Acute carbamoylated erythropoietin reduces social stress-induced anxiety and depression-related behaviors

Abstract

The hormone and trophic factor Erythropoietin (EPo) promotes red blood cell production and has neurotrophic effects, modulating behavior and promoting neural plasticity, like neurogenesis. Modifying EPo by attaching a carbamoyl group (cEPo) results in similar neuronal effects without erythropoietic actions. We hypothesize that neuroplastic and learning effects of cEPo may be dependent on its action in dorsal dentate gyrus of the hippocampus, where neurogenesis occurs. The Stress Alternatives Model (SAM), a 4-day social stress and decision-making paradigm with a large novel aggressor, that provides opportunities to avoid interaction via escape routes. Early, male test mice display stable Escape (avoiding aggression) or Stay (acquiescence to aggressor) behavioral phenotypes. In these studies, mice were given a single intracerebroventricular (icv; 100 ng), or single intra-dentate gyrus (iDG; 10 ng) injection of cEPo or vehicle. By Day 4, 30% of icv cEPo treated mice and 37.5% of iDG cEPo treated mice reversed their phenotype (Stay to Escape). Mice receiving vehicle injections did not change. Normalization of social preference, and reduction in fear freezing behavior, after cEPo treatment, coincide with transcriptional changes of orexin receptors (Hcrtr1 & Hcrtr2) in the hippocampus, including phenotype- and treatment-dependent alterations in learning-associated molecular signaling molecules. Hippocampal Hcrtr1 moderately colocalize with EPo receptors (Epor) in the dentate gyrus, while Hcrtr2 is expressed with Epor in CA₁. Anxiolytic actions of cEPo during social interaction indicate mechanisms that influence learning in stressful situations, suggesting that cEPo may be a novel agent for treatment of comorbid conditions related to anxiety, depression, and PTSD.

Graphical Abstract

1. Introduction

Major depressive and anxiety disorders are the most common psychiatric diseases (Kraus et al., 2019), and are underdiagnosed and undertreated, in part because currently available treatments are incompletely effective, have unconsidered side effects (Cascade et al., 2009; Zimmerman et al., 2010), and may be addictive, produce prolonged withdrawal, or often are compulsively abused (Davies and Read, 2019; Jauhar and Hayes, 2019). Depression, anxiety and Post-Traumatic Stress Disorder (PTSD) are highly comorbid (Spinhoven et al., 2011; Zlotnick et al., 1999), and affect a high proportion of modern populations, but are also underserved by existing pharmacotherapies. The result is that the quality of and satisfaction with these treatments is poor to moderate (Young et al., 2001). One third of patients are treatment-resistant (fail to respond to two or more antidepressants), and moreover, 23% of suicide victims are on an antidepressant at the time of death (Al-Harbi, 2012; David A. Mrazek et al., 2014; Gerhard et al., 2016). While ketamine has recently gained attention for its rapid antidepressive efficacy, its therapeutic effects are short lasting (2–3 weeks), and has potential for abuse due to its reinforcing qualities (Liu et al., 2016). Given the limitations of current antidepressant medications, it is imperative that new drugs with greater efficacy and fewer side effects are tested and made available.

Erythropoietin (EPo) is a multifunctional protein found peripherally and in the central nervous system. While the tissue protective and blood cell proliferating functions are well known and described, its neurotrophic effects are less understood (Kirkeby et al., 2008; Montero et al., 2007). The cell-proliferative and angiogenic capacity of this trophic factor are evident as neurogenesis and gliogenesis in hippocampal dentate gyrus (DG) and subventricular zone (SVZ), and in vascularization throughout (Kaneko et al., 2013; Osredkar et al., 2010; Yang et al., 2011). In addition, EPo has been demonstrated to promote cognitive function (Almaguer-Melian et al., 2024) and to limit depression (Duman and Newton, 2013; Li et al., 2017; Ma et al., 2016; Miskowiak et al., 2008; Miskowiak et al., 2009; Miskowiak et al., 2012). However, as a treatment for affective disorders, EPo has the disadvantage of potentially lethal side-effects through overproduction of red blood cells. A carbamoylated (the addition of a carbamoyl groups) version of EPo (cEPo) lacks the erythropoietic function (Leist et al., 2004), but remains angiogenic, and has shown demonstrable neurotrophic activity (Oh et al., 2012; Rothschadl et al., 2023; Thomas Tayra et al., 2013; Tiwari et al., 2021; Tiwari et al., 2019). Although several clinical studies have demonstrated the therapeutic effects of EPo in neuropsychiatric disease populations, pharmacokinetic studies with EPo or cEPo have not, as yet, established an optimal dose-response relationship in modulating behavior. Furthermore, the precise brain levels required for behavioral and neuroprotective effects are unknown. Therefore, demonstrating behavioral effects after direct cEPo administration into the brain and specific hippocampal circuits is an essential step in understanding the CNS effects of cEPo and developing well-informed translational studies. Progress must additionally be qualified by the information that effective antidepressant treatments also often exert anti-inflammatory actions along with enhanced neurotropic signaling and synaptic plasticity (Krupp et al., 2024; Troubat et al., 2021).

Although the anti-inflammatory effects of cEPo are significantly less well studied than those of EPo (Bond and Rex, 2014), and the precise mechanism and signaling pathway through which cEPo offers protective effects against inflammation in the CNS have not been elucidated, research from stroke and pain models suggests that a reduction in microglial activation plays a role (Rahmani et al., 2022; Villa et al., 2007). Behavioral stress models have been shown to produce enhanced Akt₂ and Mtor gene expression in BLA and hippocampal CA₃ (John et al., 2025), the proteins from which regulate glucose metabolism in glia, macrophage activity, cytokine production, and inflammation (Levenga et al., 2021; Zhang et al., 2021), all of which can be reversed by drugs or activities that reverse behavioral stress (John et al., 2025; Krupp et al., 2024). Tissue-protective and behavioral outcomes of cEPo in hippocampus likely encompass neuroprotective, neurotrophic, and anti-inflammatory effects, which leads to the suggestion that neurotrophins underlie the causal mechanisms related to the development of depression and other affective disorders (Levy et al., 2018). Thus, the neurotrophic and angiogenic, but not erythropoietic, cEPo presents a good candidate for therapeutic effectiveness for affective disorders (Duman and Newton, 2013; Miskowiak et al., 2012; Sathyanesan et al., 2018).

Recent analyses of classic tests for anxiety and depression using current animal model paradigms suggest that the results have little translatable relevance for trials in clinical populations, with the caveat that tests including social measures have promise (Blanchard et al., 2013; Curtiss et al., 2017; Haller and Alicki, 2012; Haller et al., 2013; Holsboer and Ising, 2010; Keifer and Summers, 2016; Nestler and Hyman, 2010). The limitations of current therapies and animal model systems suggest that more research is necessary to provide novel approaches, new models, and therapies are necessary to specifically address the complexity of social interactions in tests of anxiety, depression, and PTSD (Blanchard and Blanchard, 1989a, b; Keifer and Summers, 2016; Robertson et al., 2015; Yaeger et al., 2022c).

