CRH-expressing brain regions (Table 1).Table 1Quantitative analysis of afferent projections to PVT from 18 CRH-expressing brain regions. Numbers represent the range of retrogradely-labeled CRH-expressing neurons found in each region, across all mice.Table 1. Regions with CRH-expressing neuronsAnterior PVT (n = 4; 2 males, 2 females)Posterior PVT (n = 6; 4 males, 2 females)Main olfactory bulb0 to 10 to 1Prelimbic area0 to 10Infralimbic area00Nucleus accumbens00Bed nucleus of the stria terminalis00Substantia innominata00Central amygdala00Basolateral amygdala00Basomedial amygdala0 to 110 to 2Nucleus of reuniens00Lateral hypothalamic area00 to 3Paraventricular nucleus of the hypothalamus00Subiculum00Periaqueductal gray00Parabrachial nucleus8 to 336 to 22Barrington nucleus2 to 61 to 6Nucleus of the solitary tract00Inferior olivary complex00

Experimental animals

2.1

Ai14 (B6.Cg-Gt(ROSA)26Sor^tm14(CAG-tdTomato)Hze^/J; Jax #007914, MGI:J:155793), CRH-CRE (B6(Cg)-Crh^tm1^(cre)Zjh/J; Jax #012704, MGI:J:177261), TRAP2 (Fos^tm2.1(icre/ERT2)Luo^/J; Jax #030323, MGI:J:250613), and NuTRAP (B6; 129S6-Gt(ROSA)26Sor^tm2(CAG-NuTRAP)Evdr^/J; Jax #029899, MGI:J:347707) mice were purchased from Jackson Laboratories and bred in-house. Crhr1-FlpO mice were generated in collaboration with the lab of Dr. Nicholas J. Justice (Hardy et al., 2024) and were also bred in-house. All mice were housed with their littermates under standard conditions at 72 °F and 42% humidity on a 12-h light–dark cycle (Lights ON: 0700; Lights OFF: 1900). All experiments were conducted according to the National Institutes of Health guidelines for animal care and use and were approved by the Institutional Animal Care and Use Committee (AUP-24-104) of the University of California-Irvine (UCI).

Limited bedding and nesting (LBN) and activity-dependent genetic tagging with TRAP

2.2

ELA was induced with our limited bedding and nesting paradigm (LBN) during postnatal days (PN) 2 to 9 (Molet et al., 2016; Walker et al., 2017). In LBN cages the corncob bedding was reduced to a thin layer, and the nesting material was reduced to half. LBN cages were fitted with a fine-gauge plastic-coated aluminum mesh (cat #4700313244; McNichols Co., Tampa, FL) elevated approximately 2.5 cm from the cage floor. To induce genetic labeling of activated cells on postnatal day (P)6, TRAP2 x Ai14 or TRAP2 x NuTRAP pups were briefly separated from the dams and placed on a heating pad to subcutaneously administer 60 mg/kg of tamoxifen (Sigma; #T5648; MilliporeSigma, St. Louis, MO) dissolved in corn oil (Sigma; #C8267, MilliporeSigma, St. Louis, MO).

Viruses and surgeries

2.3

Adult mice (PN60–PN180) were first anesthetized with isoflurane (3-5%) with a 1.0L/min oxygen flow rate using an isoflurane tabletop unit. Mice were then placed onto the headpiece of the robotic stereotaxic apparatus, with continuous 1.5% isoflurane anesthesia. The mice were given 0.1 mg/kg of buprenorphine and, under aseptic conditions, an incision was made to expose the skull and estimate the injection site coordinates according to the position of the bregma and lambda.

For single retrograde tracing experiments, CRH-CRE mice received 200 nl unilateral injections of AAV2rg-hSyn-DIO-EGFP (AAVrg; #50457, Addgene, Waterloo, MA) into the left hemisphere of the anterior PVT (AP: 0.40, ML: 0.37, DV: 3.55) or posterior PVT (AP: 1.50, ML: 0.30, DV: 2.90) at a 6° angle. For dual AAV tracing experiments, CRH-CRE mice received 100 nl unilateral injections of AAV2rg-CAG-tdTomato (AAVrg; #59462, Addgene, Waterloo, MA) into either the left anterior or posterior PVT, and 100 nl of anterograde AAV1-hSyn-DIO-EGFP (AAV1; #50457, Addgene, Waterloo, MA) into the left hemisphere of the parabrachial nucleus (AP: 5.00, ML: 1.40, DV: 3.60 at a 0° angle).

