Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Jul 28.
Published in final edited form as: Pharmacol Biochem Behav. 2026 Jun 12;266:174229. doi: 10.1016/j.pbb.2026.174229

Spontaneous oxycodone withdrawal alters behavior and oligodendrocyte-related gene expression in mice

Olivia M Law 1,**, Jonathan C Gewirtz 1, Jessica L Verpeut 1,**
PMCID: PMC13404751  NIHMSID: NIHMS2189816  PMID: 42285420

Abstract

Negative affective signs are induced during withdrawal from acute opioid injection and increase in magnitude over repeated injections. Recent evidence indicates that opioid use also causes alterations in oligodendrocyte differentiation and myelination. To investigate these phenomena in mice, we examined spontaneous withdrawal following 10 days of oxycodone administration (30 mg/kg/day).

Following a single injection, oxycodone increased exploratory locomotion and elevated thermal pain thresholds, consistent with the acute effect of opioid drugs. Twenty-four hours after oxycodone injection, mice interacted less in an open field, but did not differ from controls in the three-chamber social preference task. The following day (i.e., at the 48-hour withdrawal point), oxycodone-withdrawn animals showed decreased marble burying. Tissue harvested from the medial prefrontal cortex (mPFC) after this test was analyzed for expression for 5 genes associated with regulation of oligodendrocyte cell lineage progression. RT-qPCR revealed decreased expression of Tcf7l2 and Klk6.

These findings indicate behavioral and neurobiological alterations in response to oxycodone. The increased locomotion and antinociception seen following an acute oxycodone dose corroborates past rodent studies. Decreased marble burying during withdrawal represents a novel finding and may reflect anhedonia. The observed changes suggest that alterations in myelin-related gene expression accompany affective signs of acute withdrawal after repeated opioid exposure.

Keywords: social behavior, oxycodone, machine learning, marble burying, oligodendrocytes

1. Introduction

Opioid use disorder (OUD) constitutes a major public health concern in the United States. Approximately 8.6 million people reported misusing prescription opioids in 2023, and about 76% of drug overdose deaths involved opioids (CDC, 2025; Substance Abuse and Mental Health Services Administration, 2024). Given its widespread use and high abuse liability (Cicero et al., 2013; Katz et al., 2008), the semi-synthetic opioid oxycodone is an important focus for investigating the mechanisms underlying opioid dependence.

A major challenge in the treatment of OUD is the management of affective opioid withdrawal symptoms which may affect the risk of relapse. For example, anxiety (Pergolizzi Jr et al., 2020) and anhedonia (Huhn et al., 2016) are common symptoms of opioid withdrawal in humans. Furthermore, social behavior deficits are associated with these negative affective states and with dependence on opioids and other addictive drugs (Christie, 2021; Rokach, 2005; Volkow et al., 2011). Thus, OUD is associated with social isolation, and addiction can result in the devaluation of previously reinforcing stimuli, such as social interactions, in favor of the abused drug. Social isolation during withdrawal has also been demonstrated in rodent models of OUD. For example, mice display a decreased preference for social cues over novel objects in the three-chamber task during precipitated withdrawal following repeated, escalating-dose morphine injections (Ozdemir et al., 2024) and during spontaneous, protracted morphine withdrawal (Pomrenze et al., 2022). Similarly, mice display decreases in freely moving social behavior during morphine abstinence (Bravo et al., 2020), which is dependent on activity of ventral tegmental area (VTA) neurons expressing the mu opioid receptor (Jo et al., 2024).

Importantly, while these studies have demonstrated gross decreases in social interaction during periods of opioid abstinence no studies to date have investigated the specific behavioral repertoires of mice undergoing opioid withdrawal nor the social dynamics of groups. To this end, we first utilized Social Leap Estimates Animal Poses (SLEAP) to track individual poses automatically through machine learning algorithms (Pereira et al., 2019, 2022). This, and similar machine learning programs (Mathis et al., 2018), allow for greater precision and specificity in tracking animal movements across time, and may identify behavioral patterns that are not readily observable by eye. Additionally, the use of machine learning helps to reduce potential researcher biases. Thus, applying machine learning to animals during withdrawal may help identify the nature and extent of behavioral changes during opioid withdrawal to a greater extent than the use of traditional observational measures (Klibaite et al., 2022; Truong et al., 2024; Verpeut et al., 2023). Conversely, machine learning software for social behavior tracking is not typically designed for the analysis of social behavior among more than two animals. Therefore, in order to investigate group dynamics, we also used Behavioral Observation Research Interactive Software (BORIS) to allow for manual logging of specific social behaviors occurring between mice.