Animal responses in the Stress Alternatives Model (SAM) provide a window onto development of behavior associated with anxious, depressive and PTSD attributes (Blanchard et al., 2013; Robertson et al., 2015; van der Kolk, 2006; Yaeger et al., 2022b; Yaeger et al., 2022c; Yehuda and LeDoux, 2007), and the mechanisms of decision-making that produce resilient and susceptible phenotypes (Smith et al., 2014; Yaeger et al., 2020; Yaeger et al., 2022b). We note that social defeat models rarely involve females, in part because the behavioral reactions are difficult to attain, and often broadly inefficient (Harris et al., 2018; Huhman et al., 2003; Iñiguez et al., 2018; Newman et al., 2019; Solomon, 2017; Summers et al., 2021; Takahashi et al., 2017; Yaeger et al., 2022a). Since males and females are diagnosed at distinctively different levels for affective disorders, and commit suicide in very different rates (Kessler, 2003), to allow future studies to focus on both sexes the SAM was designed to include female subjects (Summers et al., 2021; Yaeger et al., 2022a). In an oval SAM arena with apical escape routes, novel larger aggressive individuals interact with smaller adult test subjects (Fig. 1). Test animals self-select one of two phenotypes: Escape or Stay, which exhibit stress-resilient (social engagement) and susceptible (social avoidance) responses to social interaction/preference tests (SIP) and differences in reactive plasma glucocorticoid concentrations, also distinguished by sex (Smith et al., 2016; Staton et al., 2018; Yaeger et al., 2024b; Yaeger et al., 2022b). Production of two easily identified phenotypes via a simple dichotomous choice, as occurs in the SAM, results in distinctively different behaviors, gene expression, and neurochemical responses, which develop over a short time course (Carpenter et al., 2023; Robertson et al., 2015; Smith et al., 2014; Yaeger et al., 2022b), and are dependent on specific types of learning (Carpenter and Summers, 2009; Smith et al., 2016; Summers et al., 2017; Yaeger et al., 2020; Yaeger et al., 2022c). Resilient status is confirmed for Escape animals, because anxiolytic drugs (corticotropin releasing factor type 1 receptor [CRF₁] antagonist antalarmin, orexin 1 receptor [Orx₁R] antagonist SB-674042, and orexin 2 receptor [Orx₂R] agonists [Ala¹¹, d-Leu¹⁵]–Orx_B and YNT-185) promote Escape behavior in Stay animals (Carpenter et al., 2023; Smith et al., 2016; Staton et al., 2018; Yaeger et al., 2022b; Yaeger et al., 2022c). Alternatively, anxiogenic drugs (Yohimbine, an α₂ adrenoreceptor antagonist, and Orx₂R antagonist MK-1064) promote Stay behavior in Escape phenotype mice, clearly suggesting that Stay responses represent stress-susceptible behavior. Behaviors reflecting motivation to escape the SAM are also modified by stress-related neuromodulatory events (Staton et al., 2018; Yaeger et al., 2024a; Yaeger et al., 2022b). A fear conditioning protocol, during which a tone (conditioned stimulus, CS) precedes aggressive interaction (unconditioned stimulus, US), also produces associative learning (conditioned response, CR) in the SAM. Although a trace period is used between the condition stimulus (cue = tone) and US (aggression), this association yields Pavlovian conditioning, which occurs concomitantly with contextual conditioning (Yaeger et al., 2022b). Contextual and cued responses (CR = freezing) to environment or tone alone in conditioned Stay mice is reduced by intra-basolateral amygdala (intra-BLA) injection of an Orx₁R antagonist (Yaeger et al., 2022b). This model was specifically designed to overcome limitations with the most commonly used tests for anxiety and depression relative to clinical translation (Curtiss et al., 2017; Haller and Alicki, 2012; Haller et al., 2013).

Fig. 1. The Stress Alternatives Model (SAM) results in phenotype establishment after two days of social stress.

A) The SAM is a 4-day paradigm in which test mice are conditioned to a tone (left) before an opaque cylinder is lifted and animals must decide whether to Escape from (top right) or Stay with (bottom right) a large social aggressor. By the end of Day 2, test mice commit to Escape or Stay behavioral phenotypes. B) Experimental design for behavioral trials includes stereotaxic surgeries for cannula implantation followed by a recovery and handling period before the beginning of the SAM. On Day 3, 1 hr (icv) or C) 24 hrs (iDG; see map for injection sites, black for vehicle, red for iDG cEPo) before SAM exposure, mice were administered cEPo into the lateral ventricle or directly into the dentate gyrus (C). Following SAM social interaction on Day 4, mice were exposed to Social Interaction/Preference and Fear Conditioning tests on Day 5.

Our purpose was to make use of the SAM to examine the putative anxiolytic and anti-depressive properties of acute cEPo treatment in this social defeat/avoidance model. Previous work on EPo, cEPo, and other neurotrophins are typically examined over weeks of administration. We hypothesized that a single dose of cEPo delivered intracerebroventricularly (icv, 100ng) would produce anxiolytic and antidepressive responses; and that they would be similar to a single dose of cEPo delivered intracranially, directly into the DG (intra-DG or iDG). We hypothesized additionally that acute icv or iDG cEPo would increase the proportion of stress-resilient phenotype Escape individuals, increasing motivation for, and actual Escape, as well as increasing the celerity of Escape. We therefore, also expected cEPo treated Stay animals to exhibit increased social preference, and decreased cued fear conditioning (freezing) prior to each SAM trial, and also during testing for conditioned responses (CR) after all SAM trials are finished.

2. Methods

2.1. Subjects and housing

Adult male C57BL/6NHsd mice (6–8 weeks old) weighing ~22–28 g were obtained from Envigo (Indianapolis, IN; n=194) and acclimated for a seven day period in groups five, after which animals were singly housed in rooms held at 22°C and 35% relative humidity for the remainder of the experiments. Food and water were provided ad libitum. Pharmacological manipulations (n = 95) included bilateral stereotaxic surgeries, performed where guide cannula (26 ga cut to 1.0 mm [icv] and 1.65 [dg]) were directed at the lateral ventricle (icv) or hippocampal dentate gyrus (intraDG). A separate set of retired male breeder Hsd:ICR mice (CD1, n = 18) weighing ~50 g (Envigo) were individually housed, and used to initiate aggression in the Stress Alternatives Model (SAM; Fig. 1A).

Mice were subjected to a 12:12 light-dark cycle (scotophase, lights off at 6 pm), and behavioral experiments were performed during scotophase (the animals’ active phase is during darkness). Two days (48 h) after surgeries, test subjects (C57BL/6NHsd mice) were handled daily for 7 days before SAM exposure and behavioral testing the five proceeding days (Fig. 1B). All procedures (surgery and behavioral testing) were performed in a manner that minimized suffering. The number of animals used was in accordance with the National Institutes of Health’s Guide for the Care and Use of Laboratory Animals (NIH Publications No. 80–23) and approved by the Institutional Animal Care and Use Committee of the University of South Dakota.

2.2. Stereotaxic surgeries

Mice were anesthetized using isoflurane (2% at 1.0 L/min flow rate) before bilateral icv or intraDG guide (iDG) cannula (PlasticsOne, Roanoke, VA; 26 ga cut to 1.85 [icv] or to 2.8 [iDG] mm) implantation. To help control for the potential of neuroinflammation following surgical implantation of cannulae into the brain (Seyer et al., 2016), mice were provided a recovery period (7 days) before behavioral testing. The possibility of cannula implantation into the lateral ventricle has been examined relative to TNF_α or IL₆ cytokine production and cognitive function, both of which were unaffected (Seyer et al., 2016). Importantly, we have not, as yet measured neuroinflammation related to intra-DG cannula placements, although the distance of cannula passage is less for this surgery. Cannula placement was performed using the following stereotaxic coordinates: −0.50 mm AP, −1.0 ML, and −1.85 mm DV for intracerebroventricular (icv) implantation and −1.40 mm AP, ±0.60 ML, and 2.0 mm DV for intra-DG (iDG) implantation. During surgery and for ~45 minutes post-surgery, mice were kept on a warming pad to maintain core body temperature. Immediately following surgery and 24 hours after surgical procedures, mice were provided pain relief in the form of subcutaneous injections of the analgesic ketorolac (5 mg/kg).

