Adolescent Intermittent Ethanol (AIE) sensitized fever in male Sprague Dawley rats exposed to Poly I:C in adulthood

Abstract

Fever plays an indispensable role in host defense processes and is used as a rapid index of infection severity. Unfortunately, there are also substantial individual differences in fever reactions with biological sex, immunological history, and other demographic variables contributing to adverse outcomes of infection. The present series of studies were designed to test the hypothesis that a history of adolescent alcohol misuse may be a latent experiential variable that determines fever severity using polyinosinic:polycytidylic acid (poly I:C), a synthetic form of double-stranded RNA that mimics a viral challenge. Adult male and female Sprague Dawley rats were injected with 0 (saline) or 4 mg/kg poly I:C to first establish sex differences in fever sensitivity in Experiment 1 using implanted radiotelemetry devices for remote tracking. In Experiments 2 and 3, adolescent males and females were exposed to either water or ethanol (0 or 4 g/kg intragastrically, 3 days on, 2 days off, ~P30-P50, 4 cycles/12 exposures total). After a period of abstinence, adult rats (~P80–96) were then challenged with saline or poly I:C, and fever induction and maintenance were examined across a prolonged time course of 8 hours using implanted probes. In Experiments 4 and 5, adult male and female subjects with a prior history of adolescent water or adolescent intermittent ethanol (AIE) were given saline or poly I:C, with tissue collected for protein and gene expression analysis at 5 hours post-injection. Initial sex differences in fever sensitivity were minimal in response to the 4 mg/kg dose of poly I:C in ethanol-naïve rats. AIE exposed males injected with poly I:C showed a sensitized fever response as well as enhanced TLR3, IκBα, and IL-1β expression in the nucleus of the solitary tract. Other brain regions related to thermoregulation and peripheral organs such as spleen, liver, and blood showed generalized immune responses to poly I:C, with no differences evident between AIE and water-exposed males. In contrast, AIE did not affect responsiveness to poly I:C in females. Thus, the present findings suggest that adolescent binge drinking may produce sex-specific and long-lasting effects on fever reactivity to viral infection, with preliminary evidence suggesting that these effects may be due to centrally-mediated changes in fever regulation rather than peripheral immunological mechanisms.

1. Introduction

Alcohol use typically begins during adolescence, with early onset of drinking associated with a substantially greater risk for development of alcohol use disorder (AUD) later in life (Hingson et al., 2006; Sartor et al., 2016). According to Substance Abuse and Mental Health Services Administration (SAMHSA), 5.8 million underage drinkers aged 12 to 20 were past month alcohol users, while 3.5 million of this age group reported binge drinking in the past moth in 2022 (SAMSHA, 2023). Binge pattern of drinking (at least five drinks for males and four drinks for females within a 2-hour period, resulting in blood alcohol levels above 0.08 g/dL) is particularly harmful for the developing adolescent brain. Findings from human longitudinal studies suggest that adolescent (Chung et al., 2018; SAMHSA, 2023; Schulteis et al., 2008) binge drinking is associated with long-term neural and cognitive alterations. Adolescents with a history of binge drinking often demonstrate impairments in executive function, attention, and memory as well as enhanced risky decision making and impulsive behavior (Ewing et al., 2014, Lees et al., 2020; Spear, 2018). Alcohol-induced innate immune activation of the brain may play a substantial role in these alterations (Coleman & Crews, 2018; Pascual et al., 2018) given results of animal studies that have demonstrated the long-term impact of adolescent intermittent ethanol (AIE) exposure on immune function (Cui et al., 2014; Erickson et al., 2019). Furthermore, AIE-associated immune alterations may have lifelong deleterious effects on the ability of an individual’s immune system to defend against pathogenic events (Deak et al., 2022).

Inflammation and other components of host defense are mediated by cytokines, a category of signaling proteins synthesized by immune cells. Chronic consumption of ethanol can induce systemic inflammation in both humans and laboratory rodents through interaction with Toll-like receptors (TLRs), such as TLR3 and TLR4, which initiate the production of pro-inflammatory cytokines (Alfonso-Loeches et al., 2010; Erickson et al., 2019; Montesinos et al., 2016; Vetreno & Crews, 2012). As such, chronic ethanol exposure can have a profound impact on neuroimmune reactivity to pathogens and, as a result, may increase susceptibility to bacterial and viral infection (Lewis et al., 2023; Wood et al., 2013). Indeed, HIV, HPV, viral respiratory infections, and pneumonia are more prevalent among humans with a history of alcohol misuse, and these individuals are likewise more vulnerable to the inflammatory effects of ethanol itself (Szabo & Saha, 2015). Given the prevalence of drinking initiated in early development, it is important to public health to examine the mechanisms by which adolescent alcohol exposure can mediate the adult response to infection.

Across species, challenges to the immune system invoke a set of physiological and behavioral responses referred to as the sickness syndrome (Saper et al., 2012). Components of the sickness syndrome include anhedonia, fatigue, body aches, and fever; this combination optimizes immune function to combat infection (Saper et al., 2012). Fever is generally a prostaglandin- and cytokine-mediated response to inflammation that originates in the CNS with input from the periphery and serves to stimulate the immune system by elevating core body temperature (Evans et al., 2015; Romanovsky et al., 2005; Saper et al., 2012). Compartments of the hypothalamus, such as the paraventricular nucleus (PVN) and medial preoptic area (mPOA) integrate inputs from other brain regions such as the nucleus of the tractus solitaris (NTS) and peripheral immune organs (e.g. spleen, liver, etc.) and white blood cells in order to regulate this response. This hyperthermia increases the replication and activity of white blood cells and creates a hostile environment for pathogens to replicate, ultimately decreasing inflammation and the duration of infection (Djaldetti & Bessler, 2015). The febrile response is initiated when pathogen-associated molecular patterns (PAMPs) interact with specific pattern-recognition receptors (PRRs) on cells of the innate immune system, triggering the production of pro-inflammatory cytokines with pyrogenic properties (Evans et al., 2015).

TLRs are a family of membrane-associated PRRs that have been widely studied for their roles in fever induction and the effects of alcohol on host defense (Donnadieu-Rigole et al., 2016; Muralidharan et al., 2014; Romanovsky et al. 2006). TLR4, the receptor for gram-negative lipopolysaccharide (LPS) that mediates inflammatory response against gram-negative bacteria, may play a substantial role in ethanol-induced neuroinflammation given ethanol-associated activation of TLR4 signaling (Alfonso-Loeches et al., 2010; Erickson et al., 2019; Montesinos et al., 2016). AIE exposure has been shown to upregulate TLR4 gene expression in male rats (Vetreno & Crews, 2015), whereas TLR4 knockout female mice did not show AIE-associated neuroinflammation (Montesinos et al., 2015). AIE exposure of male rats also has been shown to upregulate TLR3 (Vetreno & Crews, 2012), which is expressed on endosomes within cells and binds double stranded RNA (dsRNA) molecules characteristic of viral pathogens, or, in experimental conditions, a synthetic form of dsRNA, polyionosinic:polycytidylic acid (poly I:C), which serves as a viral mimetic (Fortier et al., 2004). As TLRs recognize and bind PAMPs, cells of the innate immune system increase production of pro-inflammatory cytokines (Interleukins [IL]-1β and IL-6 in particular) that in turn increase levels of prostaglandin E2 (PGE2) in the anterior preoptic area of the hypothalamus (AH-POA). The ultimate effect of PGE2 signaling in the AH-POA is the elevation of core body temperature – fever – to combat infection (Evans et al., 2015; Lazarus et al., 2007).

Sex differences in baseline cytokine expression and fever are evident in both humans and rodents (Klein & Flanagan, 2016), and appear to share similarities across TLR receptor families. When TLR4 receptors were stimulated with LPS, males had a stronger fever response than females as well as greater pro-inflammatory cytokine expression (Asai et al., 2001; Klein & Flanagan, 2016; Marriott et al., 2006; Murakami & Ono, 1987). Similar findings were evinced when poly I:C was used to activate TLR3 receptors in mice, with males demonstrating greater elevations of cytokines and chemokines in the hippocampus than females after intraventricular administration of poly I:C (Coelho et al., 2021; Posillico et al., 2021). Importantly, acute ethanol administered within a few hours of Poly I:C suppressed cytokine reactivity in both sexes (Glover et al., 2011), an effect that appears to be independent of ethanol-induced corticosterone release (Glover et al., 2009; Glover & Pruett, 2006). In contrast, some studies have shown that chronic ethanol led to heightened cytokine reactivity to Poly I:C challenge (Lawrimore & Crews, 2017; Qin & Crews, 2012). Activation of TLR3 signaling with poly I:C increased ethanol intake in male mice, with no change or even decreased ethanol consumption evident in their female counterparts (Warden et al., 2018, 2019). We have also observed sex differences in long-term effects of exposure to ethanol during adolescence, with AIE impairing LPS-induced cytokine induction within circulating leukocytes in adult male but not female rats (Vore et al., 2017). Additionally, a modified AIE exposure in male rats induced impairment in LPS-induced fever (Telles et al., 2017), as well as decreased the mass of brown adipose tissue, the primary site of febrile thermogenesis in rodents (Cruz et al., 2020). In contrast, the effects of AIE exposure on TLR3-mediated neuroimmune pathways remain underexplored, and it is not known whether these effects are sex-dependent. Therefore, the present series of studies was designed to assess effects of AIE in males and females using an alcohol exposure model that has previously been shown to affect cytokine expression in both circulating immune cells and the CNS (Gano et al., 2019; Vore et al., 2017, 2021) on adult responses to TLR3 agonist challenge by examining detailed time courses of fever as well as brain, liver, spleen, and blood cytokine as well as chemokine expression associated with peaks of the acute fever response.