Along with the behavioral sequelae related to opioid exposure and withdrawal in mice, it is crucial to better characterize the neurobiological adaptations that underlie them. In this regard, there has been a heightened interest recently in the role of myelination in OUD. White matter deficits are commonly seen in human patients diagnosed with substance use disorders, including opioid use disorder (Hampton et al., 2019). For example, oxycodone-exposed female rats show demyelination, primarily in the cortex (Fan et al., 2018). This could occur through apoptosis or reduced proliferation or maturation of oligodendrocytes, which form the myelin sheath. Conversely, a blockade of oligodendrogenesis prevented morphine conditioned place preference and reduced dopamine release in the nucleus accumbens, suggesting that the process of myelination promotes the development of drug-seeking behavior (Yalçın et al., 2024). We have also found that rats classified as vulnerable or resilient to withdrawal-induced anhedonia (WIA) show differential expression in the medial prefrontal cortex (mPFC) of the transcription factors Tcf7l2 and Sox2, both of which play important roles in myelination-related processes (Liu et al., 2025). Thus, there is considerable evidence of alterations in myelination-related processes following short- or long-term opioid exposure, although the directionality of effects has varied, depending on the paradigm.

The aim of the present study was to examine the effects of acute oxycodone use and withdrawal following repeated oxycodone administration. We hypothesized that spontaneous, acute withdrawal from oxycodone would result in greater anxiety-like behavior, reduced social behavior, and alterations in oligodendrocyte-related gene expression.

2. Methods

2.1. Animals

C57BL/6J mice (n = 66, males: n = 30, females: n = 36; Jackson laboratories) at postnatal day 60 (P60) were used to examine acute oxycodone effects or treatment withdrawal impacts following repeated administration of oxycodone. Mice were injected once per day for 10 days and examined for changes in anxiety, social behavior, and oligodendrocyte gene expression during acute (24 or 48 hour) withdrawal (n = 21, males: n = 12, females: n = 9). Additional subsets of adult mice (~P60) were used to examine acute locomotor (n = 12) and antinociceptive (n = 33) effects of oxycodone. For all experiments, mice were housed together in groups of three with ventilated top cages (Allentown, LLC) under a 12-hour reverse light-dark cycle. This allowed for procedures to occur during the day during the active cycle of the mice. Radio frequency identification (RFID) microchips (United Information Devices) were implanted one day prior to experimentation for individual identification. All microchips were injected subcutaneously under light anesthesia (1.5%) between the shoulder blades at the scruff of the neck. Each microchip consisted of a unique ID number which was used to identify each animal when scanned by the RFID reader. Groups were designed with either two oxycodone-treated animals and one saline-treated animal or one oxycodone-treated animal and two saline-treated animals. Mice were fed an ad libitum diet (Teklad Diets, Madison, WI). All procedures were approved by the Institutional Animal Care and Use Committee at Arizona State University.

2.2. Drug Preparation and Administration

Oxycodone hydrochloride (National Institute on Drug Abuse) was dissolved in 0.9% sterile saline to create working stock solutions. Mice exposed to repeated administration treatment received subcutaneous injections of either 30 mg/kg/day of oxycodone (n = 12) or 0.9% saline vehicle control (n = 9) for a total of 10 days. Subsets of drug naïve mice received a single dose of 5, 10, 20, or 30 mg/kg of oxycodone or saline for acute experiments.

2.3. Hot Plate Task

The hot plate task is commonly used to measure thermal antinociception. This task was used to characterize the potency and time course of multiple doses of oxycodone as a basis for selecting a treatment regimen for the current study. Mice were injected with either 5, 10, 20, or 30 mg/kg of oxycodone or saline vehicle control. At 15-, 30-, and 45-minutes post-injection, mice were placed under a clear Plexiglass beaker and time to withdraw a paw from the hotplate was measured. The hot plate was maintained at 55° C with temperature frequently measured using an infrared thermometer. Mice were removed from the hotplate if they failed to withdraw a paw after thirty seconds to avoid tissue damage.