2.3. cEPo Administration

The cEPo molecule was prepared as previously described (Sathyanesan et al., 2018). Treatment with cEPo was given either icv (100 ng) or directly into the dentate gyrus (iDG; 10 ng). The effects of icv or iDG cEPo injections were compared to vehicle-treated (n = 17 for icv-cEPo Vehicle; n=15 for iDG-cEPo Vehicle) control animals that underwent cannula implantation surgeries, were administered artificial cerebrospinal fluid (aCSF), and were exposed to the same testing conditions and procedures as cEPo-treated mice. On Day 3 of the behavioral design (Fig. 1B), mice were infused into the lateral ventricles or bilaterally into the dentate gyrus (300 nL/side) with their designated treatment, diluted in aCSF, an hour (icv) or 24 hours (iDG) before social interaction in the SAM.

Artificial cerebrospinal fluid (aCSF; 8.59 g NaCl, 0.201 g KCl, 0.279 g, CaCl₂, 0.16 MgCl₂, 0.124 g NaH₂PO₄, 0.199 g Na₂HPO₄/L H₂O) was mixed and brought to a physiological pH (~7.33) using NaOH before being filtered, degassed, and stored at 4°C. Drugs were infused using injector cannulae (33 ga cut to 1.85 mm [icv] and 2.0 [iDG], extending 0.85 mm below the guide cannula) placed into implanted guide cannulae, and injecting with a 1.0 μL digital syringe (Model 7101 Zero Dead Volume, Knurled Hub 2.75”, 22GA Needle; Hamilton Company, Reno, NV) at a rate of 0.5 μL/min. After drug administration, the injector and syringe were left in place for 90 sec.

2.4. Social stress and decision-making paradigm

In the SAM paradigm (Fig. 1A), social conflict between a larger novel CD1 mouse and smaller C57BL/6NHsd male mouse takes place for five minutes each day, over four days, during which test animals may shorten interaction with the aggressor by escaping through size-limited tunnels at the ends of an oval open field arena. Prior to the social interaction, a tone given as a conditioned stimulus (CS) during isolation in the SAM apparatus allows for Pavlovian Fear Conditioning of test subjects to the upcoming social interaction (unconditioned stimulus, US). As distinct and stable phenotypes (active avoidance, Escape, and accepting confrontation, Stay) determined over 45 experiments with 98% reliability (Robertson et al., 2015; Yaeger et al., 2022b) are established on Day 1 or 2, drug manipulation before Day 3 allows for within-sample and between group behavioral comparisons for phenotypes and drug controls (vehicle) during the SAM (Days 3 & 4) (Robertson et al., 2015; Smith et al., 2014; Smith et al., 2016; Staton et al., 2018; Yaeger et al., 2022b). In this model unique behavioral patterns and phenotypes develop over time. The design allows for within sample before and after treatment comparisons, which allow for identification of pre- and post-traumatic neuromodulatory plasticity. These qualities of the SAM present development of specific phenotypes, one of which (Stay) is associated with stress-vulnerability and failure to recover from trauma, as with PTSD (Yehuda and LeDoux, 2007). The Escape phenotype represents stress resilience. All procedures were performed in accordance with the Guide for the Care and Use of Laboratory Animals (NIH Publications No. 80–23) and approved by the USD Institutional Animal Care and Use Committee.

2.5. Experimental Design

For these experiments, cEPo dosages were administered icv 1 hour prior, or directed at the DG 24 hours prior (based on previous and preliminary experimentation; (Yaeger et al., 2020; Yaeger et al., 2022b) to SAM interaction on Day 3 (Fig. 1B). Behavioral measurements were taken during the active phase (scotophase/dark cycle) and includes two measures of freezing (1. socially induced - associated with aggressive conflict during SAM interactions; and 2. in response to fear conditioning [condition stimulus, CS = tone + US = aggression]), attention to the escape route, and social preference as measured in the Social Interaction/Preference (SIP) test. Brains were collected and used for visual representations of mRNA (using RNAScope) or relative changes in gene expression of Arc, Bdnf, Creb1, Egr1, Hcrtr1 and Hcrtr2.

2.6. In situ hybridization – RNAScope

Sections of fresh frozen brains (coronal; 20 μm; relative to bregma AP −1.40 to −2.0) were placed in cold (4°C) 10% formalin for 20 min and subsequently washed (2x for 1 min) in 1x phosphate buffer solution (PBS), before dehydration with ethanol (50% x 1, 70% x 1, and 100% x 3; 5 min each with the final ethanol being kept at −20°C overnight). The following day, proteins were digested using a protease treatment and rinsed with dH2O. Brain sections were incubated for two hours with RNAscope (Advanced Cell Diagnostics, Newark, CA) probes (Hcrtr2, Cat. No. 460881; Hcrtr1, Cat. No. 46663; Epor, Cat No. 412351) in a hybridization oven (ACD HybEZ II oven) set to 40°C. Fluorophores were linked to probes and signaling was enhanced through application of a series of amplification buffers (RNAscope Fluorescent Multiplex Detection Reagents). Finally, tissue was briefly stained with DAPI (20 sec) and cover-slipped. Image acquisition was performed with a fluorescence microscope (Nikon A1R; 10x/0.30 Plan Fluor and 20x/0.75 Plan Apo VC Nikon objectives) and NIS Elements software. The hippocampus was identified from images and analyzed using QuPath 4.0 and ImageJ programs.

2.7. Quantitative reverse transcription PCR (RT-qPCR)

Brains of experimental mice were carefully dissected, and the hemispheres were separated. Tissues were gently rinsed with RNAlater stabilization solution, and the hippocampal regions were carefully dissected under a microscope. Total RNA was extracted using an RNAqueous micro kit (Invitrogen). The concentration and purity of RNA at 260/280 nm were determined using a NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific, USA), then 200ng RNA was reverse transcribed into cDNA (Applied Biosystems High-Capacity cDNA Reverse Transcription Kit, USA) using thermal cyclers (Techne Prime). Gene expression analyses were performed by quantitative real-time PCR (applied biosystems QuantStudio 5 using 500 nM of each forward and reverse primers and SYBR green Universal PCR master mix (Gendepot, USA and Applied Biosystems, USA). Primers for Arc, Bdnf, Creb1, Egr1, Hcrtr1 (4351370, Mm01185776_m1), Hcrtr2 (4351370, Mm01179312_m1), and Gapdh (4453320, Mm99999915_g1) as the housekeeping gene were designed to amplify gene targets using the Primer3 program (https://bioinfo.ut.ee/primer3-0A0/). The expression levels of each gene target were normalized with the housekeeping gene panel, and the fold change of transcription was quantified using the relative quantification 2^-ΔΔCt method (Livak and Schmittgen, 2001).

2.8. Statistical Analyses

Experimental designs and statistical analyses and were based on a priori hypotheses, for the purpose of avoiding combinatorial exponential expansion of error from multiple tests (Veazie, 2006). This statistical pre-planning allows for a wider range of multiple comparison analyses across hypothetical designs. Analysis made use of two-way ANOVA for cEPo x Phenotype (Stay x Escape) designs, one-way ANOVA drug-dependent SAM-dependent behavioral responses (Days 1–4). Comparisons between two treatments (Vehicle, icv cEPo, iDG cEPo) within a given phenotype (Escape or Stay) or gene expression were investigated by Student’s t-tests. Normality and variance were tested using the Shapiro-Wilk and Brown-Forsythe tests respectively. The results are reported without α adjustment (Feise, 2002; Jennions and Moller, 2003; Moran, 2003; Nakagawa, 2004; Perneger, 1998; Rothman, 1990) based on a priori hypothesis driven exclusion from combinatorial effects (Veazie, 2006). Significant effects between groups for one-way analyses were examined with Student–Newman–Keuls post hoc analyses (to minimize Type I error) and Duncan’s Multiple Range Test (to minimize Type II error).