2. General Methods

2.1. Subjects.

All subjects were male and female Sprague-Dawley rats born in our colony at Binghamton University from breeders originally acquired from Envigo. Litters were culled to 10 pups on postnatal day (P)1, with equal sex ratios when possible. Animals were weaned on P21 and pair-housed with non-littermates with continuous ad libitum access to food and water. The vivarium was maintained at 22±1°C on a 12:12 light cycle (lights on at 0700) with humidity set to vary naturally between 20–60% resulting in average humidity levels of 30–40%. Cage mates were assigned to the same experimental conditions, and no more than 1–2 rats of the same sex from any given litter were used per experimental group. All animals were handled (2–3 minutes per day for two days) prior to each phase of the experiment to reduce experimenter-induced stress on the day of procedures. All procedures were approved by the Institutional Animal Care and Use Committee at Binghamton University (protocol 847–21) and rats were treated in accordance with Public Health Service (PHS) policy.

2.2. Adolescent intermittent ethanol (AIE) exposure.

Starting in adolescence (P28–32), rats received once-daily intragastric (i.g.) exposures to 4.0 g/kg ethanol (AIE groups) or the equivalent volume of water vehicle (intubation of tap water; VEH groups). All animals received 3 consecutive days of i.g. intubations followed by 2 days of remaining undisturbed in the home cages. Collectively, each of these 5-day periods formed a cycle, and the entire exposure regimen consisted of 4 cycles (12 total ethanol intubations) from P28–32 to P48–52. Ethanol solutions (20% v/v) were mixed fresh daily using 95% ethanol stock and tap water. Animals were weighed in the morning of each exposure 60+ min prior to intubation. Though no blood ethanol concentrations (BECs) were quantified in these experiments for fear that collection procedures would influence later fever responsivity, prior work from our group has repeatedly observed BECs in response to the initial dose typically range from 175–200 mg/dl and drop by ~25% after 6 intubations (mid-point of AIE). A timeline of this procedure and other experimental manipulations can be found in Supplemental Figure 1.

2.3. Adult poly I:C challenge.

Poly I:C is a commonly used experimental model of viral infection and has additional relevance due to its use as an adjuvant in many vaccine formulations (Martins et al., 2015). Low molecular weight poly I:C (InvivoGen, San Diego, California) was reconstituted using sterile water from the manufacturer in accordance with instructions provided, and diluted using sterile, pyrogen-free physiological saline (InvivoGen, San Diego, California). Drug was prepared fresh and used within 2 hours. Poly I:C was injected intraperitoneally (i.p.) at a dose of 4 mg/kg, in a volume of 1 mL/kg, with equivalent volumes of saline delivered as vehicle for control group injections.

2.4. Adult probe implantation (Experiments 1–3).

In adulthood (P73–97 across experiments), rats were anesthetized with isoflurane (1–3% in oxygen) for the implantation of telemetry probes (PTD 4000 E-MITTER, STARR Life Sciences, Oakmont, PA) for monitoring of body temperature and locomotor activity. Probes were inserted either abdominally (Experiments 1, 2a, 3a) or at the nape of the neck (Experiments 2b, 3b). For abdominal insertion, the midline abdomen was shaved, and a sterile #10 surgical blade was used to make a small incision (~1cm) superficial to midline, with scissors used to further open the peritoneal cavity as previously described (Barnum et al., 2008; Deak et al., 2005). The sterilized telemetry probe was inserted into the abdominal cavity and sutured (3–0 absorbable sutures) to the inside of the abdominal wall. Additional sutures were used to close the abdominal wall, and skin was wound clipped (9 mm autoclips) shut. Animals were monitored daily for signs of infection, and wound clips were removed on day 7 provided full healing. Testing took place a week after clip removal. After piloting procedures in our lab showed that subcutaneous probe insertion at the nape of the neck would allow for equally responsive monitoring of body temperature but would require a much less invasive surgical procedure, animals in Experiments 2b and 3b were implanted in this manner. Scissors were used to make a cut just above shoulder blade level, and the sterile probe was inserted subcutaneously. Wound clips were used to close the opening, and similarly to what was described before, rats were monitored for 7 days and clips were removed a week prior to experimentation. For abdominal surgeries, rats were given buprenorphine (0.05 mg/kg i.p.) prior to surgery, with 3 post-operative doses on a 12-hr recovery schedule. For studies in which probes were implanted at the nape of the neck, a ring block procedure using bupivacaine (2.5 mg/kg 25% without epinephrine, subcutaneous) was utilized to mitigate pain prior to incision.

2.5. Body temperature (Experiments 1–3).

For all experiments where body temperature was monitored, the timeline for testing was identical. Rats were tested as adults (P83–120 across experiments) in counterbalanced squads of 8 per day, with handling performed and weights recorded in the days leading up to testing. On test day, rats were moved to the testing room an hour after lights-on (0800). Each single-housed rat’s home cage (including water and food) was placed into a sound-attenuating chamber on top of a platform that scanned the probe continuously, recording core body temperature as well as locomotor activity data. After 3 hours of baseline recording, rats were injected with poly I:C or saline vehicle (1100). After injection, rats were monitored for 8 hours and were returned to the colony at lights-off (1900).

2.6. Tissue collection (Experiments 4 & 5).

Trunk blood and brain tissue were harvested using rapid unanesthetized decapitation after the final adult challenge. Trunk blood was collected in EDTA-coated Vacutainers (BD Vacutainers, VWR cat. no. VT6450, Radnor, PA), and plasma was separated via refrigerated centrifugation and stored at −80°C until time of use. Brains were extracted and stored at −80°C until subsequent tissue dissection as described elsewhere using a cryostat (Gano et al., 2017). Briefly, brain regions were excised using biopsy punches using a Paxinos and Watson rat brain atlas for guidance (2004). The NTS was extracted starting at Bregma 11.52 mm using 2 bilateral 2 mm wide × 2 mm deep punches; the mPOA was excised starting at Bregma 0.12 mm using 3 × 1.2 mm wide × 1 mm deep punches; the PVN was excised starting at Bregma −1.08 mm using 3 × 1.2 mm wide × 1 mm deep punches.

2.7. Real time RT-PCR (Experiments 4 & 5).

Once dissected, brain regions were analyzed using real time RT-PCR. All the procedures are described in detail in Doremus-Fitzwater et al. (2015). Briefly, tissue was placed into a 2.0 mL Eppendorf tube containing 500 μL Trizol RNA reagent and a 5 mm stainless steel bead, and then homogenized using a Qiagen Tissue Lyser II. Following homogenization, RNA was extracted using Qiagen’s RNeasy mini kit (catalog #74106), according to manufacturer’s instructions. Synthesis and storage of cDNA included a DNAse treatment step, and probed cDNA amplification was performed and captured in real time using the CFX384 Real-Time PCR Detection System (Bio-Rad, #185–5485). All PCR data were adjusted using the 2^−ΔΔC(T) method (Livak & Schmittgen, 2001) and are shown here relative to a stable reference gene (cyclophilin) and expressed as percent of the basal control group. Of the housekeeper genes tested (GAPDH, beta-actin, cyclophilin), cyclophilin showed fewest effects of manipulation and was selected as the most stable gene for PCR adjustment. Housekeeper graphs and analyses can be found in supplemental data (Supplemental Figure 2). Genes of interest included cFos as a marker of general cellular activation; Iκ-Bα as a marker of Nf-kB activation; and cytokines IL-1β, IL-6, and TNFα.

2.8. Plasma Magpix cytokine panel (Experiments 4 & 5).

Plasma samples were assayed on first thaw using a magnetic bead-based multiplex protein bio-plex in accordance with manufacturer’s instructions for plasma samples (23-plex, Bio-Rad Magpix, cat no. 171-K1001M; Bio-Rad, Hercules, CA). This multiplex assayed 23 targets of interest. The targets examined included Tumor necrosis factor (TNF) α, Interleukins(IL)-6, IL-1β, IL-1α, IL-2, IL-4, IL-5, IL-6, IL-7, IL-10, IL-12p70, IL-13, IL-17, IL-18, Granulocyte-colony stimulating factor (G-CSF), Granulocyte-macrophage colony-stimulating factor (GM-CSF), C-X-C motif ligand 1 (CXCL1/GRO/KC), Interferon (IFN) γ, chemokine (C-C motif) ligand 2/macrophage colony-stimulating factor (CCL2/MCP-1), CCL3/macrophage inflammatory protein (MIP) 1α, CCL20 (aka MIP3α), CCL5/regulated upon activation normal T cell expressed (RANTES), and vascular endothelial growth factor (VEGF).

2.9. Experiment 1 design and procedures.

Naïve male and female rats were raised in our colony until weaning as described in general methods, and then left alone to mature to adulthood. As adults, all rats received abdominal probe insertion surgery and, after recovery, were randomly assigned to experimental conditions of being acutely challenged with either 0 (saline) or 4 mg/kg poly I:C. The design of this experiment was a 2 (sex) × 2 (challenge) factorial (n = 6/group for males, 5/group for females, N = 22), with all animals included in the final analysis.

2.10. Experiment 2 design and procedures.

This experiment was designed to test an impact of AIE on the fever response to an acute challenge with Poly I:C in adult male subjects. Males from our colony were exposed to water vehicle or ethanol during adolescence and implanted with either abdominal (Experiment 2a) or subcutaneous (Experiment 2b) telemetry probes in adulthood. The purpose of testing two distinct sites for telemetry probe implantation was to determine whether it would be feasible to utilize subcutaneous probe implantation to achieve sensitive measures of body temperature. Subcutaneous probe implantation has the advantage that it is a far less invasive procedure than abdominal surgery, can be performed with a local analgesic, and is less likely to change immunological status of the peritoneal cavity into which Poly I:C will eventually be injected. Testing was performed as indicated in general methods. The design of each experiment was a 2 (adolescent exposure) × 2 (adult challenge) factorial (n = 8–9/group, N = 33 for Exp. 2a; n = 9–11/group, N = 39 for Exp. 2b), with all animals included in the final analysis.