2.4. Elevated Open Field

2. 4. 1. Acute Administration

The acute effects of oxycodone in the elevated open field arena was measured in a subset of mice. First, mice were placed in the elevated open field arena (Klibaite et al., 2022; Verpeut et al., 2023) for a 15-minute habituation period and were allowed to freely explore. Afterwards, mice immediately received a single injection of either 30mg/kg of oxycodone or saline before being placed back in the open field and filmed for an hour using a machine vision camera (60 fps, Allied Vision) under infrared light.

Mice were tracked using the machine learning algorithm, Social LEAP Estimates Animal Poses (SLEAP) (Pereira et al., 2019, 2022). SLEAP creates trained models to infer the position of similar labeled nodes or keypoints across frames. For mice in the open field, a total of 12 points were used, including the nose, ears, neck, centroid, all four paws, tail base, tail mid, and tail tip. The trained model consisted of a total of 734 frames with a precision score of 93% and a recall score of 92%. This model was then applied to all of the videos to calculate position, speed, and distance traveled across all frames.

2.4.2. Repeated Administration

The elevated open field was also used as a measure during the repeated administration study. In this study, groups of three cagemates were tested simultaneously in the open field in order to assess both individual and group social behavior (Figure 3A). This arena was empty aside from a plexiglass hut in the corner which was used to determine whether the animals would hide more frequently. Mice were first placed together in the open field arena at baseline for 30 minutes prior to oxycodone treatment. After the 10th day of oxycodone administration, animals were retested at the 24-hour withdrawal timepoint for 30 minutes. Individual behavior was analyzed using SLEAP. This multi-animal trained model consisted of a total of 548 labeled frames with a precision score of 92% and recall score of 88%.

Fig. 3. Open Field Task with Individual Behavior Metrics.

Fig. 3.

Open in a new tab

(A) Cartoon image of the open field configuration (figure created with BioRender.com). Three animals were placed in the open field together. The red box represents a hut that the animals could enter or climb to explore. The skeleton is a representation of the nodes used for automated pose tracking using Social LEAP Estimates Animal Poses (SLEAP). (B) Exemplar of tail-base tracks at baseline for one control animal (black) and tracks overlayed with velocity (with faster velocity represented by lighter color). (C) Boxplots representing the log-transformed distance traveled from baseline (baseline = B) to test at 24-hr withdrawal, control (n = 9) and oxycodone animals (n = 12). (D) Boxplots representing the log-transformed time spent in the center and around the hut from baseline to test (mixed model ANOVAs). **p < 0.01

Group social behavior was hand-scored using Behavioral Observation Research Interactive Software (BORIS) (Friard & Gamba, 2016). The labeled social behaviors between two mice comprised nose-to-nose touching, anogenital sniffing, paired exploration, nose-to-body sniffing, and pursuit (Figure 4A). For social behaviors that were led by one mouse (i.e., one mouse was sniffing the body of another mouse who did not reciprocate), only the instigator was counted as showing the social behavior. For behaviors that involved both mice (i.e., both mice touching noses), the social behavior instance was counted as occurring for each mouse. The total number of social behavior events and the duration of social behavior were computed for each mouse.

Fig. 4. Open Field Group Social Behavior.

Fig. 4.

Open in a new tab

(A) Cartoon depictions of each of the social behaviors that were annotated using BORIS (figure created with BioRender.com). (B) Boxplots of log-transformed number of social behaviors at baseline and test for all of the social behaviors combined. (C) Boxplots of log-transformed durations of social behaviors at baseline and test for all social behaviors combined. Mixed model ANOVAs with Bonferroni-corrected post-hoc tests were used for analysis. **p < 0.01, ***p < 0.001

2.5. Three-Chamber Social Task

The three-chamber task was used to assess social preference. Mice were injected on the 9th day, then tested in the three-chamber task the next morning approximately 24 hours later. After the completion of the three-chamber task, mice were then given their final injection. Similar to Badura et al., 2018, the apparatus consisted of three connected plexiglass chambers. This procedure includes two phases: a habituation phase and a test phase. During the habituation phase, two empty wire cups were placed in the center of the left and right chambers. Mice were released in the center (empty) chamber and allowed to freely explore all chambers for 15 minutes. During the test phase, a novel object was placed under one cup, and a same-sex and age-matched novel mouse was placed under the opposite cup. The novel mouse and novel object positions were balanced across animals to control for any possible side bias. In this phase, mice were also allowed to explore for 15 minutes. Each animal was filmed using SpinView and a machine vision camera (Blackfly S USB3, Teledyne), and all tests took place under red light.