3. Results

Typical for the SAM paradigm, untreated or vehicle treated groups produce two stable behavioral phenotypes characterized by their response to larger aggressive conspecifics (CD1 mice), in which they either Escape from or Stay in the oval social interaction arena (Robertson et al., 2015). In previous experiments including over 1,500 mice, the ratio of Escape to Stay mice is typically 50:50, which was true for these experiments. This is a self-selection process, and although animal cohorts may vary, overall Stay and Escape phenotypes, which are mostly determined during Day 1 interactions (John et al., 2025) , are typically balanced.

3.1. icv or iDG cEPo treatment promotes Escape behavior

On Day 2 of SAM social interaction, when behavioral phenotypes are stable, and prior to injection on Day 3, no mice of the Stay phenotype expressed Escape behavior, as expected. However, after icv and intra-DG cEPo injection, some of these same mice began to express Escape behavior on Day 3 (8% icv; 20% intra-DG) and Day 4 (day after injection 30% icv; 38% intra-DG; Fig. 2A, B), but not in vehicle treated mice.

Fig. 2. cEPo produces Escape in Stay mice and reduces the latency to Escape.

A) Administration of icv-cEPo (100ng) increases escapes for Stay animals (n = 11) on Days 3 and 4. B) Administration of iDG-cEPo (10ng) increases escapes for Stay animals (n =11) on Days 3 and 4. C) In the Escape phenotype only, mean (± SEM) latency (s) to escape decreases from Day 1 to 2 for both vehicle-treated animals (n = 12, p ≤ 0.009) and icv-cEPo-treated animals (n =12, F_3,24 = 8.33, p ≤ 0.001). Additionally, icv-cEPo-treated animals decreased latency to Escape on Days 2 to 4 (n = 12, p ≤ 0.007) and Days 3 to 4 (n = 12, p ≤ 0.008). D) In Escape mice, significantly decreased latency (mean ± SEM) to escape (p ≤ 0.003) occurs for both vehicle and cEPo-treated animals over four days. Vehicle-treated animals escaped faster on Days 1 to 3 (n = 6, p ≤ 0.009) and Day 1 compared to Day 4 (n = 6, p ≤ 0.004). Animals treated with iDG-cEPo escaped quicker on Days 1 to 2 (n = 8, p < 0.001). The assessments of iDG-cEPo escape celerity on Days 1–3 (n = 8, p < 0.001), Days 1 to 4 (n = 8, p < 0.001), Days 2 to 3 (n = 8, p ≤ 0.043), and Days 2 to 4 (n = 8, p ≤ 0.026) are within sample comparisons, taken before treatment (Days 1 and/or 2) with those taken after iDG-cEPo treatment (Days 3 and/or 4).

While latency to escape (time spent with the CD1 aggressor), was significantly longer on Day 1; typical for the SAM paradigm, and Day 2 escape time is significantly faster (latency is reduced; F_3,64 = 22.5, p < 0.001), regardless of treatment, icv cEPo treatment significantly improves the homogeneity of response for latency to active avoidance (F_3,24 = 8.33, p < 0.001; Fig. 2). Celerity to escape was improved by Day 4, in within-sample and post-treatment analyses, but only in icv-cEPo treated animals, for which a significant reduction in time to escape was measured compared to Day 1 (p < 0.001; within-sample, pre- to post-drug comparison) Day 2 (p ≤ 0.027; within-sample, pre- to post-drug comparison), and Day 3 (p ≤ 0.028; post-drug treatment comparison, Fig. 2C). Similarly, in intra-DG cEPo treatment, a significant reduction in escape latency exists when comparing Days 2 (p < 0.001; within-sample, pre- to post-drug comparison) and 3 to Day 4 (p < 0.001; post-drug treatment comparison, Fig. 2D). Additionally, both icv- and iDG-cEPo treatment appear to have reduced the variability of escape responses, leading to more precise statistical outcomes. In previous experiments, additional improvement in active avoidance time on Day 4 is typically only observed with potent pharmacological or physical (exercise) anxiolytic treatments (Smith et al., 2016). This is also the time point at which the greatest percentage of Stay animals revert to Escape phenotype behavior (Fig. 2A, B).

3.2. Motivation to Escape

Time spent attending the escape route is enhanced by icv and iDG cEPo treatment; and is typically greater in Escape mice (Staton et al., 2018; Yaeger et al., 2022b). Thus, this single cEPo treatment increased attention to the escape route, but only in Stay mice converting to Escape phenotypic behavior (icv: t₁₁ = 2.5, p ≤ 0.038; iDG: t₁₄ = 4.8, p ≤ 0.00031; Fig. 3A, B, see red circle), suggesting that the anxiolytic effect of cEPo had a dramatic effect on motivation. As such, it is a quantitative measure of motivation to escape, which significantly grows for Escape mice over time, in vehicle treated mice through Day 4 (icv experiment: n=12, p < 0.001; iDG experiment: n = 10, p ≤ 0.033; Fig. 3A, B). This increase in motivation to escape does not occur for mice of the Stay phenotype. For cEPo treated Escape mice, increased motivation to escape is statistically significant through the time of treatment, on Day 3, for both icv and iDG experiments, but only in Stay-to-Escape transitioning mice thereafter. Interestingly, the motivation to escape on Day 4 appears to be reduced in both Stay and Escape icv cEPo injected mice (Fig. 3A), except when considering Stay mice that convert to Escape behavior. Additionally, following intra-DG cEPo injection, while the percent time attentive to the hole remains unchanged for Escape mice on Days 3 and 4, there appears to be a decrease in motivation to escape for Stay mice, which is not statistically significant. However, this decrement among Stay mice, appears to be solely due to the fact that iDG-cEPo treatment produced and anxiolytic effect in some animals, and for those mice it also stimulated reversion to Escape behavior, and thereby stimulated time spent attending to the escape route, which was significantly greater for both icv (t₁₁ = 2.5, p < 0.038, see red circle) and intra-DG injections (t₁₄ = 4.8, p < 0.00031; Fig. 3A, B – see red circle).

Fig. 3. cEPo increases motivation to Escape for escaping Stay mice on Day 4 of the SAM.