2.11. Experiment 3 design and procedures.

Females from our colony were exposed to AIE or water vehicle and implanted with either abdominal (Experiment 3a) or subcutaneous (Experiment 3b) telemetry probes in adulthood. Testing was performed as indicated in general methods. The design of each experiment was a 2 (adolescent exposure) × 2 (adult challenge) factorial (n = 6–8/group, N = 28 for Exp. 3a; n = 8–10/group, N = 35 for Exp. 3b), with all animals included in the final analysis.

2.12. Experiment 4 design and procedures.

The goal of Experiments 4–5 was to identify candidate immune genes/proteins that may contribute to the sensitized fever response in AIE-exposed males. Our approach was to utilize a combination of RT-PCR and multiplex protein analysis to identify candidate immune signaling molecules in two key immunological organs (spleen, liver) and CNS sites involved in fever regulation (mPOA, NTS) or stress reactivity (PVN). In Experiment 4, males from our colony were exposed to water or ethanol in adolescence. In adulthood, these experimental subjects were challenged with saline or poly I:C (4 mg/kg), and tissue was collected for analysis (brains and peripheral organs for PCR, blood for protein analysis via multiplex) at 5 hours post-injection, a time point that was found to show established peak fever in Exp. 1–3. The design was a 2 (adolescent exposure) × 2 (adult challenge) factorial, n = 8 per group, N = 32. with no more than one animal per group removed from selected brain regions if found to be in violation of Grubb’s test. Half of the mPOA was lost due to technical error, shown here as n = 2–6/group, N = 16.

2.13. Experiment 5 design and procedures.

Females from our colony were exposed to water vehicle or ethanol during adolescence. In adulthood, experimental subjects were challenged with 4 mg/kg poly I:C or given a control injection of saline, and tissue was collected (brains and peripheral organs for PCR, blood for protein analysis via multiplex) at 5 hours post-injection, a time point that was found to show established peak fever in Experiments 1–3. The design was a 2 (adolescent exposure) × 2 (adult challenge) factorial (n = 8 per group, N = 32), with all females were kept for analysis, except for the few removed from selected brain regions, if found to be in violation of Grubb’s test (no more than 1/group).

2.14. Statistical analyses.

Body temperature data in Experiments 1–3 was adjusted to baseline (the average of the last two sampling periods before injection – 30 min and 1hr before challenge) and was analyzed using multiple approaches to assess the various stages of fever responses. Raw data was also initially assessed and also supported the outcomes presented here; these analyses can be found in supplemental figure 4 and the supplemental data file. First, data were analyzed using 2 (adolescent exposure: VEH vs AIE) × 2 (adult challenge: saline vs. poly I:C) × 22 (time) mixed ANOVAs (with adult challenge and adolescent exposure as between-subject variables and time as a repeated measure) to establish the main effect of acute poly I:C challenge, referred to as the “omnibus ANOVA” as our most overarching initial analysis (in Experiment 1, sex was analyzed instead of adolescent exposure as a between-ss measure). The initial reaction to injection was analyzed using 2 (sex in Exp1, adolescent exposure in Exp 2/3) × 2 (challenge: saline vs. poly I:C) × 5 (time: last 2 time points of baseline, next 3 time points/ 1.5 hours after injection) mixed ANOVAs. Data were then analyzed first using 2 (sex or adolescent exposure) × 22 (time) ANOVAs in poly I:C groups only, and then in smaller parcels to determine effects on various parts of the fever response. The rising phase of the fever response (1.5 – 4 hours after challenge) was analyzed using 2 (sex or adolescent exposure) × 8 (time: last 2 time points of baseline, 6 time points of fever) mixed ANOVAs, and the maintenance phase of fever (4.5 – 7.5 hr after challenge) was analyzed with 2 (adolescent exposure) × 9 (time: last 2 time points of baseline, 7 time points of fever). Post hoc tests for main effects of time were done using Dunnett’s test to compare time points of interest to the last two time points during baseline (to be maximally conservative, only those that differed from both were considered significant). Peak fever, peak fever change from baseline, latency to reach peak fever, and area under the curve (AUC) were analyzed for the poly I:C groups only using independent group Student’s t-tests. Activity data were analyzed using 2 (sex: M vs. F or adolescent exposure: VEH vs AIE) × 2 (adult challenge: saline vs. poly I:C) × 11 (time) mixed ANOVAs, and post hoc testing was performed using the Newman-Keuls test. For Experiments 4–5, PCR and Magpix data were analyzed using 2 (adolescent exposure: VEH vs AIE) × 2 (adult challenge: saline vs. poly I:C) ANOVAs, with post hoc testing for interactions performed using Tukey’s test.

3. Results

3.1. Experiment 1 (Figure 1)

Figure 1. The 4 mg/kg poly I:C (or saline vehicle) febrile time course for Experiment 1 ethanol-naïve males and females.

Data are shown as (A) adjusted to baseline (final two timepoints of the “acclimation” time period) with mean ± SEM shown for each time point. Other metrics such as (B) peak fever in degrees Celsius, (C) peak degrees Celsius change at the highest point of fever as compared to baseline, (D) latency to reach the peak fever in hours, and (E) total area under the curve for the duration of fever are shown with mean and quartile ranges marked via violin plot. Activity data (F) is shown in 1-hr bins. Statistical markers are decoded in the figure, with all p < 0.05.

3.1.1. Fever time course.

The omnibus ANOVA revealed main effects of challenge (F_{1, 18} = 32.13, p < 0.0001, ηp² = 0.64) and time (F_{21, 378} = 25.74, p < 0.0001, ηp² = 0.59, as well as sex × time (F_{21, 378} = 2.10, p < 0.01, ηp² = 0.10), and challenge × time (F_{21, 378} = 20.93, p < 0.0001, ηp² = 0.54) interactions.

During just the injection phase, there was a main effect of time, (F_{4, 72} = 10.07, p < 0.00001, ηp² = 0.36), and a sex × time interaction (F_{4, 72} = 3.63, p < 0.01, ηp² = 0.17). Specifically, only females showed an elevated core body temperature in the first half-hour time point post-injection relative to baseline (p < 0.001), with female core body temperature significantly higher than that of males (p < 0.01).

The poly I:C-injected groups were analyzed with the full ANOVA yielding a main effect of time (F_{21, 189} = 31.78, p < 0.0001, ηp² = 0.78). A follow-up Dunnett’s test comparing all time points to the last half-hour of baseline showed that starting at 2.5 hours after injection, males and females showed significant increases in body temperature (p < 0.001 at 2.5hr, < 0.0001 all time points after). The rising phase of fever and the maintenance phase of fever both revealed effects of time only (respectively, F_{7, 63} = 33.26, p < 0.0001, ηp² = 0.78, F_{8, 72} = 33.98, p < 0.0001, ηp² = 0.79). Post hoc tests indicated that fever became significant at 2.5 hours after injection (p < 0.0001) and remained so for the duration of testing, further confirming results from the overall ANOVA (see Figure 1A).

There were no effects of sex on peak, peak change from baseline, latency, or AUC (Figure 1B, C, D, and E, respectively).

3.1.2. Activity data.

The ANOVA of locomotor activity during testing revealed a main effect of time (F_{10, 180} = 25.56, p < 0.0001, ηp² = 0.58) and a time × challenge interaction (F_{10, 180} < 4.06, p < 0.0001, ηp² = 0.18) (Figure 1F). Post hoc analysis indicated that both saline and poly I:C groups showed an initial burst of movement in the first hour of testing that was significantly higher than at any other time point (p < 0.001 for all; likely an exploration effect as they were relocated to the testing environment), and that during the first hour post-injection, both groups moved more than during the last hour of their baseline (p < 0.01). However, the poly I:C group continued to show increased activity through the 3-hour time point post-injection (body temperature analysis indicated this to be the time of fever onset, which is similar to prior studies that have shown a period of scurrying and activity prior to settling in for illness behavior in this type of paradigm (Romanovsky et al., 1996, 2005). The only difference between saline and poly I:C groups was evident two hours after injection, at which point the poly I:C-injected animals were moving around whereas the saline rats have settled.

3.1.3. General summary.

Time of fever onset (2.5 hours after challenge) as well as other fever metrics did not differ in males and females. The only sex difference noted was that females experienced a transient elevation in temperature immediately after injection, while males did not. All poly I:C-injected animals showed more locomotor activity after injection, as expected, which abated at around the time of fever onset and remained stable across testing, as did the fever itself.

3.2. Experiment 2a (Figure 2)

Figure 2. The 4 mg/kg poly I:C (or saline vehicle) febrile time course for Experiment 2a VEH or AIE males with an abdominal probe.

Data are shown as (A) adjusted to baseline (final two timepoints of the “acclimation” time period) with mean ± SEM shown for each time point. Other metrics such as (B) peak fever in degrees Celsius, (C) peak degrees Celsius change at the highest point of fever as compared to baseline, (D) latency to reach the peak fever in hours, and (E) total area under the curve for the duration of fever are shown with mean and quartile ranges marked via violin plot. Activity data (F) is shown in 1-hr bins. Statistical markers are decoded in the figure, with all p < 0.05.

3.2.1. Fever time course in males.

The omnibus ANOVA of core body temperature (Figure 2A) revealed main effects of challenge (F_{1, 31} = 18.45, p < 0.001, ηp² = 0.37) and time (F_{21, 651} = 33.57, p < 0.0001, ηp² = 0.52). Significant interactions were observed for adolescent exposure × adult challenge (F_{1, 31} = 8.91, p < 0.01, ηp² = 0.22), with Tukey’s post hoc indicating a higher temperature in poly I:C-injected males with a history of AIE than other groups (p < 0.05), time × adult challenge (F_{21, 651} = 25.96, p < 0.0001, ηp² = 0.46), and a three-way interaction between adolescent exposure, adult challenge, and time (F_{21, 651} = 1.97, p < 0.01, ηp² = 0.06).