For visualizing the dendrites of CRHR1-expressing neurons, Crhr1-FlpO mice received 200 nl of AAV-Ef1a-Coff/Fon-GCaMP6F (AAV8; #137124, Addgene, Waterloo, MA) in the left hemisphere of the anterior PVT (AP: 0.40, ML: 0.50, DV: 3.40) at an 8° angle.

Once the virus was administered, the micropipette was left in the injection site for at least 10 min and slowly retracted. The surgical site was closed using tissue adhesive (Vetbond; 6804221, 3M, St. Paul, MN). The mice were monitored for five days post-operatively. Mice given AAVs expressing EGFP or tdTomato were perfused six weeks after surgery to visualize the fluorophores and reduce the need of immunohistochemistry (IHC) for fluorescence amplification. Mice given AAV-Ef1a-Coff/Fon-GcaMP6F required fluorescence amplification. The injection site was assessed in all mice and mice with suboptimal virus expression (viral expression outside the targeted region or no viral expression in the targeted region) were excluded from analysis.

Perfusion and sectioning

2.4

TRAP2 x Ai14 mice were perfused at P14, TRAP2 x NuTRAP mice were perfused at P30, and CRH-CRE x Ai14 mice were perfused at P56. Mice were deeply anesthetized with a lethal dose of Euthasol (# ANDA 200-071, Virbac, Westlake, Texas) administered intraperitoneally. Then they were perfused transcardially via the left ventricle with phosphate buffered saline (PBS) followed by 4% paraformaldehyde (PFA) in 0.1 M phosphate buffer (PB, pH 7.4, 4 °C). Brains were then extracted and post-fixed in 4% PFA overnight or for 4 h for experiments involving IHC at 4 °C. Following post-fixation, the brains were immersed in 30% sucrose for two days for cryoprotection. Brains intended for cell counting were coronally sectioned at 35 μm thickness, while brains intended for IHC were coronally sectioned at 25 μm thickness using a Leica CM1900 cryostat (Leica Microsystems, Germany) and placed in PBS or AntiFreeze solution for storage. Sections were mounted on coated microscope slides and cover-slipped with mounting medium containing the nuclear stain DAPI (Vectashield; H-1000, Vector, Burlingame, CA), or ToPRO3 (Slide Mount; #032347, Bioenno Lifesciences, Santa Ana, CA).

Immunohistochemistry

2.5

All IHC was performed on free-floating sections after washing and permeabilization in 0.01 M phosphate buffer solution containing 0.3% Triton X (PBS-T) for 10 min, 3 times (3 × 10 min). The sections were treated with PBS-T solution containing 0.3% hydrogen peroxide for 30 min, followed by washes in PBS-T (3 × 10 min). Sections were treated with PBS-T containing 5% normal donkey serum for 1 h to block non-specific binding of the secondary antibody. After 1 h, the sections were washed with PBS-T (3 × 10 min) and then incubated in PBS-T solution containing the primary antibody Goat anti-CRHR1 (1:3000; #SC-1757, Santa Cruz Biotechnology, Dallas, Texas) or Chicken anti-GFP (1:2000; #GFP-1020, Aves Lab, Davis, California) for 1 h on a shaker. The sections were then placed in a 4 °C refrigerator to further incubate for four days if immunolabeling for CRHR1, or overnight if immunolabeling for GFP. After incubation, the sections were washed in PBS-T (3 × 5 min) and incubated in a secondary antibody solution with Donkey anti-Goat 488 (1:400; A11055, Invitrogen, Waltham, Massachusetts) or Donkey anti-Chicken 488 (1:500; A78948, Invitrogen, Waltham, Massachusetts) for 90 min. The sections were washed one last time in PBS-T (3 × 10 min) and then mounted onto gel-coated slides.

When conducting IHC for CRH, the process began with treating the tissue with PBS-T containing 5% normal goat serum for 1 h to block non-specific binding. After an hour, the sections were washed with PBS-T (3 × 10 min) and then incubated in PBS-T solution containing the primary antibody Rabbit anti-CRH (1:20,000; Sawchenko, Salk Institute, La Jolla, California) for 1 h on a rotating table. The sections were placed in 4 °C to further incubate for a week. After incubation, the sections were washed 3 times in PBS-T for 5 min and underwent tyramide signal amplification. Sections were incubated in HRP-conjugated goat anti-rabbit IgG (1:500; G21234, Invitrogen, Waltham, Massachusetts) for 90 min, followed by washes in PBS-T (3 × 10 min). The sections were incubated in Fluorescein Tyramide amplification buffer (1:150; NEL741001KT, Akoya Biosciences, Marlborough, MA) on ice for 5 min, and then washed with PBS-T (3 × 10 min). The sections were then mounted onto gel coated slides or washed again to perform a second round of IHC.