A) Vehicle (n = 21, F_1,57 = 9.3, p < 0.001; Phenotype Effect: F_1,57 = 24.7, p < 0.001), icv-cEPo treated (n = 22, F_3,51 = 2.7, p ≤ 0.05; Phenotype Effect: F_1,51 =40.3, p < 0.001), Escape, (n = 24, F_3,63 = 9.6, *p < 0.001) and Stay (n = 19, F_3,39 = 4.5, *p ≤ 0.008) groups had increasing motivation to escape over time, but were differentially attentive to the escape routes. Vehicle-treated Escape animals were more attentive than vehicle Stay mice on all days (n = 21, Day 1: t₁₉ = 2.3, ^#p ≤ 0.030; Day 2: t₁₉ = 2.7, ^#p ≤ 0.013; Day 3: t₁₉ = 3.7, ^#p < 0.001; Day 4: t₁₉ = 3.8, ^#p < 0.001). Vehicle Escape were more motivated by escape holes on Days 1 to 3 (t₁₁ = 3.7, *p ≤ 0.003), Days 1 to 4 (t₁₁ = 4.9,*p < 0.001) and Days 2 to 4 (t₁₁ = 3.1 *p ≤ 0.011). Injection of cEPo (icv or iDG) reduces anxiety in some Stay mice (), which allowed escape, but not all cEPo Stay mice escaped (). Stay icv cEPo-injected mice that switched to Escape on Day 4 exhibited a significant increase in escape motivation (), paying more attention to the Escape Route (inverted open triangle + red circle: t₁₁ = 2.5, ⁺p ≤ 0.038) than icv cEPo-treated Stay mice that did not escape (open triangle, dashed brick-red circle ). Escape icv cEPo-treated animals were more attentive to the escape routes on all days compared to Stay counterparts (n = 22, Day 1: t₂₀ =3.0, ^#p ≤ 0.007; Day 2: t₂₀ = 2.9, ^#p ≤ 0.009; Day 3: t₂₀ = 3.2, ^#p ≤ 0.005; Day 4: t₁₇ = 3.5, ^#p ≤ 0.003), with the exception of Stay mice switching to Escape. B) Escape and Stay, vehicle-treated (Phenotype Effect: n = 16, F_1,36 =9.5, p ≤ 0.009) and iDG-cEPo-treated mice (n = 23, F_1,39 = 17.2, p < 0.001) were differently attentive to the escape routes, with motivation to escape growing over time (Vehicle: F_3,36 = 4.1, p ≤ 0.025; Escape: n = 13, F_3,27 = 2.8, p ≤ 0.05). A treatment effect of iDG-injected cEPo was observed in the Stay mice (n = 26, F_{2, 45} = 4.7, p ≤ 0.027), such that Stay iDG cEPo-injected mice that switched to Escape on Day 4 () exhibited a significant increase in attention to the Escape Route (inverted open triangle/red circle: t₁₄ = 4.8, ⁺p ≤ 0.00031) than icv cEPo-treated Stay mice that did not escape (). Vehicle-treated Escape mice were more attentive to escape routes on Days 1–3 compared to Stay counterparts (Day 1: t₁₄ = 2.9, ^#p ≤ 0.012; Day 2: t₁₄ = 2.9, ^#p ≤ 0.012; Day 3: t₁₄ = 2.8, ^#p ≤ 0.015). Escape cEPo-treated were more attentive than Stay cEPo-treated counterparts on all days (Day 1: t₂₁ = 3.9, ^#p < 0.001; Day 2: t₁₉ = 3.1, ^#p ≤ 0.006; Day 3: t₂₁ = 5.6, ^#p < 0.001; Day 4: t₁₁ = 3.9, p ≤ 0.003), with the exception of Stay mice switching to Escape.

3.3. Social preference and avoidance

After completing the SAM protocol, during the Social Interaction/Preference Test (SIP) on Day 5, Escape phenotype mice spend significantly more time attending to the social target than Stay mice (Phenotype effect: F_1,35 = 9.5, p ≤ 0.004; Fig. 4A, B, C). This behavioral pattern has been demonstrated previously, for untreated and for vehicle treated animals, as demonstrated here (t₁₅ = 2.2, *p ≤ 0.045; Fig. 4B). Treatment with icv cEPo did not affect social preference in Escape mice, which continued to show more social preference than Stay mice (t₂₀ = 2.2, p ≤ 0.05; Fig. 4B), but increased social preference in Stay mice, and thus significantly increased time spent in contact with the social target (t₁₁ = 2.2, ⁺p ≤ 0.046; Fig. 4B). Similarly, in iDG injected Stay mice, cEPo treatment produced increased social preference (t₁₇ = 2.2, ⁺p ≤ 0.047; Fig. 4C). However, the degree of social preference was still less than for Escape mice, but not strictly significantly so (t₂₁ = 1.9, p ≤ 0.076). While the consequences of the SIP tests showed very similar trends for both icv and iDG treatments, we noted that the absolute level of behavior differed between vehicle treated animals for different experimental cohorts (icv vs iDG). It is possible that the surgery necessary for cannula placements, produced neuroinflammation which biased the results (Seyer et al., 2016) . Examining this possibility for icv surgeries, Seyer and colleagues suggested that although cytokines may be elevated immediately surrounding cannulae, that neither TNF_α and IL₆ were upregulated, and spatial and recognition memory (related to our tests) were not adversely affected (Seyer et al., 2016) . However, GFAP and CD_11b mRNA expression was elevated very close to cannulation. With that in mind, despite that fact that the trends for SIP and FC are very similar for both icv and iDG injections, and cannulation distances were very comparable, it may be that neuroinflammation modified behavioral outcomes.

Fig. 4. Social interaction in Stay mice is increased by cEPo.

A) Preference for social interaction yields more time spent near the jar with the social target (Phenotype effect: F_1,35 = 9.5, p ≤ 0.004). B) Stay mice treated icv with cEPo spent more time near the social target than those treated with vehicle (n = 10, icv-cEPo Stay; n = 8, vehicle Stay; t₁₁ = 2.2, ⁺p ≤ 0.046). Vehicle Escape and icv-cEPO Escape animals spent more time near the social target than their Stay counterparts (n = 12, vehicle Escape; vehicle Stay, n =8, t₁₅ = 2.2, *p ≤ 0.045; icv-cEPo Escape n = 12; icv-cEPo Stay n = 10, t₂₀ = 2.2, p ≤ 0.05). C) Similarly, iDG-cEPo treatment in Stay mice promoted more spent time near the social target than vehicle-treated Stay animals (n = 13, iDG-cEPo Stay; n= 10, Vehicle Stay; t₁₇ = 2.2, ⁺p ≤ 0.047). Although vehicle-treated Escape animals spent more time near the social target than Stay animals (*p ≤ 0.05) this relationship is lost after iDG treatment (t₂₁ = 1.9, p ≤ 0.076).

3.4. Fear conditioning responses (contextual and cued)

During the fear conditioning (FC) protocol each day (Fig. 5A), prior to socially aggressive interactions with a larger CD1 in the SAM (i.e. during isolation in an opaque cylinder) freezing time was measured, before (contextual conditioning) and after (cued) the 5 s tone (CS). During training, both contextual and cued responses are relatively high, even on Day 1, rise a little over time, and remain generally high (John et al., 2025). Contextual and cued conditioned freezing responses are positively correlated with freezing during aggressive SAM social interactions, but only in vulnerable Stay mice (John et al., 2025) . Freezing on Day 5, during the conditioned response test (CR; Fig. 5A, B; no large aggressive CD1 present outside the cylinder) vehicle Stay and Escape (icv experiments, Stay, n = 9, t₈ = 3.6, p ≤ 0.007; iDG experiments, Escape: n = 5, t₄ = 5.9, p ≤ 0.004; Stay: n = 10, t₉ = 4.9, p < 0.001) displayed more cued compared to contextual freezing. Following icv and iDG cEPo injections there was a cEPo-induced reduction in conditioned freezing. Specifically, icv cEPo treatment in Stay mice significantly (t₁₄ = 3.35, *p ≤ 0.0048) reduced cued freezing in the CR test. For intracranial iDG injection during CR, cEPo diminished cued freezing compared to vehicle treatment (t₂₀ = 3.7, *p < 0.001), even while in Stay mice treated iDG with vehicle (see above: t₉ = 4.9, p < 0.001) or cEPo (n = 12, t₁₁ = 3.3, p ≤ 0.007) cued freezing increased compared to contextual.

Fig. 5. Cue-induced freezing is reduced by icv and iDG-cEPo injection.

A) Fear Conditioning response test on Day 5 measured contextual (time in opaque cylinder dividing test and aggressor mice) and cued (tone) fear freezing, in which B, C) Vehicle Stay and Escape (icv experiments, Stay, n = 9, t₈ = 3.6, p ≤ 0.007; iDG experiments, Escape: n = 5, t₄ = 5.9, p ≤ 0.004; Stay: n = 10, t₉ = 4.9, p < 0.001) displayed more cued compared to contextual freezing. B) Importantly, icv-cEPo treatment in Stay mice significantly (t₁₄ = 3.35, *p < 0.0048) reduced cued freezing in the CR test. C) Additionally during CR, cEPo iDG injection decreased cued freezing compared to vehicle treatment (t₂₀ = 3.7, *p < 0.001), even though in Stay mice treated iDG with cEPo cued freezing was increased compared to contextual freezing (n = 12, t₁₁ = 3.3, p ≤ 0.007).