In the post-injection period, the ANOVA revealed a main effect of time (F_{4, 124} = 21.70, p < 0.00001, ηp² = 0.41), as well as interactions of adolescent exposure × adult challenge (F1, 31 = 7.08, p < 0.05, ηp² = 0.19), and adolescent exposure × adult challenge × time (F_{4, 124} = 2.99, p < 0.05, η² = 0.09). Post hoc tests revealed minor differences that did not indicate a coherent pattern: the VEH-saline group was elevated for an hour after injection (p < 0.01), the AIE-poly I:C group was elevated for half an hour (p < 0.01), and the rest did not show significant changes associated with injection. The only group difference was evident at 1.5 hours post- injection, with the VEH-poly I:C group having lower body temperature than VEH-saline controls (p < 0.05).

The ANOVA of core body temperature in the poly I:C groups revealed main effects of adolescent exposure (F_{1, 15} = 6.27, p < 0.05, ηp² = 0.29), with higher levels observed in the AIE group, and time (F_{21, 315} = 35.29, p < 0.0001, ηp² = 0.70). During the rising fever phase, core body temperature differed as a function of adolescent exposure (F_{1, 15} = 6.72, p < .05, ηp² = 0.31) and time (F_{7, 105} = 50.66, p < 0.0001, ηp² = 0.77). There was also an interaction of these two variables (F_{7, 105} = 2.88, p < 0.01, ηp² = 0.16). The interaction indicated that in the VEH group, fever became significant starting at the 3-hour time point (p < 0.05 for first time point, then p < 0.001), whereas the AIE group showed significant increases in core body temperature starting at 2.5 hours post-injection (p < 0.001). Starting at 3 hours, the AIE group showed significant elevation above VEH (p < 0.05, < 0.05, < 0.01 at each consecutive point). During the maintenance phase, there was a main effect of time (F_{8, 120} = 42.51, p < 0.0001, ηp² = 0.74), with body temperature remaining elevated at all time points (p < 0.0001).

3.2.2. Other fever metrics.

Peak fever, peak degree change, or latency did not differ as a function of adolescent history (Figure 2B, C, D). There was a significant effect in AUC (Figure 2E), with greater AUC seen in the AIE group (t₁₅ = −2.28, p < 0.05).

3.2.3. Activity data.

There was a main effect of time (F_{10, 310} = 30.26, p < 0.0001, ηp² = 0.49) and an adult challenge × time interaction (F_{10, 310} = 2.36, p < 0.05, ηp² = 0.07). Both saline and poly I:C groups were the most active during the first half an hour of testing (Figure 2F). After this, saline groups showed an elevation of activity in the first hour after injection (p < 0.01) above all other time points but then remained stable the rest of testing, and the same pattern was evident for the poly I:C groups. There were no differences between saline and poly I:C group activity at any time point.

3.2.4. General summary.

Adult males with a history of AIE showed a higher fever response to poly I:C than their control counterparts exposed to water during adolescence (VEH). Some (overall fever) but not all (latency, peak) measures indicated that not only was fever higher, but also started earlier in the AIE group relative to VEH controls. Overall, the fever response was similar to that evident in ethanol-naïve animals. Activity was affected by the relocation for testing and the injection itself, but showed no effects of illness or circadian rhythm in this experiment.

3.3. Experiment 2b (Figure 3)

Figure 3. The 4 mg/kg poly I:C (or saline vehicle) febrile time course for Experiment 2b VEH or AIE males with a subcutaneous probe.

Data are shown as (A) adjusted to baseline (final two timepoints of the “acclimation” time period) with mean ± SEM shown for each time point. Other metrics such as (B) peak fever in degrees Celsius, (C) peak degrees Celsius change at the highest point of fever as compared to baseline, (D) latency to reach the peak fever in hours, and (E) total area under the curve for the duration of fever are shown with mean and quartile ranges marked via violin plot. Activity data (F) is shown in 1-hr bins. Statistical markers are decoded in the figure, with all p < 0.05.

3.3.1. Fever time course.

The omnibus ANOVA revealed main effects of adult challenge (F_{1, 35} = 13.24, p < 0.001, ηp² = 0.27) and time (F_{21, 735} = 33.62, p < 0.0001, ηp² = 0.49), as well as a challenge × time interaction (F_{21, 735} = 11.68, p < 0.0001, ηp² = 0.25).

The ANOVA of core body temperature changes as an initial response to injection (Figure 3A) revealed a main effect of time (F_{4, 140} = 50.32, p < 0.0001, ηp² = 0.59), with all groups showing significant elevation in body temperature during the first two time points/one hour after injection (p < 0.0001).

The ANOVA of core body temperature in the poly I:C groups revealed a main effect of time (F_{21, 399} = 16.35, p < 0.0001, ηp² = 0.46). In the rising phase of fever, there was a main effect of time (F_{7, 133} = 15.72, p < 0.00001, ηp² = 0.45), with both groups showing fever by 2 hours post-injection (p < 0.05 for 2 and 2.5hr) and every time point after (< 0.0001). During the fever maintenance phase, a main effect of time (F_{8, 152} = 25.74, p < 0.0001, η² = 0.58 was noted, with significant elevation of core body temperature evident at all time points (p < 0.0001).

3.3.2. Other fever metrics.

The AIE males reached significantly higher peak fever than VEH controls (t₁₉ = −2.34, p < 0.05) in raw degrees C (Figure 3B), though there was no difference when adjusted to change from baseline (Figure 3C) and no difference in latency (Figure 3D). AIE males achieved higher AUC for fever than VEH males (Figure 3E; t₁₉ = −2.11, p < 0.05).

3.3.3. Activity data.

In activity (Figure 3F), there was a main effect of time (F_{10, 350} = 17.07, p < 0.0001, η² = 0.33) and an adult challenge × time interaction (F_{10, 350} = 3.92, p < 0.0001, ηp² = 0.10). The saline groups showed higher than baseline activity during the first hour of testing (p < 0.0001) and during the first hour post-injection (p < 0.05). The poly I:C groups also showed highest activity at the start of testing for a 1-hour period (p < 0.0001), but after injection remained active for far longer than the saline group – up through 3 hours after injection (p < 0.0001, < 0.001, < 0.05 for each consecutive hour). The two challenge conditions differed from one another at 2 hours post-injection (p < 0.05).

3.3.4. General summary.

Though the effects of AIE did not quite reach significance when analyzing the entire time course, significantly higher peak fever in raw data and higher AUC achieved by AIE exposed males indicate that the overall pattern replicated the data from Experiment 2a. Likewise as seen in one of the previous experiments, there was more activity in the poly I:C injected animals that settled around the time of fever onset.

3.4. Experiment 3a (Figure 4)

Figure 4. The 4 mg/kg poly I:C (or saline vehicle) febrile time course for Experiment 2b VEH or AIE females with an abdominal probe.

Data are shown as (A) adjusted to baseline (final two timepoints of the “acclimation” time period) with mean ± SEM shown for each time point. Other metrics such as (B) peak fever in degrees Celsius, (C) peak degrees Celsius change at the highest point of fever as compared to baseline, (D) latency to reach the peak fever in hours, and (E) total area under the curve for the duration of fever are shown with mean and quartile ranges marked via violin plot. Activity data (F) is shown in 1-hr bins. Statistical markers are decoded in the figure, with all p < 0.05.

3.4.1. Fever time course in females.

The omnibus ANOVA of core body temperature changes (Figure 4A) showed main effects of adolescent exposure (F_{1, 24} = 4.93, p < 0.05, ηp² = 0.17), with lower levels seen in AIE females than VEH; adult challenge (F_{1, 24} = 23.01, p < 0.0001, ηp² = 0.49), with higher levels seen in females injected with poly I:C, and time (F_{21, 504} = 14.72, p = 0.0001, ηp² = 0.38). Interactions were noted between adolescent exposure and time (F_{21, 504} = 1.72, p < 0.05, ηp² = 0.07) as well as adult challenge and time (F_{21, 504} = 14.49, p < 0.0001, ηp² = 0.38).

Immediately post injection, core body temperature changed as a function of time (F_{4, 96} = 14.90, p < 0.00001, ηp² = 0.55), with all groups showing significant elevation during all time point (p < 0.0001 for first two time points, p < 0.05 for the last).

The ANOVA of core body temperature in the poly I:C groups revealed a main overall effect of time (F_{21, 252} = 17.51, p < 0.0001, ηp² = 0.59). During the fever rising phase, there was only a main effect of time, (F_{7, 84} = 22.03, p < 0.00001, ηp² = 0.65), with elevation observed for both adolescent exposure conditions starting at 2.5 hours post-injection (p < 0.001). During the maintenance phase, there was a main effect of time as well (F_{8, 96} = 27.90, p < 0.0001, ηp² = 0.70).

3.4.2. Other fever metrics.

There were no effects of adolescent exposure on peak, peak degree change, latency, or AUC (Figure 4B, C, D, E, respectively).

3.4.3. Activity data.

There were main effects of time (F_{10, 240} = 11.47, p = 0.00001, ηp² = 0.32) and an adult challenge × time interaction (F_{10, 240} = 2.74, p < 0.01, ηp² = 0.10) in the activity data (Figure 4F). The poly I:C groups showed the pattern of elevated activity during the first two hours of baseline and the first two hours after injection (all p < 0.05). The saline groups were active during the first hour of testing, one hour after injection, and also sporadically at 3 and 7 hours after injection (all p < 0.05). At no point did saline vs poly I:C differ.

3.4.4. General summary.

Overall, AIE exposure did not significantly affected the poly I:C response in adult females. Peak appeared similarly to males in Experiment 2. The visual effect of slightly lower levels in the AIE group can be explained by one female that may have been a non-responder – however, the statistics for the time course as well as tertiary measures confirm no lasting AIE effects in females.