Confocal imaging

2.6

Sections used for viral tracing were scanned using a Zeiss AxioScan.Z1 equipped with a Colibri camera and Zen AxioScan 2.3 software. A set of 18 brain regions was selected based on extensive literature indicating their robust expression of CRH. 35 μm sections containing these regions were imaged in a 70 μm interval, with a 20x magnification lens, using a confocal microscope (LSM 510, Zeiss, Germany), generating qualitative data. Sections used for determining proximal projections were first imaged at 60x magnification, at 0.5-μm intervals. 40x and 20x magnification images were then captured for regions of interest in 16 bits, 1024 μm × 1024 μm frame size. The images were processed using Fiji (ImageJ).

Analysis

2.7

Retrogradely labeled neurons were counted manually using Fiji. Only neurons with a visible nuclear stain were counted. An average of 108 sections were analyzed per mouse. Numbers in Table 1 represent the minimum and maximum number of retrogradely labeled neurons found in each area of interest across all mice, after injection of a CRE-dependent virus in the anterior or posterior PVT of CRH-CRE mice.

Results

3

PVT CRHR1 neurons are activated during early life and distinguish ELA from typical experiences

3.1

We first recapitulated the activation of PVT by early-life experiences, and the preferential activation of PVT cells expressing CRHR1 by ELA vs typical rearing (Kooiker et al., 2023). We used activity-dependent genetic labeling (TRAP2) in neonatal mice, administering tamoxifen on P6, the midrange of the ELA rearing period. This allowed cells that were activated in the subsequent 24-36 h to be labeled permanently. Using this approach, we found populations of activated and labeled “TRAPed” neurons throughout the brain in both CTL and ELA conditions, as in our prior work (Kooiker et al., 2023; qualitative data of follow-up experiment in Fig. 1A). We then conducted immunostaining for CRHR1 and recapitulated the discovery that a fraction of PVT TRAPed neurons express CRHR1 (Fig. 1B), and that a higher proportion of TRAPed PVT neurons expressed CRHR1 in ELA mice (40-60% of TRAPed neurons) compared to CTL mice (20% of TRAPed neurons) (Kooiker et al., 2023).Fig. 1PVT CRHR1 neurons are activated during early life and distinguish stress from typical experiences. (A) Coronal sections from male TRAP2 x NuTRAP mice raised in a control (left) or ELA (right) environment. Tamoxifen administration at P6 induced permanent expression of an EGFP tag in all activated (TRAPed) neurons. (B) TRAPed neurons expressing tdTomato (red) and CRHR1 (green) in the PVT of control (top) and ELA (bottom) TRAP2 x Ai14 male mice.Fig. 1

PVT CRHR1 neurons are abutted by CRH + axon terminals

3.2

The fact that a large fraction of TRAPed neurons expressed CRHR1 in ELA mice but not in control mice raised the possibility that the activation of these neurons was triggered by the ligand for the CRHR1 receptor, CRH. Therefore, we searched for local PVT CRH cells and axon terminals using both IHC and the CRH-Cre mouse line crossed with a tdTomato reporter line. CRH cells were present in PVT (Fig. 2A). Next, we used a transgenic mouse expressing FLPo under the control of the CRHR1 promoter and delivered a FLP-dependent virus that expresses EGFP in the PVT, to visualize CRHR1 expressing neurons. This approach revealed CRH-labeled axons that densely innervate the EGFP-expressing neurons in the PVT (Fig. 2B, qualitative data from 4 male mice). Indeed, dendrites of the CRHR1 neurons seem to form synaptic contact with CRH-labeled axons (Fig. 2C). However, the origin of the CRH-positive axons remained unclear.Fig. 2PVT CRHR1 neurons are abutted by CRH+ axon terminals. (A) CRH-expressing neurons in the PVT of a CRH-CRE x Ai14 mouse. tdTomato was visualized via direct fluorescence while CRH (green) is visualized with IHC. (B) EGFP-expressing cells in the PVT of a Crhr1-FlpO mouse, after delivery of a Flp-dependent AAV expressing EGFP, and CRH visualized with IHC (red). (C) Zooming in, CRH-labeled axons densely innervate the PVT and abut CRHR1-expressing neurons (n=4 males).Fig. 2