3.5. Gene expression is modified by icv or iDG cEPo treatment

In DG of mice treated with vehicle, Escape produced down regulation of Arc, Creb1, and Hcrtr2 gene expression (Fig. 6A). The Stay phenotype produced significantly elevated Egr1 and Hcrtr1 mRNA. The addition of Day 3 cEPo injection significantly reduced Stay Creb1 expression similar to vehicle Escape. Injection of cEPo also significantly stimulated Egr1 expression in Escape mice. The gene expression of Hcrtr1 of cEPo-injected mice was stimulated in Escape mice, and remained elevated in Stay mice, similar to vehicle-treated Stay mice. Additionally, in vehicle Escape mice, Hcrtr2 expression was significantly reduced compared to cage controls (Fig. 6A).

Fig. 6. Gene expression is significantly modified by acute cEPo treatment.

A) In the DG, cEPo treatment produced elevated expression of Egr1 and Hrcrt1 in Escape mice (n = 5, p < 0.05). Escape animals had decreased Hcrtr2 expression (Vehicle Escape; n = 5, p < 0.05; iDG-cEPo Escape; n = 5, p < 0.05). B) In hippocampal CA₁, iDG-cEPo treated Escape mice have higher Bdnf expression compared to other counterparts (n = 5, p < 0.05). Vehicle Stay animals had higher Hcrtr1 gene expression but not Hcrtr2 (n = 5, p < 0.05). Escape and iDG-cEPo stay animals had higher Hcrtr2 gene expression compared to other animals (Vehicle Escape; n = 5, p < 0.02; iDG-cEPo Stay; n = 6, p < 0.05). CC = cage control, EV = Escape Vehicle-treated, SV = Stay Vehicle-treated, EC = Escape cEPo-treated, SC = Stay cEPo-treated.

In the CA₁ region of the hippocampus, Stay phenotype alone promoted upregulation of Hcrtr1 expression, and Escape phenotype increased Hcrtr2 expression, in mice injected with vehicle (Fig. 6B). Injection of cEPo resulted in elevated Bdnf and diminished Creb1 expression in Escape phenotype mice, and elevated Hcrtr2 mRNA in Stay phenotype mice (Fig. 6B).

3.6. Hippocampal orexin-expressing cells produce Epor

Expression of mRNA for Epor (the EPo receptor) was extremely widespread in the dorsal hippocampus, with the greatest expression in DG (Fig. 7C, D, E). Thus, glutamatergic granule cells expressed Epor densely (in over 50% of these neurons). Although somewhat more sparsely, the excitatory pyramidal neurons of the cornu ammonis, such as the CA₁, also evinced dense expression of Epor (~ 25%), known to be elicited by learning (Almaguer-Melian et al., 2024). The hilus of the DG, expressed significant Epor, and there was Epor expression throughout the interstitial tissue.

Fig. 7. In hippocampus, gene expression for EpoR colocalizes with Orx₁ receptors in the DG and Orx₂ receptors in CA₁.

DG and CA₁ labeling A) Hcrtr₁, B) Hcrtr₂, C) Epor, and D) together as a merged image with DAPI (scale bar = 100 μm). E) In DG, there are significantly more Epor⁺ cells relative to cells expressing Hcrtr₁ or Hcrtr₂ (n = 12; F_2,9 = 46.47, p < 0.0001; Hcrtr₁+ vs Epor+: t₆ = 5.8 ***p ≤ 0.0011; Hcrtr₂+ vs Epor+: t₆ = 17.0 ****p < 0.0001). Examining Epor⁺ colocalization in DG, F) approximately 20% of Epor⁺ cells also express Hcrtr₁ (n = 12), with potentially fewer colocalizing Hcrtr₂⁺. In CA1, G) more cells contain Hcrtr₂ and Epor expression relative those expressing Hcrtr₁ (n = 12; F_2,9 = 7.605, p ≤ 0.0116; Hcrtr₁⁺ vs Hcrtr₂⁺: t₆ = 5.7 *p ≤ 0.0013; Hcrtr₁⁺ vs Epor⁺: t₆ = 3.3, *p ≤ 0.0169). H) Approximately 30% of Epor⁺ cells in CA1 also express Hcrtr₂ (n = 12; F_2,9 = 23.52, p ≤ 0.0003; Hcrtr₁⁺ vs Hcrtr₂⁺: t₆ = 5.0 ***p ≤ 0.0024; Hcrtr₂⁺ vs Hcrtr₁⁺ & Hcrtr₂⁺: t₆ = 5.0, ***p ≤ 0.0025). F, H) Very few cells colocalize expression of Epor⁺, Hcrtr₁⁺ and Hcrtr₂⁺.

The Orx receptors genes, Hcrtr1 and Hcrtr2, were well represented in DG, hilus, and CA regions (Yaeger et al., 2024a). Gene expression for Hcrtr1 was highly colocalized with Epor in the hilus and margins of the DG. The expression of Epor was significantly greater (F_2,9 = 46.47, p ≤ 0.0001; Hcrtr₁⁺ vs Epor⁺: t₆ = 5.8 ***p ≤ 0.0011; Hcrtr₂⁺ vs Epor⁺: t₆ = 17.0 ****p < 0.0001) than that of either Orx receptor. Approximately 20% of neurons in DG margin or hilus reflect colocalized expression of Epor and Hcrtr1. Only ~5% of Epor positive neurons also express Hcrtr2 or both Orx receptors (Hcrt₁⁺ vs Hcrt₁⁺ & Hcrt₂⁺: t₆ = 2.0, p ≤ 0.0938; Hcrt₁⁺ vs Hcrt₂⁺: t₆ = 2.0, p ≤ 0.0985). In CA1, more neurons express either Epor or Hcrtr2 (F_2,9 = 7.6, p ≤ 0.012; Hcrtr₁⁺ vs Hcrtr₂⁺: t₆ = 5.7 *p ≤ 0.0013; Hcrtr₁⁺ vs Epor⁺: t₆ = 3.3, *p ≤ 0.0169) than Hcrtr1, and ~30% of Epor positive cells in CA₁, also express Hcrtr2 (F_2,9 = 23.5, p ≤ 0.003; Hcrtr₁⁺ vs Hcrtr₂⁺: t₆ = 5.0 ***p ≤ 0.0024; Hcrtr₂⁺ vs Hcrtr₁⁺ & Hcrtr₂⁺: t₆ = 5.0, ***p ≤ 0.0025) compared to less than 5% colocalization with Hcrtr1 or both Orx receptors.