All three-chamber videos were analyzed using SLEAP. A total of eight node points were labeled, including the nose, ears, neck, centroid, tail base, tail middle, and tail tip. The trained model consisted of a total of 223 labeled frames with a precision score of 92% and a recall score of 96%. For all SLEAP models that were used to predict animal tracks per frame each video, predictions with a score of less than 1 were removed.

2.6. Marble Burying

The marble burying task is frequently used to measure anxiety- and compulsive-like behavior. Mice were tested in the task at approximately 48 hours withdrawal. They were placed in individual cages filled with Sani-chips under red light and allowed 10 minutes to habituate to the cages with no marbles inside. After the 10 minutes had elapsed, mice were briefly removed and a 5×4 grid of marbles was placed in each cage. The mice were then returned to the cages and filmed for 15 minutes. After mice were removed from the cage, experimenters who were blind to condition then counted the number of marbles buried 2/3 or more in each cage.

2.7. Reverse Transcription Quantitative Polymerase Chain Reaction (RT-qPCR)

Tissue collection took place immediately following the marble burying task. Mice were briefly anesthetized with an overdose of carbon dioxide (2.5–5%) and decapitated. The prefrontal cortex was immediately dissected and homogenized in 1 mL of RNA STAT-60 (Tel-Test) using a sonicator. Samples were frozen on dry ice and stored at −80° C.

For RNA extraction, 200 μL of chloroform was added to each sample. Samples were left on ice for 5 minutes after mixing and then centrifuged at the maximum RCF for 10 minutes at 4° C. With the sample now separated into three distinct layers, 350 μL of the aqueous layer was collected and transferred to an Eppendorf tube. An equivalent volume of 70% ethanol was added to the same tube and mixed. The mix was then transferred to a spin column and centrifuged at 11,000 RCF for 30 seconds at 4° C. The RNeasy Mini Kit (Qiagen) was used for RNA preparation. For DNAse digestion, 350 μL of RW1 buffer was added to the spin column and centrifuged at 11,000 RCF for 30 seconds at 4° C. Next, 80 μL of incubation mix consisting of DNAse 1 Stock Solution and Buffer RDD (Qiagen) was added to the spin column. The samples were then incubated for 15 minutes at room temperature. Following the incubation period, 350 μL of RW1 was added to the spin column which was then centrifuged at 11,000 RCF for 30 seconds at 4° C. Afterwards, 500 μL of RPE buffers was added to spin columns (2x) and spun at 11,000 RCF for 30 seconds at 4° C and then for 2 minutes. Samples were spun at the maximum RCF for 1 minute at 4° C to completely dry the spin column. Finally, RNA was resuspended by adding 30 μL of nuclease-free water and spinning at 11,000 RCF for 1 minute at 4° C to elute the RNA. RNA concentration and purity were quantified using a nanodrop spectrophotometer (Thermofisher Nanodrop One).

For cDNA conversion, a pre-anneal mixture was created, consisting of 1μL of oligo dT (5′-TTTTTTTTTTTTTTTTTTTT-3′; Sigma-Aldrich, Burlington MA, USA), RNA based on 2μg as quantified with a nanodrop, and nuclease free water for a total volume of 11.75μL. The mix was centrifuged for 630 RCF for 1 minute at 4° C. Samples were then run in a thermocycler at 70° C for 3 minutes, then at 4° C for 2 minutes. The post-anneal mix (n + 1 samples) was prepared, consisting of 4 μL of 5×1st strand buffer, 2 μL of 10mM dNTPs, 2 μL of 100mM DTT, and 0.25 μL of RT Smart MMLV. For each sample, 8.25 μL of post-anneal mix was added. Samples were centrifuged at 630 RCF for 1 minute at 4° C. Next, samples were added to the thermocycler and ran at 42° C for 60 minutes, 70° C for 15 minutes, and 10° C for 5 minutes. Finally, samples were diluted with 160 μL of nuclease-free water. Samples were stored at −20° C.