CRH axon terminals might arise from somata in PVT-adjacent nuclei, such as the paratenial thalamic nucleus

3.3

CRH-expressing neurons populate other thalamic nuclei including the paratenial nucleus, which borders the anterior portion of the PVT (aPVT). CRH x Ai14 sections underwent immunohistochemistry for CRHR1 and we searched for tdTomato-labeled neurons from the paratenial nucleus that might project and innervate the CRHR1 expressing neurons in PVT (Fig. 3A, n = 4 males). We observed paratenial-origin tdTomato-positive axons projecting to the PVT, potentially contacting CRHR1 labeled neurons.Fig. 3CRH+ axon terminals may originate from adjacent nuclei, such as the paratenial thalamic nucleus. (A) Confocal image of sections from CRH-CRE x Ai14 mice that underwent IHC for CRHR1. tdTomato-labeled neurons (red) are seen in the paratenial nucleus. CRHR1-expressing neurons are visible in the aPVT, especially in the lateral regions (green). Upon higher magnification, tdTomato-labeled neurons from the paratenial nucleus are projecting to the aPVT and their axons are in close proximity to CRHR1-expressing neurons in the PVT (n=4 males).Fig. 3

The PVT receives afferent CRH projections from pontine nuclei

3.4

The PVT receives inputs from many regions, ranging from the prelimbic area to the nucleus of the solitary tract (Li and Kirouac, 2012; Kirouac, 2025). To identify potential CRH-expressing projections from these regions, we administered CRE-dependent retrograde adeno-associated viruses (AAVs), and specifically AAV2, into the left anterior PVT or posterior PVT of CRH-CRE mice (Fig. 4A) and searched for retrogradely labeled neurons in 18 regions containing CRH-expressing neurons. We consistently found retrogradely labeled neurons in the pontine region, particularly in the parabrachial nucleus (PBN) and Barrington nucleus (BN) (Fig. 4B–E and Table 1) of both male and female mice (aPVT n = 2 males, 2 females; pPVT n = 4 males, 2 females). Retrogradely labeled neurons tended to localize in the lateral anterior portion of the PBN, (Fig. 5D). In contrast, retrogradely labeled neurons in BN clustered at the posterior end of the nucleus (Fig. 5E).Fig. 4PVT receives afferent CRH projections from pontine nuclei. (A) Schematic of administering a CRE-dependent retrograde AAV expressing EGFP, into either anterior (aPVT) or posterior (pPVT). (B–C) Confocal images of the injection site after administration of a CRE-dependent retrograde AAV expressing EGFP (aPVT n= 2 males, 2 females; pPVT n= 4 males, 2 females). (D–E) Retrogradely labeled neurons were found in the parabrachial nucleus and Barrington nucleus.Fig. 4. Fig. 5Dual anterograde/retrograde tracing of PBN CRH+ neurons demonstrate the existence of the projection. (A) Schematic of administering a CRE-dependent anterograde AAV1 (green) into the parabrachial nucleus in combination with a CRE-independent retrograde AAV2 (red) into the pPVT or aPVT of a CRH-CRE mouse. (B) Injection site in the parabrachial nucleus, showing successful labeling of CRH-CRE expressing neurons with EGFP. (C–D) In the same mice, the aPVT ((C), n=4 males) or pPVT ((D), n= 4 males, 2 females) was injected with a CRE-independent retrograde AAV. (E–F) EGFP-labeled axons from the parabrachial nucleus appeared in both the aPVT (E) and pPVT (F). (G) Overview image of the viral expression in the parabrachial nucleus. (H) Higher magnification of image in G showing dually labeled neurons on the lateral side of the parabrachial nucleus.Fig. 5

Dual anterograde/retrograde tracing demonstrates the existence of a CRH-expressing projection from the PBN

3.5

Retrograde tracing identifies potential candidate cells projecting to the candidate region. To validate the existence of a CRH-expressing projection from PBN to PVT we performed concurrent anterograde and retrograde tracing to find colocalization of both retrograde and anterograde viruses in the same PBN neurons (Fig. 5A). We injected a CRE-dependent anterograde AAV1 expressing EGFP into the left, lateral PBN and a CRE-independent retrograde AAV2 into either the anterior PVT (n = 4 males) or posterior PVT (n = 4 males, 2 females) of CRH-CRE mice (Fig. 5B–D). The anterograde injection identified EGFP-positive axons innervating both PVT regions (Fig. 5E and F), and the retrograde virus administered into either anterior or posterior PVT delineated PVT-projecting neurons in the PBN. A few neurons co-expressed both markers, an anatomical connection of PBN to both anterior and posterior PVT (Fig. 5G and H).