4. Discussion

Acute treatment with the non-hematopoietic EPo molecule cEPo, delivered either via cerebrospinal fluid in the brain (icv), or directly into the hippocampus, specifically intra-DG, produced anxiolytic and antidepressive effects (Miskowiak et al., 2008) in mice tested in the SAM. Specifically, when delivered on Day 3 of the SAM into the ventricles (icv) or the DG, stress-susceptible Stay mice begin to Escape, similar to previous experiments using anxiolytic and/or antidepressive drugs (Carpenter et al., 2023; Krupp et al., 2024; Smith et al., 2016; Staton et al., 2018; Yaeger et al., 2022b). Similarly, Escape mice begin to escape faster after icv or intra-DG cEPo treatment, compared within samples to Day 2 before treatment, as has been seen with anxiolytic or antidepressive drugs, but also by anxiolytic and/or antidepressive activities, such as exercise or learning (Almaguer-Melian et al., 2024; Carpenter and Summers, 2009; Miskowiak et al., 2012; Miskowiak et al., 2015; Smith et al., 2016; Staton et al., 2018; Summers et al., 2017; Yaeger et al., 2022c). For Escape mice, following heavy early attention to the escape route, motivation appears to be high, but following successful escape attempts, motivation is unchanged in icv or intra-DG cEPo treatments, late during SAM training (Day 4), after escape is well established. In contrast, this motivation is generally low for Stay animals, and only enhanced on Day 4 after cEPo treatment, which lowers anxiety, allows for escape in some animals, and enhances motivation, but is limited to the subset of Stay mice that change phenotype to Escape. These effects may, in part, be related to anti-inflammatory actions caused by cEPo (Levenga et al., 2021; Rahmani et al., 2022; Villa et al., 2007; Zhang et al., 2021), perhaps similar to those of EPo (Bond and Rex, 2014). We have demonstrated that the SAM protocol promotes neuroinflammation in the stress-vulnerable Stay phenotype only, commensurate with elevated plasma and central TNF_α, and reversible with the psychedelic drug (R)-DOI, as well as Akt₂ and Mtor gene expression (John et al., 2025; Krupp et al., 2024). In addition, social preference is increased in cEPo treated Stay mice. These Stay mice are significantly more fearful and exhibit elevated rates of cued freezing on training Days 2–4, and then again during the CR test (when only the CS [tone], and the US [aggressor], is present) on Day 5. This elevated cued fear conditioning is blocked by cEPo treatment. Finally, acute cEPo treatment also influences gene expression in the DG and CA₁ of the hippocampus.

The concepts of stress-vulnerability and resilience have generated greater examination recently related to temporary outcome versus long-term state expression, inflexible responsiveness, the possibility of global resilience expression, non-translational classical tests, ecological, ethological and evolutionary relevance, and multiple stress-related phenotypes (Ayash et al., 2023; Ayash et al., 2020; Blanchard et al., 2013; Haller and Alicki, 2012; Haller et al., 2013; Kalisch et al., 2015; Lyons et al., 2023; Mancini and Bonanno, 2009). For these reasons the SAM paradigm was designed to confront the challenges associated with multifactorial etiologies of stress, stress-induced behaviors, and the concepts of stress-vulnerability and resilience. Importantly, these multifactorially derived stress-states have the same basic etiologies and challenges as those of affective disorders in humans, like anxiety, depression, and PTSD. Stability of resilient and vulnerable phenotypes, as well as the celerity of onset, can be important. Our model has not only been demonstrated to have very rapid onset of phenotypic traits, but also stability of expression (John et al., 2025; Yaeger et al., 2024a). The fear conditioning element of our model is produced by an unpredictable US, aggression from a larger CD1 mouse, which by the nature of the US, presents the stimulus as inflated and deflated, at the whim of the aggressive animal. What is more, in our model, we measure not only freezing as a CR, but freezing during training, freezing during the aggressive interaction, escape from the aggressive interaction, motivation to escape, but also startles and jumping. These results are all phenotypically consistent. Stress-vulnerable mice (Stay) always show greater freezing (of all types), more startles, and more jumps. Only stress-resilient mice (Escape) exhibit increased attention to and exploration of the escape route (increased motivation), as well as escape behavior (John et al., 2025; Smith et al., 2014; Yaeger et al., 2024a; Yaeger et al., 2022c). The SAM protocol was specifically designed to be ethologically, ecologically, and evolutionarily relevant (Lyons et al., 2023; Smith et al., 2014) with versions of the SAM for trout, hamsters, and rats adjusted to meet these considerations (Arendt et al., 2012; Carpenter et al., 2023; Carpenter and Summers, 2009; Robertson et al., 2015; Summers et al., 2017). Our work in the SAM, examining generalization effects of PTSD in this animal model, suggest that carryover effects dramatically change the outcomes of classical tests like OFT and EPM (Yaeger et al., 2022c). Primary behavioral outcomes in the SAM are all directly related to the aggression - primarily the threat of aggression (John et al., 2025), and in the social interaction preference test (SIP) after the SAM, we are directly comparing social aversion with either avoidance of aggression or submission to it. Thus, all the tests in the SAM are integrated, and the choice of phenotype is strictly self-selected and bi-directional, Escape or Stay. Because of the two distinctive phenotypes we have been able to ascertain part of the neurocircuitry involved, and demonstrate that stress-vulnerability is dependent on anxiousness, and resilience the lack of it. Thus, anxiolytic drugs (such as CRF₁ antagonist, Orx₁ antagonists, Orx₂ agonists, NPS, and the psychedelic drug (R)-DOI) and anxiolytic behavior (exercise and familiarity) can reverse phenotype from Stay to Escape, and that anxiogenic drugs (yohimbine, Orx1 agonists) can change Escape mice into stress-vulnerable stay mice (Carpenter et al., 2023; Carpenter and Summers, 2009; Krupp et al., 2024; Smith et al., 2014; Smith et al., 2016; Staton et al., 2018; Summers et al., 2017; Yaeger et al., 2024a; Yaeger et al., 2022b; Yaeger et al., 2022c).

The neurocircuitry involved with anxiety, depression, and PTSD involves stress regulating regions of the brain, including prefrontal cortex, the amygdala and extended amygdala, and hippocampus (Korzan and Summers, 2021; Yaeger et al., 2020). Neurotrophic activity associated with affective disorders and stress-related behavior has been clearly demonstrated in hippocampus (Adlard and Cotman, 2004; Andero and Ressler, 2012; Arendt et al., 2012; Kozlovsky et al., 2007; Ninan, 2013). Previous work has also suggested EPo or cEPo activity in hippocampus is associated with decreased anxious and depressive behavior, as well as increased cognition (Almaguer-Melian et al., 2024; Girgenti et al., 2009; Miskowiak et al., 2012; Miskowiak et al., 2015; Saad et al., 2019; Sathyanesan et al., 2018) potentially through effects on gene expression, neural and/or synaptic plasticity (Girgenti et al., 2009; Kamal et al., 2011; Oh et al., 2012; Sathyanesan et al., 2018), and effects on learning and memory (Adamcio et al., 2008; Almaguer-Melian et al., 2024; Miskowiak et al., 2008; Miskowiak et al., 2012; Miskowiak et al., 2015; Newton and Sathyanesan, 2021). Most studies involving neurotrophin treatment take place over weeks of administration (Adamcio et al., 2008; Miskowiak et al., 2015) , but here we demonstrate that a single dose of cEPo delivered icv (100 ng) or iDG (10 ng) has acute effects on anxious and depressive behaviors, such as increased motivation for active social stress avoidance, enhanced stress avoidance (Escape and celerity to escape), increased preference for social interaction, and reduced cue-induced fear conditioned freezing (Carpenter et al., 2023; Carpenter and Summers, 2009; Smith et al., 2014; Smith et al., 2016; Staton et al., 2018; Yaeger et al., 2022b; Yaeger et al., 2022c).