The selected myelination-related genes include Mbp, Olig1, Plp1, Klk6, and Tcf7l2. Myelin basic protein (Mbp) is an essential component of the myelin sheath in mature oligodendrocytes. Oligodendrocyte transcription factor 1 (Olig1) is involved in oligodendrocyte development, maturation, and remyelination. Proteolipid protein 1 (Plp1) is the main structural component of the myelin sheath. Kallikrein-related peptidase 6 (Klk6) is a serine protease expressed in oligodendrocytes, neurons, and astrocytes in the CNS and has been implicated in processes such as demyelination and neuroinflammation (Mella et al., 2020). The Wnt transcription factor 7 like 2 (Tcf7l2) is highly expressed in pre-myelinating oligodendrocyte cells and is explicitly associated with active myelination and re-myelination (Fu et al., 2011; Guo & Wang, 2022).

For qPCR, a 3 μM FW/RV primer mix was prepared for each of the primers used. Primers were obtained from KiCqStart™ SYBR® Green Primers (MilliporeSigma, Burlington, MA, USA) for mouse genes Klk6 (Gene ID: 19144; RefSeq ID: NM_001164696), Olig1 (Gene ID: 50914, RefSeq ID: NM_016968), Mbp (Gene ID: 17196, RefSeq ID: NM_001025251), Plp1 (Gene ID: 11823, RefSeq ID: NM_011123), Tcf7l2 (Gene ID: 21416, RefSeq ID: NM_001142918), and Rpl13a (Gene ID: 22121, RefSeq ID: NM_009438). Next, 3.5 μL of cDNA were added into qPCR 96-well plates, followed by 6.5 μL SYBR/primer mix for a total volume of 10 μL per well. Plates were sealed and run on an ABI Viia7 qPCR instrument. The RT-qPCR run protocol was as follows: 95° C for 3 minutes, 95° C for 15s, 60° C × 1 minute (40x), 95° C for 1 minute, 55° C for 30s, and 95° C for 30s. Samples were run in triplicate, and only samples with a standard deviation of less than 0.5 were included.

2.8. Statistical Analyses

All statistical and data analyses were performed using RStudio or Python. A mixed model ANOVA with latency and time as the factors was used to assess overall differences in the hot plate task. Follow-up between and within-subject ANOVAs or their non-parametric alternatives (Kruskal-Wallis or Friedman’s test) with Bonferroni-corrected post hoc tests where warranted, were conducted to assess latency to withdraw the paw in the hotplate test between doses and at three post-injection timepoints (15, 30, and 45 minutes post-injection. Distance traveled and time spent in the open field after acute oxycodone was analyzed using an independent sample’s t-test or Welch’s t-test. Additionally, a mixed-model ANOVA was used to assess differences between the control and experimental group during the first and second 5-minute time bins. For the repeated administration open field analyses, distance-traveled and time-spent variables were log-transformed, and a mixed-model ANOVA was used to assess the effect of timepoint (baseline vs test) and experimental condition (control vs oxycodone). If warranted, Bonferroni-corrected post-hoc tests were conducted. Mixed model ANOVAs with log transformed data were also used to examine whether total social behavior counts and duration depended on timepoint and condition. Additionally, mixed model ANOVAs with log-transformed data were used to examine the effect of timepoint and condition on each individual social behavior duration and count with Bonferroni-corrected post hoc tests. Paired sample t-tests were used to test for time spent with the novel mouse compared to the novel object in the three-chamber task. To examine social preference and total distance traveled between control and oxycodone animals, independent-sample t-tests were conducted. Welch’s t-test was used to compare the number of marbles buried between control and oxycodone animals. RT-qPCR analysis was performed by using independent sample t-tests or Wilcoxon rank sum tests. Linear regression models were used to assess the association between significant behavioral and neural changes. No sex differences were found, so the sexes were combined for each test. For all statistical tests, the significance level was set at p < 0.05.