The Barrington nucleus is a source of CRH inputs to PVT

3.6

The Barrington nucleus (BN) is densely populated with CRH+ neurons (Fig. 6A; Vincent and Satoh, 1984; Verstegen et al., 2017) and is thus poised to integrate stress-related signals with micturition. Our initial brain-wide assessment of PVT afferents uncovered retrogradely labeled neurons in the BN (Fig. 6B). Therefore, we used the dual anterograde/retrograde approach to validate the BN to PVT CRH projection in the mouse (Fig. 6C). Indeed, after administration of the retrograde AAV into the posterior PVT, and anterograde AAV in the BN area, we found neurons expressing both viruses in the BN (n = 4 males, 2 females).Fig. 6The Barrington nucleus is a source of CRH inputs to the PVT. (A) The Barrington nucleus of a CRH-CRE mouse labeled with a CRE-dependent anterograde AAV expressing EGFP (green) (n= 4 males, 2 females). (B) Injection site in pPVT showing expression of tdTomato (red) after administering the CRE-independent retrograde AAV2 (n=4 males, 2 females). (C) Overview image of retrograde and anterograde viral expression in the Pontine region. (D) Neurons dually labeled by both the retrograde and anterograde AAV were found sporadically throughout the Barrington nucleus.Fig. 6

Discussion

4

This series of experiments collectively demonstrate how CRH, originating both within and outside of PVT can contribute to the activation of PVT CRHR1 neurons during ELA, conveying environmental signals to this key regulatory node. We first recapitulate the robust activation of CRHR1-expressing neurons during ELA. We then identify CRH expressing neurons in the PVT with the use of two independent methods (IHC and transgenic reporter mice). These local cells are candidates for local CRH release. Next, we discover new populations of CRH-expressing neurons in the adjacent paratenial nucleus, that seem to communicate with CRHR1 PVT cells. Finally, CRH neurons in the pontine region, specifically the BN, project to the PVT and could potentially relay maternal-derived sensory input from peripheral nerve endings to the PVT.

While the profound effect of both positive and adverse early-life experiences on life-long motivated behaviors is implied in the human and proven in experimental animal models, how these experiences reach brain regions regulating adult behaviors is unclear. A second question is how transient early life experience can lead to persistent changes in adult behavior. Seeking to identify brain regions encoding early-life experiences, and especially those discriminating ELA from typical rearing conditions, we employed targeted recombination of active populations (TRAP) and identified the PVT as a critical node both robustly activated and uniquely distinguishing ELA from typical rearing (Kooiker et al., 2023). In addition, pilot data (Kooiker et al., 2024) suggest that the same neurons in PVT that are activated early in life also contribute to the disruptions of adult reward behaviors. These facts provide impetus for the studies described here.

The PVT's contribution to gating reward and stress behavior as a function of prior stress has been initially investigated in adult rats by Seema Bhatnagar, then in Mary Dallman's lab (Bhatnagar and Dallman, 1998), who found that activation of neurons in the posterior PVT was regulated by a history of prior adult stress. Studies by our group indicate that prior stress could involve the PVT even when it is as remote as events during the first week of life.

Surprisingly, the population of PVT neurons specifically overrepresented among ELA-activated cells expressed the CRH receptor type 1, CRHR1. These findings provided the impetus for the current work, searching for the origin of CRH that might be released during ELA to activate its cognate receptor.

While PVT contains cell populations that express CRH (Itoga et al., 2019; Engelke et al., 2021; Kooiker et al., 2023), the relation of these cells to neurons expressing CRHR1 has been unclear. To address this question, we used immunohistochemistry to visualize CRH somata and axons in mice with genetically labeled CRHR1 neurons. We found an abundance of CRH-labeled axons that made it difficult to distinguish underlying CRH-labeled cell bodies as the origin of the plethora of axons. Nevertheless, we establish that CRH terminals are in relative proximity to CRHR1 neurons, and CRH can interact with its receptor via synaptic contact or volume transmission, allowing the CRH ligand to travel and bind to CRHR1 without synaptic contact (Chen et al., 2012).