3.5. Experiment 3b (Figure 5)

Figure 5. The 4 mg/kg poly I:C (or saline vehicle) febrile time course for Experiment 2b VEH or AIE females with subcutaneous probe.

Data are shown as (A) adjusted to baseline (final two timepoints of the “acclimation” time period) with mean ± SEM shown for each time point. Other metrics such as (B) peak fever in degrees Celsius, (C) peak degrees Celsius change at the highest point of fever as compared to baseline, (D) latency to reach the peak fever in hours, and (E) total area under the curve for the duration of fever are shown with mean and quartile ranges marked via violin plot. Activity data (F) is shown in 1-hr bins. Statistical markers are decoded in the figure, with all p < 0.05.

3.5.1. Fever time course (Figure 5a).

The omnibus core body temperature ANOVA revealed main effects of time (F_{21, 651} = 24.82, p < 0.0001, ηp² = 0.44), and adult challenge (F_{1, 31} = 12.22, p < 0.01, ηp² = 0.28), as well as an adult challenge × time interaction (F_{21, 651} = 12.57, p < 0.0001, ηp² = 0.29).

Immediately following injection, core body temperature differed as a function of time (F_{4, 124} = 46.31, p < 0.0001, ηp² = 0.60), with all groups showing elevation during the first half-hour after injection (p < 0.0001). There was also a main effect of time during the rising phase (F_{7, 119} = 14.21, p < 0.00001, ηp² = 0.46), with fever seen at 2 hours post-injection (p < 0.05), 2.5 hours (p < 0.01), and every time point onward (p < 0.0001). During the maintenance phase of fever, again only an effect of time was detected (F_{8, 136} = 11.08, p < 0.00001, η² = 0.39), with elevation seen at all time points (p < 0.0001).

3.5.2. Other fever metrics.

Peak, peak degree change, latency, or AUC did not differ as a function of adolescent exposure (Figure 5B, C, D, E, respectively).

3.5.3. Activity data.

In activity data (Figure 5F), there was a main effect of time (F_{10, 310} = 20.48, p < 0.0001, ηp² = 0.39) and a challenge × time interaction (F_{10, 310} = 5.09, p < 0.00001, ηp² = 0.14). As in prior experiments, the saline groups showed higher activity during the first hour after testing began (p < 0.0001) and first hour post-injection (p < 0.001) relative to baseline. Poly I:C groups showed increased activity only in the first hour of testing (p < 0.001) but not after injection. Adult challenge conditions differed during the first hour of testing, with lower levels of activity seen in poly I:C injected females (p < 0.05).

3.5.4. General summary.

AIE did not affect any of the measures under investigation, largely confirming findings of Experiment 3a and continuing to stand in contrast to findings in AIE males in Experiment 2.

3.6. Experiment 4 (Table 1, Figure 6)

Table 1.

Gene expression as percent of control (VEH-saline group) and Magpix protein (pg/mL) findings for Experiment 4 are reported here.

Region	Target	VEH-saline	AIE-saline	VEH-poly I:C	AIE-poly I:C
RT-PCR gene expression (% of control: VEH-saline group)
Spleen	IL-6	101.1 ± 5.6	114.4 ± 8.1	218.7 ± 49.6 *	201.5 ± 34.2 *
	TNFα	105.2 ± 13.0	83.5 ± 8.7	108.3 ± 17.8	132.1 ± 18.0
Liver	IL-6	103.3 ± 10.8	351.4 ± 137.9	54,443.3 ± 23,566.2 *	18,090.1 ± 7,615.7 *
	TNFα	160.0 ± 124.9	2,729.9 ± 2,517.0	1,661.9 ± 669.0	1,483.8 ± 385.3
NTS	IL-6	103.0 ± 9.4	105.9 ± 7.7	218.0 ± 49.9 *	406.8 ± 132.8 *
	TNFα	101.7 ± 7.7	131.4 ± 17.2	185.8 ± 62.0 *	324.6 ± 74.1 *
PVN	IL-6	150.0 ± 12.1	100.0 ± 13.1	189.1 ± 23.5 **	209.7 ± 27.8 **
	TNFα	104.7 ± 11.3	116.8 ± 8.7	236.6 ± 49.7 *	319.7 ± 71.2 *
mPOA	IL-6	101.1 ± 15.0	103.1 ± 14.3	204.9 ± 56.4	239.8 ± 111.4
	TNFα	113.1 ± 52.9	102.8 ± 8.4	322.3 ± 119.7 *	421.5 ± 182.4 *
Magpix protein multiplex (pg/mL)
Plasma	IL-1α	204.8 ± 35.2	170.0 ± 38.7	202.5 ± 41.5	239.2 ± 31.3
	IL-1β	54.8 ±15.2	60.2 ± 18.8	80.1 ± 15.3	84.9 ± 19.4
	IL-2	1431.0 ± 433.3	1743.7 ± 162.2	1278.8 ± 308.0	1385.4 ± 290.4
	IL-4	407.3 ± 86.6	408.4 ± 97.9	446.9 ± 87.7	525.3 ± 78.9
	IL-5	568.2 ± 81.2	570.1 ± 79.6	590.9 ± 75.5	652.1 ± 57.5
	IL-6	366.0 ± 119.0	464.7 ± 143.8	369.6 ± 120.9	450.9 ± 125.4
	IL-7	64.6 ± 15.7	61.8 ± 19.2	81.5 ± 15.1	82.0 ± 15.5
	IL-10	224.7 ± 47.0	222.0 ± 53.5	284.4 ± 38.9	284.2 ± 43.6
	IL-12p70	519.0 ± 154.0	653.0 ± 1637.	538.9 ± 130.3	549.2 ± 145.8
	IL-13	289.6 ± 87.6	355.0 ± 81.6	275.0 ± 90.6	355.9 ± 81.2
	IL-17	42.4 ± 11.8	60.9 ± 14.4	55.1 ± 11.9	52.3 ± 11.6
	IL-18	817.5 ± 132.6	1167.6 ± 169.9	1166.3 ± 221.4	1040.0 ± 214.9
	G-CSF	19.5 ± 5.1	17.3 ± 5.0	22.5 ± 5.0	22.8 ± 4.6
	GM-CSF	75.8 ± 19.9	92.2 ± 21.9	74.5 ± 18.9	84.3 ± 18.9
	GRO/KC (CXCL1)	81.1 ± 22.8	66.4 ± 19.8	225.5 ± 87.7 *	259.6 ±106.1 *
	IFN-γ	512.1 ± 114.6	519.7 ± 144.5	450.3 ± 115.4	624.6 ± 99.6
	M-CSF	31.6 ± 10.1	43.6 ± 12.6	41.5 ± 12.4	36.1 ± 10.2
	MCP-1 (CCL2)	548.1 ± 58.2	509.1 ± 41.7	961.6 ± 194.8 **	1030.6 ± 166.1 **
	MIP-1α (CCL3)	21.1 ± 4.9	21.1 ± 5.7	164.4 ± 52.2 ***	154.4 ± 42.7 ***
	MIP-3α	22.2 ± 3.3	20.0 ± 3.1	116.9 ± 43.3 **	151.5 ± 48.8 **
	RANTES	220.6 ± 40.3	217.0 ± 50.9	743.8 ± 244.3 **	778.3 ± 181.2 **
	TNFα	1609.3 ± 445.1	2228.5 ± 220.5	1799.4 ± 360.4	1604.4 ± 348.5
	VEGF	93.2 ± N/A	55.9 ± 22.3	43.0 ± 15.2	55.5 ± 19.4

Here, males with VEH or AIE adolescent history were challenged with a saline or 4 mg/kg poly I:C injection 5 hr prior to tissue collection.

Asterisks denote main effects of challenge (* p < 0.05, ** < 0.01, *** < 0.001); no effects of adolescent history or interactions were observed.

All effects are bolded for additional emphasis/visibility. For a full list of target acronym meanings, please see general methods, 2.8 Magpix subsection.

Figure 6. Gene expression following 0 (saline) or 4 mg/kg poly I:C challenge 5 hr after injection in Experiment 4 VEH or AIE males.

Toll-like receptor 3 (TLR3), nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor, alpha (IkBα), and Interleukin(IL)-1β are shown for spleen (A-C), liver (D-F), nucleus of the solitary tract (NTS; G-I), paraventricular nucleus of the hypothalamus (PVN; J-L), and the medial preoptic area (mPOA; M-O). Statistical markers are decoded in the figure, with all p < 0.05.

3.6.1. Spleen.

There was a main effect of adult challenge in multiple targets, with poly I:C increasing expression of TLR3 (F_{1, 26} = 12.66, p < 0.01, ηp² = 0.33; Figure 6A), IκBα (F_{1, 26} = 4.24, p < 0.05, ηp² = 0.14; Figure 6B), and IL-1β (F_{1, 26} = 17.44, p < 0.001, ηp2 = 0.40; Figure 6C). IL-6 expression was significantly elevated by poly I:C (main effect of adult challenge: F_{1, 26} = 11.75, p < 0.01, ηp² = 0.31; Table 1).

3.6.2. Liver.

There was a main effect of adult challenge on TLR3 expression (F_{1, 28} = 6.98, p < 0.05, ηp² = 0.20; Figure 6D), with higher expression in the poly I:C groups. IκBα (Figure 6E) showed no effects, and IL-1β (Figure 6F) showed an effect of adult challenge (F_{1, 27} = 8.13, p < 0.01, ηp² = 0.20). There were no effects on TNFα, and for IL-6, there was a main effect of challenge (F_{1, 26} = 7.34, p < 0.05, ηp² = 0.22; Table 1).