3. Results

3.1. Acute Administration Analyses

3.1.1. Hot Plate

Latency to withdraw the forepaw was compared at each time point (15, 30, and 45 minutes post-injection) between control, 5-, 10-, 20-, and 30-mg/kg doses (Figure 2). There was a significant main effect of dose (F(4, 28) = 9.93, p < 0.001) and time (F(1.54, 43.22) = 14.32, p < 0.001), as well as an interaction effect, F(6.17, 43.22) = 2.85, p = 0.19. At 15 minutes post-injection, there was a significant effect of dose on latency, χ2(4) = 19.05, p < 0.001 (Kruskal-Wallis). Mice that received 30 mg/kg had a significantly higher latency compared to control animals (Dunn’s post-hoc test, Bonferroni-corrected, p = 0.009). Additionally, all other doses also showed a higher latency compared to control animals (5 mg: p = 0.024, 10 mg: p = 0.048, 20 mg: p = 0.008). At 30 minutes post-injection, there was a significant effect of dose on latency, F(4, 28) = 3.84, p = 0.013. The 30-mg animals showed a significantly higher latency than control animals (post-hoc test, Bonferroni-corrected, p = 0.012). Lastly, at 45 minutes post-injection, there was a significant effect of dose on latency, χ2(4) = 13.352, p = 0.0097 (Kruskal-Wallis), but no significant differences in latency between any doses.

Fig. 2. Hot Plate Task.

Fig. 2.

Open in a new tab

Boxplots of latency to withdraw the paw (seconds) measured at 15, 30, and 45 minutes post-injection between the 0 (control), 5, 10, 20, or 30 mg/kg dose of oxycodone. Effects were analyzed using between and within subject ANOVAs and Bonferroni corrected post-hoc tests. At 15 minutes post-injection, all doses of oxycodone had higher latencies than control animals. Only the 30 mg/kg oxycodone animals showed higher latencies than control animals at both 15- and 30-minutes post-injection. *p < 0.05, **p < 0.01

Comparing across time within each dose, there were no significant differences across time among control, 20-, or 30 mg-dosed mice. As expected, there was a significant effect of timepoint on latency for both the 5- (χ2(2) = 7.19, p = 0.028, Friedman test) and 10-mg (χ2(2) = 7.41, p = 0.025, Friedman test) treated mice. These results indicate that the 5 and 10 mg doses have a shorter antinociceptive effect compared to the higher doses, as expected. The 30 mg/kg dose produced full antinociception for up to 30 minutes. Based on these results, the rest of the study focuses on the 30 mg/kg dose.

3.1.2. Open Field

Oxycodone and other opioids are known to have impacts on locomotion. To analyze these changes in our mice both after acute exposure and withdrawal, an open field was used to measure velocity, total distance traveled, and time spent in different areas of the arena (Supplementary Figure 1). Oxycodone-injected animals showed an increased total distance traveled relative to saline animals (W = 0, p = 0.002). There was no significant difference in total distance traveled in the center of the arena or in time spent in the center of the arena (Supplementary Figure 1B).

Further analyses were conducted to ascertain potential differences in distance traveled across time between oxycodone and control animals (Supplementary Figure 1B-C). When examining differences between the first and second five-minute time bins, a significant interaction was found between time and condition, F(1,10) = 12.53, p = 0.005 (Mixed ANOVA). Specifically, the oxycodone mice showed greater locomotion in the first five minutes compared to controls (p < 0.001) and compared to the level of locomotion in the next five minutes (p = 0.009). Therefore, oxycodone administration produced hyperlocomotion in the open field.

3.2. Repeated Administration Analyses

3.2.1. Open Field Individual Behavior

To analyze group dynamics in the open field, we first analyzed locomotion behavior in individual mice. Distance traveled and time spent in different areas of the arena were computed and compared between oxycodone-withdrawn (n = 12) and control animals (n = 9) (Figure 3A-B). Differences in distance traveled and time spent in different sections of the elevated open field arena were examined between oxycodone and control animals at baseline versus the test (during withdrawal). For total distance traveled across the entire 30 minutes, there was a significant main effect of timepoint (F(1,19) = 14.744, p = 0.001) with total distance traveled decreasing from baseline to test, but no main effect of group or interaction effect. Distance traveled in the center of the arena also decreased from baseline to test (F(1,19) = 6.923, p = 0.165) but there were no differences in group nor an interaction effect. For distance traveled around the hut, there was again only a significant main effect of timepoint (F(1,19) = 5.762, p = 0.027) with a significant decrease from baseline to test (Figure 3C). There were no significant differences between groups in time spent (seconds) in the center of the arena (Figure 3D). Additionally, there were no significant differences in time spent around the hut.