Beyond the PVT itself, we investigated the paratenial nucleus, an understudied region often coupled with the PVT when determining its connectivity and function (Chen and Su, 1990; Van der Werf et al., 2002; Vertes and Hoover, 2008). Using mice in which CRH neurons are visible via a fluorescent marker (tdTomato) we found labeled axons projecting from the paratenial nucleus into PVT, potentially activating PVT CRHR1 receptors through synaptic or volume transmission. Linley et al. (2025) demonstrated that the paratenial nucleus receives inputs mostly from the medial prefrontal cortex, with few brainstem afferents. Thus, it is an unlikely source of CRH-mediated sensory information during ELA.

Our tracing experiments revealed that CRH-expressing neurons projecting to PVT reside in several pontine nuclei, particularly in the lateral PBN and Barrington nucleus. The presence of PVT projections from these regions is consistent with the literature (Cornwall and Phillipson, 1988; Krout and Loewy, 2000; Li and Kirouac, 2012). We did not find any CRH-expressing retrogradely labeled neurons in some brain areas with well-established CRH expression, such as the central amygdala or the paraventricular nucleus of the hypothalamus (Table 1). These two areas have been shown to project to the PVT, but whether the projecting neurons express CRH was not reported and has remained unclear (Singh et al., 2022; Wang et al., 2025). Our inability to detect projections from these areas could also be a result of relatively small PVT area in which neurons are infected by the viral retrograde tracers, as well as the heterogeneity of PVT's input/output connections along its anteroposterior axis (Kirouac et al., 2022).

Among the brain regions containing CRH-expressing neurons that project to the PVT, the PBN and the Barrington nucleus were the major sources. The PBN is generally considered to relay negative affective stimuli (Zhu et al., 2022). For example, lateral PBN is considered a hub for pain-related exteroceptive stimuli which are relayed throughout the brain (Sato et al., 2015; Campos et al., 2018; Chiang et al., 2019). PBN also conveys interceptive information to thalamus, promoting homeostasis (Mizusawa et al., 1995; Palmiter, 2018; Engelen et al., 2023; Zhang; et al., 2024). While the role of PBN in conveying information about ELA to the PVT is unclear, it is tempting to propose that it may contribute to both physical and emotional components of the LBN paradigm (Bolton et al., 2019). Future studies should determine whether CRH-expressing neurons in PBN are themselves active during early life and whether they relay input to the PVT.

Finally, we established that the Barrington nucleus projects to PVT using dual anterograde and retrograde tracing. This nucleus is densely populated with CRH-expressing neurons, and evidence of its projection to PVT was first reported in rats in separate classical retrograde and anterograde tracing experiments (Peng et al., 2017; Otake and Nakamura, 1995; Valentino et al., 1995). Barrington nucleus regulates micturition, and CRH-expressing neurons contribute to recruiting pelvic muscles to contract the urinary wall and expel urine (Barrington, 1925; Hou et al., 2016; Ito et al., 2020). The discovery of CRH-expressing projection from this nucleus to PVT suggests additional roles: ‘Gig’ Levine and his group suggested that sensory input from the mother during the first week of life in the form of anogenital licking and grooming was crucial for micturition and also provided a stress-blunting effect during the so-called stress hyporesponsive period (Suchecki et al., 1993; Korosi and Baram, 2009). While the presence of stress reactivity throughout the first week of life is now established (Yi& Baram, 1994; Chen and Baram, 2016), the current work lends substance to the idea that, during the first weeks of life, CRH-expressing projections from Barrington nucleus to PVT convey important sensory information to a key regulating brain node (Kooiker et al., 2024). In support of this notion, Li et al. (2025) recently demonstrated that CRH-expressing neurons in this nucleus can be activated by tactile and noxious stimuli. Future work should directly determine the role of the CRH-expressing Barrington-PVT projection in conveying maternal input to her pups.

CRediT authorship contribution statement

Jorge M. Mendoza: Formal analysis, Funding acquisition, Investigation, Visualization, Writing – original draft. Amalia Floriou-Servou: Data curation, Supervision, Writing – review & editing. Yuncai Chen: Data curation, Investigation, Methodology, Validation. Cassandra Kooiker: Data curation, Investigation, Methodology. Mason Hardy: Methodology, Resources, Visualization. Tallie Z. Baram: Conceptualization, Data curation, Funding acquisition, Project administration, Supervision, Validation, Writing – review & editing.

Declaration of competing interest

The authors do not have any conflict of interest to report.