3.6.3. Nucleus of the Solitary Tract (NTS).

The ANOVA of TLR3 expression revealed a main effect of adult challenge (F_{1, 28} = 12.87, p < 0.01, ηp² = 0.31, Figure 6G), with poly I:C increasing TLR3 expression levels, as well as an adolescent exposure × adult challenge interaction (F_{1, 28} = 6.04, p < 0.05, ηp² = 0.18), with the AIE-exposed males challenged with poly I:C showing higher levels than their counterparts challenged with saline (p < 0.001) as well as VEH-exposed males challenged with poly I:C (p < 0.05). Likewise, for IκBα expression, there was both a main effect of adult challenge (F_{1, 28} = 11.15, p < 0.01, ηp² = 0.28; Figure 6HB) as well as an adolescent exposure × adult challenge interaction (F_{1, 28} = 4.55, p <0.05, ηp² = 0.14), with the AIE-poly I:C group showing higher levels than both the AIE-saline (p < 0.001) as well as the VEH-poly I:C groups (p < 0.05). There was a main effect of challenge in IL-1β (F_{1, 27} = 8.84, p < 0.01, ηp² = 0.25; Figure 6I), with higher levels overall seen after poly I:C challenge, with this effect driven by the AIE group as evidenced by adolescent exposure × adult challenge interaction (F_{1, 27} = 4.92, p < 0.05, ηp² = 0.15). Specifically, there were no differences in IL-1β expression between AIE- and VEH-exposed males challenged with saline, with poly I:C significantly increasing IL-1β only in AIE-exposed males relative to AIE-exposed animals challenged with saline (p < 0.001) and VEH-exposed males challenged with poly I:C (p < 0.05). There was a main effect of adult challenge on TNFα, with higher levels seen in poly I:C animals (F_{1, 27} =7.39, p < 0.05, ηp² = 0.21), and on IL-6 with a similar pattern (F_{1, 28} = 8.52, p < 0.01, ηp² = 0.23; Table 1).

3.6.4. Paraventricular Nucleus of the Hypothalamus (PVN).

There was a main effect of adult challenge on TLR3 expression, with higher levels evident in poly I:C-challenged animals relative to their saline-challenged counterparts (F_{1, 28} = 18.22, p < .001, ηp² = 0.39; Figure 6J). IκBα expression differed as a function of adult challenge (F_{1, 28} = 14.13, p < 0.001, ηp² = 0.34; Figure 6K), with poly I:C significantly increasing IκBα expression. There was a main effect of adult challenge on IL-1β expression, with higher levels seen in poly I:C challenged animals relative to saline-challenged counterparts (F_{1, 27} = 9.06, p < 0.01, ηp² = 0.25; Figure 1H). TNFα as well as IL-6 expression also differed as a function of adult challenge (F_{1, 28} = 14.49, p < 0.001, ηp² = 0.34 and F_{1, 28} =22.88, p < 0.0001, ηp² = 0.45, respectively; Table 1), with poly I:C significantly increasing TNFα and IL-6.

3.6.5. Medial Preoptic Area (mPOA).

Due to technical error, half of the samples for mPOA were lost at the extraction stage and unavailable for PCR, so these data are shown for descriptive purposes with no analyses (Figure 6 M, N, O).

3.6.6. Plasma protein.

Of the 23 targets examined, no effects were observed for all but 5 chemokines (all findings reported in Table 1). Main effects of adult challenge with elevation in the poly I:C groups were evident for the following chemokines: GRO/KC F_{1, 27} = 5.92, p < 0.05, ηp² = 0.18, MCP-1 F_{1, 28} = 12.37, p < 0.01, ηp² = 0.31, MIP-1α F_{1, 27} = 18.15, p < 0.001, ηp² = 0.40, MIP-3α F_{1, 28} = 11.97, p < 0.01, ηp² = 0.30, and RANTES F_{1, 28} = 12.16, p < 0.01, ηp² = 0.30.

3.7. Experiment 5 (Figure 7).

Figure 7. Gene expression following 0 (saline) or 4 mg/kg poly I:C challenge 5 hr after injection in Experiment 5 VEH or AIE females.

Toll-like receptor 3 (TLR3), nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor, alpha (IkBα), and Interleukin(IL)-1β are shown for spleen (A-C), liver (D-F), nucleus of the solitary tract (NTS; G-I), paraventricular nucleus of the hypothalamus (PVN; J-L), and the medial preoptic area (mPOA; M-O). Statistical markers are decoded in the figure, with all p < 0.05.

3.7.1. Spleen.

Main effects of adult challenge were seen in TLR3 (F_{1, 28} = 7.95, p < 0.01, ηp² = 0.22; Figure 7A), IκBα (F_{1, 28} = 27.18, p < 0.0001, ηp² = 0.49; Figure 7B), and IL-1β expression (F_{1, 28} = 27.55, p < 0.0001, ηp² = 0.49, Figure 7C), and an interaction (F_{1, 28} =13.23, p < 0.01, ηp² = 0.32) evident for IL-1β. Females exposed to VEH during adolescence showed a significant IL-1β increase following poly I:C challenge relative to their counterparts challenged with saline as well as AIE exposed females challenged with poly I:C (p < 0.0001), with no effect of poly I:C evident in AIE exposed females. TNFα expression was not affected by either adolescent exposure or adult challenge (Table 2). IL-6 expression differed as a function adult challenge (F_{1, 28} = 37.16, p < 0.00001, ηp² = 0.57; Table 2).

Table 2.

Gene expression as percent of control (VEH-saline group) and Magpix protein (pg/mL) findings for Experiment 5 are reported here.

Region	Target	VEH-saline	AIE-saline	VEH-poly I:C	AIE-poly I:C
RT-PCR gene expression (% of control: VEH-saline group)
Spleen	IL-6	101.2 ± 6.1	132.5 ± 15.4	438.0 ± 74.3 ****	301.2 ± 32.9 ****
	TNFα	106.6 ± 14.3	115.2 ± 12.2	132.3 ± 13.7	104.3 ± 12.5
Liver	IL-6	ND	105.3 ± 13.5 A	14,974.1 ± 2,683.7 B	10,623.1 ± 2,164.5 B
	TNFα	106.7 ± 15.8	88.8 ± 9.3	5,245.9 ± 806.7 ****	3,303.6n ± 417.7 ****
NTS	IL-6	110.9 ± 23.9	114.9 ± 6.5	449.3 ± 98.1 ***	456.9 ± 85.2 ***
	TNFα	71.5 ± 3.1	59.3 ± 16.5	420.0 ± 95.9 ***	386.6 ± 106.9 ***
PVN	IL-6	103.2 ± 9.4	105.6 ± 13.8	197.0 ± 22.1 ***	207.2 ± 47.2 ***
	TNFα	117.4 ± 18.1	119.0 ± 17.7	823.8 ± 108.0 ***	634.5 ± 116.8 ***
mPOA	IL-6	100.1 ± 1.8	96.5 ± 9.2	205.9 ± 22.6 ****	202.7 ± 25.4 ****
	TNFα	109.9 ± 20.5	105.5 ± 9.6	640.9 ± 95.2 ****	569.3 ± 87.7 ****
Magpix protein multiplex (pg/mL)
Plasma	IL-1α	227.4 ± 23.7	213.1 ± 26.2	213.2 ± 23.8	230.7 ± 22.8
	IL-1β	72.4 ± 10.8	62.9 ± 12.1	92.1 ± 11.5	82.1 ± 9.8
	IL-2	1,238.8 ± 213.9	1,136.3 ± 244.5	1,099.9 ± 228.7	1,152.4 ± 184.4
	IL-4	523.7 ± 54.5	475.2 ± 66.7	481.8 ± 56.8	499.6 ± 43.2
	IL-5	475.2 ± 46.5	630.3 ± 55.5	604.8 ± 47.9	646.7 ± 37.6
	IL-6	362.6 ± 57.4	414.3 ± 136.0	518.5 ± 114.3	348.5 ± 103.1
	IL-7	89.3 ± 11.2	80.2 ± 11.7	76.1 ± 10.7	81.4 ± 9.0
	IL-10	278.9 ± 29.2	268.5 ± 39.9	275.0 ± 28.8	277.0 ± 22.0
	IL-12p70	545.8 ± 103.2	530.6 ± 117.3	452.2 ± 107.0	488.6 ± 105.6
	IL-13	284.1 ± 36.7	274.7 ± 82.3	198.1 ± 57.7	180.6 ± 41.6
	IL-17	55.8 ± 9.2	51.9 ±11.3	53.0 ± 7.8	51.1 ± 6.9
	IL-18	1,175.6 ± 152.3	1,792.1 ± 464.9	1,171.0 ± 489.0	950.4 ± 161.4
	G-CSF	20.4 ± 3.1	16.9 ± 3.6	17.2 ± 3.4	16.5 ± 2.3
	GM-CSF	85.2 ± 14.9	70.2 ± 16.9	65.9 ± 14.4	72.1 ± 13.0
	GRO/KC (CXCL1)	88.0 ± 12.0	90.6 ± 14.7	1,442.3 ± 559.9 **	710.9 ± 145.0 **
	IFN-γ	532.7 ± 53.9	453.3 ± 117.7	373.2 ± 86.6	384.8 ± 69.6
	M-CSF	34.5 ± 6.6	24.9 ± 9.1	26.8 ± 6.2	24.8 ± 6.3
	MCP-1 (CCL2)	596.4 ± 51.4	617.4 ± 54.0	1747.4 ± 225.1 ****	1,433.6 ± 83.9 ****
	MIP-1α (CCL3)	27.8 ± 3.4	26.9 ± 5.2	363.8 ± 60.6 ****	311.8 ± 31.8 ****
	MIP-3α	22.4 ± 2.2	19.6 ± 2.8	201.6 ± 35.3 ****	184.3 ± 19.1 ****
	RANTES	223.9 ± 26.2	262.2 ± 46.5	1,600.0 ± 232.3 ****	1,248.4 ± 70.2 ****
	TNFα	1,484.3 ± 281.6	1,245.8 ± 258.8	1,448.7 ± 240.8	1,318.8 ± 241.0
	VEGF	38.4 ± 7.9	42.6 ± 18.5	29.0 ± 7.1	35.8 ± 19.4

Here, females with VEH or AIE adolescent history were challenged with a saline or 4 mg/kg poly I:C injection 5 hr prior to tissue collection.