3.2.2. Open Field Group Social Behavior

Next, we analyzed within-group social interactions using BORIS. Social behavior poses were defined as: anogenital sniffing, pursuit, nose-to-nose, nose-to-body, paired exploration, and side-side (Figure 4A). These behaviors were summed together for each animal (Figure 4B-C). All data were log-transformed for analyses (raw data graphs available in Supplement). For combined social behavior events, there was a significant main effect of timepoint (F(1,19) = 29.421, p < 0.001) and a significant time × group interaction (F(1,19) = 9.486, p = 0.006). This meant that mice dosed with oxycodone demonstrated a significant decrease in the number of social behavior events from baseline to test (Bonferroni, p < 0.001) and compared to control animals at baseline (Bonferroni, p = 0.006). On the other hand, there were no significant differences in combined social behavior durations.

For further analysis, social behavior event counts (Figure 5A) and durations (Figure 5B) were analyzed separately for each behavioral categorization. Social behavior counts that differed from baseline to test included anogenital sniffing (F(1,19) = 6.146, p = 0.023), nose-to-body (F(1,19) = 15.475, p < 0.001), and nose-to-nose (F(1,19) = 16.857, p < 0.001). Furthermore, nose-to-body analyses revealed an interaction effect between time × condition (F(1,19) = 17.167, p < 0.001), such that oxycodone-withdrawn animals showed a decreased number of nose-to-body interactions when compared to baseline (Bonferroni, p < 0.001) and when compared to control animals during the test (Bonferroni, p = 0.0098). Social behavior durations that were significantly different between baseline and test included anogenital sniffing (F(1,19) = 7.223, p = 0.015), side-side (F(1,19) = 4.721, p = 0.043), and nose-to-nose (F(1,19) = 13.667, p = 0.001). There was also a significant time × group interaction effect for nose-to-body duration (F(1,19) = 4.539, p = 0.046). There were no differences in pursuit or in paired exploration. Therefore, BORIS revealed distinct changes in social behavior dynamics between groups.

Fig. 5. Durations and Counts for Each Social Behavior.

Fig. 5.

Open in a new tab

(A) Boxplots of the log-transformed number of social behaviors at baseline versus test in control and oxycodone-withdrawn animals split up by behavior. Anogenital sniffing and nose-to-nose showed a significant main effect of time point (mixed model ANOVA). Nose-to-body showed a significant time point × group interaction (mixed model ANOVA) and follow-up Bonferroni-corrected post-hoc tests revealed that oxycodone-withdrawn animals showed fewer nose-to-body interactions at test compared to control animals and to their baseline values. (B) Boxplots of the log-transformed duration of social behaviors at baseline versus test in control and oxycodone-withdrawn animals split up by behavior; anogenital sniffing, nose-to-body, and nose-to-nose showed a significant main effect of time point (mixed model ANOVA). *p < 0.05, **p < 0.01, ***p < 0.001

3.2.3. Three-Chamber Task

The three-chamber task was used to measure social preference (Figure 6A). Animals were tracked using SLEAP and time spent in different areas of the apparatus was visualized using heatmaps (Figure 6B). Control animals (n = 8, t(7) = 2.946, p = 0.022, paired t-test) as well as oxycodone-withdrawn animals (n = 10, t(9) = 3.56, p = 0.006) spent more time with the novel social animal compared to the novel object (Figure 6C). Overall no differences were found in social preference between control- and oxycodone-treated mice. (t(16) = 1.226, p = 0.904) (Figure 6D). There were also no differences in total distance traveled between the two groups, t(16) = 0.997, p = 0.334 (Supplementary Figure 5). Therefore, while we see differences in freely-interacting group social behavior, social preference is not impacted.

Fig. 6. Three-Chamber Social Task.

Fig. 6.