The difficulty with our assessment of the results is that further mechanistic insights are necessary to distinguish the therapeutic potential for cEPo and a likely neurocircuitry of effect. The effects of cEPo are potentially actuated through EPo receptor homodimer or EPoR-βc (CD131; βc = common cytokine β subunit) heterodimer complexes (Brines et al., 2004; Sadanandan et al., 2023), and has been demonstrated to increase Epor and Csf2rb (Cd131, Il3rb, Il5rb) gene expression (Sadanandan et al., 2023). Our results suggest that Epor gene expression is widely exerted in the DG of the hippocampus, appearing in over 50% of DG cells. Similarly, we report that iDG-cEPo injection dramatically alters orexin Hcrtr1 and Hcrtr2 gene expression, that orexinergic activity in the DG may play an important role. Neuroprotective effects of EPo require enhanced gene expression of its receptors, Epor (Sanchez et al., 2009), suggesting that neuroprotective effects of cEPo also require gene expression of Epor. However, rapid effects of neurotrophic action (Kovalchuk et al., 2004) suggests activation of subcellular molecular signaling, relatively independent from gene expression. Relative to anxious and depressive behaviors, this kind of rapid neurotrophic action has been linked to psychiatric treatments with fast action times, such as ketamine and psychedelics which promote increased synaptic neurotrophins and receptor tyrosine kinases (TrK, such as TrK_{A or B}) (Kim et al., 2024). While EPoR activate JaK₂/STAT₅, as well as PI₃K/Akt, and ERK/MAPK intracellular molecular signaling pathways to produce erythropoiesis (Tóthová et al., 2021), it is not yet clear which of these, or other pathways, influence rapid social learning and stress-related behavioral responses.

However, while rapid actions may not require gene expression, in the BLA, an acute injection of the orexin 1 receptor (Orx₁R) antagonist SB-674042, while similarly reducing stress-related affective behaviors, also produce and increase Mapk3 gene expression (Yaeger et al., 2022b). In the experiments presented here, even with a short duration of cEPo availability in hippocampus, there was increased Egr1 and Hcrtr1 gene expression in DG in Escape phenotype mice. Interestingly, Stay mice also showed elevated Egr1 and Hcrtr1 gene expression in DG (also CA₁ for Hcrtr1) without cEPo treatment (vehicle-treated), suggesting that these effects were dependent on phenotype. In CA₁, cEPo treatment increases Bdnf, and reduces Creb expression, but only in Escape mice. Vehicle-treated Escape mice experience increased Hcrtr2 expression in CA₁, but Stay mice show this effect following cEPo treatment. These phenotype dependent effects are a distinctive advantage of the SAM defeat/avoidance paradigm. Our results suggest similarly that phenotypic-specific behavioral and gene expression results are also characteristic of icv and iDG cEPo treatments.

The effects of SAM and cEPo treatment on Hcrtr1 and Hcrtr2 gene expression in DG and CA₁ regions of the hippocampus, are occurring where EPo receptor gene (Epor) expression is greatest. In DG, Epor expression (Epor⁺) is found in approximately 50% of cells in this region, and about half that many (~25%) in CA₁. Many fewer cells in DG and CA₁, 20% or less, express Hcrtr1 and Hcrtr2, also characteristic of other limbic brain regions, such as the BLA (Yaeger et al., 2022b; Yaeger et al., 2022c). However, it is important to note that as many as 20% of DG Epor expressing cells co-express Hcrtr1, while approximately 5% of Epor⁺ cells express Hcrtr2. This pattern is reversed in CA₁, where 30% of Epor⁺ positive cells also express Hcrtr2, but less than 5% express Hcrtr1. The relatively high and region specific expression of Hcrtr1 and Hcrtr2 in hippocampal granule and pyramidal neuron regions have led us recently to propose a stress-spatial-learning related circuitry that connects the cortex, amygdala, and hippocampus to promote integration of complex behavioral outcomes associated with social interactions (Yaeger et al., 2024a). The results reported here suggest that Epor may be an important component of stress-related circuits that integrate social cognition with spatial learning to promote complex adaptive behaviors (Almaguer-Melian et al., 2024; Osborn et al., 2013; Sargin et al., 2011).

5. Conclusions

This study demonstrates anxiolytic and antidepressive cEPo effects following a singular injection. The acute action is an important distinction relative to the typical expected effects for neurotrophins, but also in comparison with widely available clinical antidepressants. It is significant that the anxiolytic and antidepressive effects are limited to stress-vulnerable individuals, which is an important component for a therapeutic treatment. Our study also demonstrates that the anxiolytic effect, revealed in stress-vulnerable animals (Stay) switching to a stress-resilient phenotype (Escape), also increases these individuals to greater motivation. Greater motivation increases activity related to attentiveness. Similarly, reduced anxiety increases social preference and decreases fear conditioning. This study also increased the circuitry-based understanding of the molecular actions of cEPo. Finally, the dramatic quantification of differences in Epor expression (particularly in DG), gives a reasonable starting place for understanding of the neurocircuitry involved in cEPo actions. The relatively high colocalization in DG with Hcrtr₁ and in CA₁ with Hcrtr₂ expression, also suggest that orexin may be playing a role in the effects of EPo or cEPo as a neuromodulator. In addition, for regions of the hippocampus, DG and CA₁, that express significant Epor gene expression, an acute single treatment of carbamoylated EPo, either icv or iDG, modifies expression of genes known to influence neuroplasticity, learning and behavior, Bdnf, Creb1, Egr1 expression, as well as orexin receptor genes, Hcrtr1 and Hcrtr2, concurrent with modifying affective behavior in a phenotypically dependent manner. Thus, in stress-vulnerable Stay mice, tested in the SAM protocol with larger aggressive CD1 mice, icv or iDG cEPo increased motivation for active social stress avoidance, enhanced stress avoidance (Escape and celerity to escape), increased preference for social interaction, reduced cue-induced fear conditioned freezing. These results suggest that the effects of cEPo can be relatively fast acting on behavior, while beginning the process of activating or down-regulating gene expression, as well as having effects from longer duration treatments.

Highlights.

• cEPo produces anti-stress behavioral changes in male mice with anxious phenotype

• A single dose of cEPo delivered icv or intra-DG produces anxiolytic effects

• cEPo-driven anxiolysis yields increased behavioral motivation for decision making

• Social preference is increased in anxious animals receiving cEPo treatment

• Cue-induced fear conditioning is reduced by cEPo treatment

Abbreviations:

aCSF

artificial cerebrospinal fluid

Arc

activity-regulated cytoskeletal gene

BDNF

brain-derived neurotrophic factor

Bdnf

brain-derived neurotrophic factor gene

BLA

basolateral amygdala

C57BL/6NHsd

a strain of black mice used for stress testing

CA₁

Cornu Ammonis 1

CC

cage control

CD1

Hsd:ICR retired breeder mice used as aggressors

CD11b

Intergrin alpha M

CeA

central amygdala

cEPo

carbamoylated erythropoietin

CR

conditioned response

CS

conditioned stimulus

Creb1

cAMP responsive element binding protein 1 gene

CS

cEPo-treated Stay mice

ΔCt

difference between target gene value of corresponding endogenous reference gene

ΔΔCt

fold change of target gene expression relative to reference sample, normalized to reference gene

CR

cued response

DG

Dentate Gyrus

EC

Escape cEPo-treated

Egr1

early growth response gene 1

EPo

erythropoietin

EpoR

erythropoietin receptor

Epor

erythropoietin receptor gene

EV

Escape Vehicle-treated

g

gram(s)

ga

gauge

Gapdh

housekeeping gene

GFAP

glial fibrillary acidic protein gene

Hcrtr1

Orexin receptor 1 gene

Hcrtr2

Orexin receptor 2 gene

icv

intracerebroventricular

iDG

single intra-dentate gyrus

NIH

National Institutes of Health

NIMH

National Institute of Mental Health

PBS

phosphate buffer solution

PTSD

post-traumatic stress disorder

s

seconds

SAM

Stress Alternatives Model

SC

Stay cEPO-treated

SIP

Social Interaction Preference

SV

Stay Vehicle-treated

SVZ

subventricular zone

TNF_α

tumor necrosis factor alpha

US

unconditioned stimulus