Asterisks denote main effects of challenge (* p < 0.05, ** < 0.01, *** < 0.001, **** < 0.0001); a system of letters is used to denote group differences in the case of an interaction, wherein groups that share a letter do not differ, whereas groups that have different letters have statistically significant disparity (p < 0.05); no effects of adolescent history were observed.

All effects are bolded for additional emphasis/visibility. For a full list of target acronym meanings, please see general methods, 2.8 Magpix subsection. Analytes that were not detected are labeled “ND.”

3.7.2. Liver.

Expression of the following genes differed as a function of adult challenge: TLR3 (F_{1, 26} = 120.84, p < 0.00001, ηp² = 0.82, Figure 7D), IκBα (F_{1, 26} = 80.68, p < 0.00001, ηp² = 0.76; Figure 7E), IL-1β (F_{1, 26} = 96.19, p < 0.00001, ηp² = 0.79; Figure 4F), TNFα (F_{1, 25} = 71.98, p < 0.00001, ηp² = 0.74; Table 2), with poly I;C significantly elevated expression of TLR3, IκBα, IL-1β, and TNFα in the liver of female rats. IL-6 in the liver (Table 2) showed an unusual pattern not observed elsewhere or in other genes – females exposed to VEH during adolescence and challenged with saline in adulthood showed no detectable IL-6, however, AIE -exposed females challenged with saline showed IL-6 expression and were therefore used as a control group for adjusting relative expression of poly I:C injected groups. In other words, a history of AIE was sufficient to raise the baseline IL-6 expression in the liver to detectable levels when assessed after a prolonged period of ethanol abstinence. Since there were no data for VEH saline group, IL-6 was analyzed using a one-way ANOVA (Group: AIE-saline, VEH-poly IC, AIE-poly I:C), with a main effect observed (F_{2, 19} = 13.57, p < 0.001, ηp² = 0.59). Post hoc analysis revealed that poly I:C significantly increased IL-6 expression relative to the AIE-saline group, at a higher significance and mean % change level for VEH-exposed (p < 0.0001) than AIE-exposed females (p < 0.01) but did not differ from one another.

3.7.3. Nucleus of the Solitary Tract (NTS).

There were main effects of challenge on the following genes, with higher expression seen after poly I:C challenge regardless of adolescent exposure history: TLR3 expression (F_{1, 22} = 7.05, p < 0.05, ηp² = 0.24; Figure 7G), IκBα (F_{1, 20} = 34.86, p < 0.0001, ηp² = 0.64; Figure 7H), IL-1β (F_{1, 20} = 8.70, p < 0.01, ηp² = 0.30; Figure 7I), TNFα (F_{1, 20} = 18.31, p < 0.001, ηp² = 0.48; Table 2), and IL-6 (F_{1, 20} = 21.57, p < 0.001, ηp² = 0.52; Table 2).

3.7.4. Paraventricular Nucleus of the Hypothalamus (PVN).

TLR3 expression (Figure 7J), differed as a function of adult challenge (F_{1, 26} = 40.21, p < .00001 ηp² = 0.61), with higher expression in poly I:C rats, and adolescent exposure (F_{1, 26} = 5.86, p < 0.001, ηp² = 0.18), with overall lower expression in AIE animals. There were main effects of challenge in IκBα (F_{1, 25} = 65.90, p < 0.00001, ηp² = 0.72; Figure 7K), IL-1β (F_{1, 25} = 29.91, p < 0.0001, ηp² = 0.54; 7L), TNFα F₁, ₂₆ = 65.28, p < 0.00001, ηp² = 0.72, Table 2) and IL-6 expression (F_{1, 26} = 16.70, p < 0.001, ηp² = 0.39; Table 2).

3.7.5. Medial Preoptic Area (mPOA).

Adult challenge with poly I:C increased expression of the following: TLR3 (F_{1, 26} = 31.02, p < 0.0001, ηp² = 0.54; Figure 7M), IκBα (F_{1, 26} = 95.82, p < 0.00001, ηp² = 0.79; Figure 7N), IL-1β (F_{1, 26} = 26.27, p < 0.0001, ηp² = 0.51; Figure 7O), TNFa (F_{1, 25} = 52.44, p < 0.00001, ηp² = 0.68; Table 2), and IL-6 (F_{1, 24} = 31.57, p < 0.0001, ηp² = 0.57; Table 2).

3.7.6. Plasma protein.

Of the 23 targets examined, no effects were observed for all but 5 chemokines (all findings reported in Table 2), which were the same ones as affected in males. The overall range was similar between the two studies as well, with only GRO/KC showing higher overall levels in females (though a direct comparison was not made statistically, given that these studies were performed separately). Poly I:C significantly elevated expression of GRO/KC (F_{1, 26} = 13.18, p < 0.01, ηp² = 0.34), MCP-1 (F_{1, 26} = 67.24, p < 0.0001, ηp² = 0.72), MIP-1α (F_{1, 26} = 87.96, p < 0.0001, ηp² = 0.77), MIP-3α (F_{1, 26} = 78.09, p < 0.0001, ηp² = 0.75), and RANTES (F_{1, 26} = 101.4, p < 0.0001, ηp² = 0.80).

4. Discussion

The present series of experiments assessed effects of adolescent ethanol exposure of males and females on adult responsiveness to the viral mimetic and TLR3 agonist, poly I:C. The most striking novel finding of these experiments is the persistent and replicable sensitization of the fever response to poly I:C in males exposed to ethanol during adolescence relative to water-exposed controls. This effect of AIE was evident in males when probes were implanted in the abdominal cavity (Experiment 2a), with AIE-exposed animals showing an earlier onset of fever as well as a higher total fever achieved as measured via AUC. In Experiment 2b, probes were implanted subcutaneously at the nape of the neck in the mid-subscapular region, above an area with a concentration of brown adipose tissue (intrascapular BAT) that has dense sympathetic innervation, contributes to thermogenesis at times of fever, and is also known to be disrupted by chronic ethanol exposure (Blaner et al., 2017; Cruz et al., 2020). This procedure differed from the abdominal surgery in terms of location (subcutaneous temperature vs. internal core temperature) as well as in the likely severity of inflammatory consequences of the implantation procedure itself. Yet, while the absolute body temperatures differed between these locations, and the intrascapular subcutaneous probe displayed more range and sensitivity to the handling involved in the testing day injections, Experiment 2b replicated the main findings of Experiment 2a, showing AIE males to have a higher peak fever as well as higher total accumulated fever (AUC). Importantly, sensitization of the fever response to poly I:C was evident in males during adulthood after a period of abstinence from AIE. The raw data illustrating the differences in overall temperature and the parallels of fever dynamics can be found in supplemental figure 4.

It is interesting to note that in the animals with subcutaneous probes, the change in temperature in the first hour of testing after being relocated to the testing room exceeded the increase provoked by immune challenge, suggesting that this site of thermogenesis may be particularly sensitive to stress manipulation. Certainly, temperature measurements in the subcutaneous space are likely to be affected by changes in skin blood flow and autonomic reactivity in ways that abdominal temperature is not. However, subcutaneous temperature recordings are also not synonymous with surface/skin temperature, which is typically recorded with infrared cameras and does not show sufficient coherence with core body temperature for immunological purposes (Meyer et al., 2017). Importantly, our work demonstrates comparable immunological effects, and sensitivity to AIE, regardless of whether probes were located subcutaneous or in the abdominal core. Thus, the present findings provide an initial validation for a less invasive approach toward assessment of immunological fever.

In contrast to males, AIE exposure of females did not affect the fever response to poly I:C. This pattern of results suggests sex-specific sensitivity to long-term consequences of AIE, with only males becoming hypersensitive to poly I:C induced febrile response. This finding adds to a list of AIE-associated adverse consequences evident only in males, including, but not limited to, disruptions of blood-brain integrity (Vore et al., 2022), increases in hippocampal pro-inflammatory cytokine expression (Nwachukwu, Healey, et al., 2022; Nwachukwu, King, et al., 2022), and attenuations of stress- and LPS-induced cytokine expression in circulating lymphocytes (Vore et al., 2017).

Another important finding of the present study is that non-manipulated, naïve to ethanol adult male and female rats demonstrated similar fever response to poly I:C (Experiment 1). This result is in agreement with findings from the study of Posillico et al. (2021) that used mice and showed no sex differences in febrile response induced by centrally administered poly I:C. In contrast, almost all studies of fever induced by poly I:C in rats used only males (Fortier et al., 2004; Kamerman et al., 2011; Yamato et al., 2014), though one female-only study by Coelho and colleagues (2021) suggests that estrogen may lower fever, as ovariectomized female rats responded to 0.3 mg/kg poly I:C with a suppressed fever relative to sham-operated controls. Given that ovariectomized females are not equivalent to males, direct comparisons of male and female rats, as in Experiment 1, are needed for better understanding sex differences and/or similarities in responding to this viral mimetic using a wide dose range of poly I:C, with particular attention paid to the differences in fever dynamics seen in abdominal temperature measures versus those that may relate to brown adipose tissue that may impact assessment using probes that are inserted subcutaneously in the mid-subscapular region.

Using separate cohorts of males and females, Experiments 4 and 5 examined consequences of AIE exposure on poly I:C-induced expression of cytokines, chemokines, and growth factors in brain regions associated with thermoregulation as well as in peripheral organs instrumental to the acute phase of viral infections (blood and liver, but not spleen) at the time point that corresponded to fever peak observed across Experiments 1–3. While the first three experiments allowed for tracking fever progression under relatively undisturbed home cage conditions for a prolonged period, in Experiments 4 and 5, tissue collection occurred at a single time point (5 hours post-injection) selected to illuminate central control of the fever response. At this time point, no significant poly I:C induced increases in serum levels of the major cytokines (IL-1β, IL-6, and TNFα) that serve as first responders to an immune challenge were evident in both males and females. Therefore, it was likely too late to observe the very first indicators of the body’s response to challenge that would have occurred in the peripheral organs during the first few hours after injection. Peripheral sources of these cytokines include dendritic cells and macrophages in various organs, and the tissue specificity of the response is dynamic and regulated by exposure to the febrile temperature being evoked (Evans et al., 2015), with production of IL-6, a major contributor to pyogenesis, shifted strongly to the liver (Jiang et al., 1999; Ostberg et al., 2000). Indeed, we observed trends for increased IL-6 at 5 hours post-injection in AIE-exposed males in multiple tissue compartments except the liver and spleen. Though these effects were not statistically significant, it might suggest a shift away from liver production of IL-6 in AIE-exposed males. The exact mechanism via which IL-6 propagates fever from the periphery remain to be fully elucidated, though it has been shown that IL-6 receptors on brain endothelial cells contribute to this effect by activating the prostaglandin pathway (Eskilsson et al., 2014; Rummel et al., 2006 and 2011). It should be noted that these studies were done using LPS as a stimulus and even less is known about the effects of poly I:C on this pathway. A number of chemokines were also elevated in poly I:C-injected animals at this time point, suggestive of ongoing fever maintenance. For example, RANTES has been shown to promote febrile responses via both injection at the preoptic area as well as following LPS administration via PGE2-dependent mechanisms (Machado et al., 2007).

In males, peripheral expression of cytokines and chemokines was largely unaffected by AIE history. Further studies employing tissue collection at earlier time points following administration of poly I:C may be necessary to test whether sensitized peripheral responses might occur earlier during the fever induction phase or whether the periphery is not involved in fever sensitization evident in AIE males. Interestingly, though we did not see any fever sensitization in females, we did observe a relic of AIE history in female livers. While control females exposed to water during adolescence had no detectable IL-6 in the liver when challenged with saline, their AIE exposed counterparts had IL-6 levels that were easily detected with PCR, while females in both adolescent exposure conditions showed comparable levels of poly I:C-induced IL-6 expression. PCR is a sensitive enough detection tool that we are able to measure IL-6 reliably in the brain where it is found at miniscule concentrations, which makes this finding in the liver unusual. Though not the focus of the present experimental series, this finding may be of relevance given the higher susceptibility of women to alcoholic liver diseases and cirrhosis (Bizzaro et al., 2023; Kezer et al., 2021).

While the present studies suggest that peripheral immune mechanisms are likely not contributing to the sensitized fever response to poly I:C in adult males with a history of AIE, the NTS gene expression data suggest that this sensitization is centrally mediated. TLR3, IkBα, and IL-1β expression induced by poly I:C were all significantly higher in AIE males relative to their water-exposed counterparts (unlike TNFα and IL-6, perhaps indicating IL-1β specificity of this effect). These results agree with prior work revealing the importance of the NTS IL-1β in the inflammatory response, as Gordon (2000) discovered that NTS-lesioned rats demonstrated an impaired fever response to IL-1β injection, whereas others have found that IL-1β enhances synaptic transmission in the NTS through PGE₂ synthesis (Marty et al., 2008). Though only the NTS effect was significant, visual inspection of the data shows slightly higher levels of IL-1β in the AIE males in the PVN and mPOA as well, suggesting that there may be more general sensitization across more brain regions that are responsible for fever regulation than just the NTS. Additionally, given the proximity of the NTS to the area postrema, it is still possible that circulating cytokines resultant from a humoral mechanism of ethanol toxicity may be involved in the signaling cascade and contributing to the NTS response. The more detailed knowledge we have of LPS-induced fever, for example, indicates that humoral and neuronal mechanisms are responsible for coordinating the multi-phase fever response to endotoxin challenge (reviewed by Roth & Blatteis, 2014). The female central cytokine response to poly I:C was not affected by adolescent exposure to ethanol, with no differences evident between AIE exposed females and their control counterparts, which matches what was observed in fever. However, it should be noted that males and females have different trajectories of response to immune stimuli. For example, other experiments that have investigated cytokine gene expression following poly I:C have shown female mice to display a delayed response to challenge compared to males, with peak expression of pro-inflammatory cytokines occurring 24–48 hours post-injection for females and only 3–8 hours for males (Qin & Crews, 2012; Warden et al., 2019). Additionally, it should be noted that we did not cardiac-perfuse animals with saline. Rats were rapidly decapitated in order to keep a precise timeline and preclude stress effects affecting measures. Therefore, it should be noted that blood-borne cytokines could affect these outcomes (Parker et al., 2000), though given the lack of effects seen in the blood we collected at this time, we believe this influence is likely to be minimal.

There were some limitations in these studies which may guide future experiments. For instance, as we have previously validated BECs achieved in the AIE model and did not collect blood for ethanol concentration analysis in these experiments. However, historical data indicates that these were in the binge range across adolescence. Another point of note is that as in most fever studies reported in the literature, animals were housed and tested at room temperature (22±1°C). It has previously been found that rodents prefer a warmer climate (higher temperature than the 22–26°C used in most vivariums) and therefore might have metabolic and other mechanisms engaged to reach a zone of thermoneutrality that may skew findings (Gordon 2017; Hankenson et al., 2018; Maloney et al., 2015). Future studies should specifically delineate fever responses achieved during cold stress from those seen in a preferred thermoneutral environment. Additionally, while the overarching results of adolescent and adult manipulations were similar across abdominal and subcutaneous probe locations, it should be acknowledged that there were some differences in fever dynamics when probes were implanted subcutaneously. Future studies will need to consider the balance of compromise between invasive surgery versus the specific type of immunological response that needs to be tracked experimentally. Finally, though we begin to draw conclusions about sex differences in these studies, further work will be necessary to determine whether AIE in females truly has no effect in adulthood, as it has been shown that febrile responses differ as a function of estrus cycle and this was not examined here. It has previously been shown that sensitivity to IL-1β changes with ovarian hormone levels and estrus cycle (Mouihate and Pittman, 1998), as well as across the lifespan at times that are associated with hormone perturbations such as those seen during pregnancy (Ashdown et al., 2006). Some of our own previous work demonstrates changes in rat brain IL-1β across aging (Gano et al., 2017), in harmony with the data suggesting that neuroimmune changes across aging as well as sex would contribute to altered pyrogenic responses (Deeter et al., 1989). However, another recent report has shown that females did not exhibit more variability in temperature as a result of ovarian cycle than males (Smarr and Kriegsfeld, 2022) and that indeed the greater determinant of temperature variability was circadian rhythm.

One of the challenges of these sets of data was determining how to handle animals that showed no or little response to poly I:C injection. The phenomenon of “non-responders” comes up frequently when dealing with LPS, and typically these are noted by researchers and either removed or included in the final dataset with 10–20% of LPS-injected animals showing little cytokine response to LPS challenge (unpublished obs.) This phenomenon has been reported in studies in which mice were challenged with LPS and forms the basis for the use of the C3H/HeJ strain in work that probes the immune mechanisms that respond to endotoxin (Grahn et al., 1987; Sultzer et al., 1993). In the present data, we had between 0–3 rats/group show either no response or limited fever response to poly I:C (e.g., less than 1° change from baseline, highest temperatures not cresting 38°C, AUC close to 0°), or, in the bioassay experiments, no change in cytokines after challenge. These animals are visually identifiable in our graphs, which are shown as supplemental data (Supplemental Figure 3). Of course, it is also always possible that some non-responses could be attributable to missed injections or other technical issues, though we are reasonably confident that is unlikely to be the case here. We have elected to keep all animals in, reasoning that perhaps AIE history itself may influence the prevalence of non-responders, positing an interesting question that future targeted studies would need to be sufficiently powered to address. We believe embracing the natural variability in response to antigen will be a cornerstone for progress in understanding individual differences in immunocompetence and is consistent with mouse substrain differences in neuroimmune sensitivity (Warden et al., 2020).

5. Conclusions

The main findings of the present experimental series are of interest for the alcohol field as well as more broadly for understanding public health. We showed that a history of alcohol exposure during adolescence sensitized the adult fever response to a poly I:C challenge exclusively in males, even after a period of prolonged alcohol abstinence of 25 days or longer. AIE exposed males also demonstrated enhanced TLR3, IκBα, and IL-1β expression in the nucleus of the solitary tract following poly I:C challenge, suggesting a central mechanism at play in the observed sensitization of the fever response. The most recent viral pandemic (coronavirus) has demonstrated that chronic alcohol use exacerbated negative outcomes following infection (Xu and Randall, 2023) and in particular revealed a strong sex bias in which men were more vulnerable to adverse outcomes of Covid-19 than women, yet the reasons for these differences remain unclear. Our data therefore raise the intriguing possibility that adolescent binge drinking may be a latent variable that predicts altered fever responses to either vaccinations which commonly use poly I:C as an adjuvant in their formulations or to viral infections such as Covid-19. Additionally, prior work has shown that poly I:C increased voluntary ethanol consumption, in mice rendered ethanol-dependent (Gano et al., 2022) as well as in non-dependent mice (Warden et al., 2018, 2019). In this way, binge drinking may exacerbate the response to viral challenge, and ultimately feed-forward into heightened alcohol consumption, which would in turn further compromise anti-viral immunity, and so forth (Gano et al., 2022, 2023; Lékó et al., 2023).

Data statement:

All data are available upon request.