Open in a new tab

(A) Cartoon depicting the three-chamber task during baseline (right before the test) and the social discrimination test (figure created with BioRender.com). (B) Representative image of one mouse’s nose tracks of the three-chamber arena extracted from SLEAP and corresponding heat map (brighter colors of yellow and red = more time spent). (C) Boxplots of time spent in seconds around the cup with the object and with the mouse (social) for control and oxycodone-withdrawn animals (paired samples t-test). (D) Boxplots of social preference ratio (time spent social/time spent social + time spent object) for control and oxycodone-withdrawn animals. *p < 0.05, **p < 0.01

3.2.4. Marble Burying

To assess anxiety and compulsive-like behaviors, the marble burying test was used (Figure 7A). There was a significant difference between the groups (t(15.967) = 3.464, p = 0.003; Welch’s t-test) with oxycodone-withdrawn animals burying fewer marbles on average compared to control (Figure 7B).

Fig. 7. Marble Burying.

Fig. 7.

Open in a new tab

(A) Images taken after the marble burying task to count the number of marbles buried. The circles indicate marbles that were counted for a representative control and oxycodone animal. (B) Boxplot of number of marbles buried ⅔ or more by group (Welch’s t-test) **p < 0.01.

3.2.5. RT-qPCR

RT-qPCR was used to analyze oligodendrocyte-related gene expression in mice that were exposed to repeated oxycodone compared to control animals. All Ct values for each gene of interest were normalized to a housekeeping gene (Rpl13a), which did not significantly differ between groups, to obtain the ΔCt value. There were no significant differences in expression of Olig1, Mbp, or Plp1 between oxycodone-treated and control animals. In contrast, oxycodone-treated animals showed a significant increase in Tcf7l2 ΔCt values (Wilcoxon rank-sum test; W = 6, p = 0.009) and in Klk6 ΔCt values (t(11) = −2.266, p = 0.045) compared to control. The formula 2^(−ΔΔCt) was used to calculate and plot relative fold change (Figure 8). To further classify whether alterations in Klk6 and Tcf7l2 are related to our behavioral outcomes during withdrawal, we used linear regression (Supplementary Figure 6). No associations were found between these genes and our significant behavioral measures (total group social behavior count and number of marbles buried).

Fig. 8. RT-qPCR for Oligodendrocyte-Related Markers.

Fig. 8.

Open in a new tab

Boxplots of relative fold change 2^(−ΔΔCt) by gene, comparing control and oxycodone animals. Oxycodone animals showed decreased expression of Klk6 (independent samples t-test) and Tcf7l2 (Wilcoxon rank-sum test). *p < 0.05, **p < 0.01

4. Discussion

The current study examined behavioral response to an acute dose of oxycodone, as well as withdrawal following repeated-administration of oxycodone in mice. All doses produced similar maximal antinociception at the first test time of 15 min, but as expected, the highest doses of 20 and 30 mg/kg had the longest durations of action. An acute 30 mg/kg dose of oxycodone produced locomotor stimulation in an open field, a well-established effect of repeated administration, even after a single injection (Vanderschuren et al., 2001). While we did not address the long term effects of this acute oxycodone dose, an assessment of more protracted behavioral and neurobiological adaptations following a single dose in the future would be valuable (Rothwell et al., 2009). Based on these results, the highest dose (30 mg/kg) was chosen as the dose for the repeated injection regimen. Although other studies have used higher doses to specifically assess physical dependence, giving too high of a dose repeatedly could lead to toxicity or other problematic secondary effects such as sedation which could reduce interpretability of behavioral findings.

While other opioid withdrawal studies have examined the frequency/duration of social behavior in general, preclinical studies rarely measure separate social behavior interactions. Therefore, this study provides deeper insight into the behavioral phenotype of opioid-withdrawn animals. In the current study, social behaviors were hand-annotated using BORIS (Friard & Gamba, 2016). Using BORIS, the number of total social behavior events were found to be decreased in oxycodone-withdrawn mice compared to their baseline values. However, there was no significant difference when comparing the oxycodone-withdrawn animals to the control animals during the test. This may be indicative of the importance of measuring individual changes across time. Interestingly, we also found decreased nose-to-body counts of oxycodone-withdrawn animals compared to baseline and control values. It is important to note that no other social behaviors significantly differed between oxycodone-withdrawn and control animals.

Specific social behavior topology during withdrawal has not been classified to this extent in prior addiction studies. As such, the ethological relevance of the nose-to-body behavior and other social behavior has yet to be fully understood in this context. One hypothesis is that this behavior represents a difference in approach and contact behavior specifically toward a familiar conspecific and could represent dominance behavior (Wesson, 2013). Further analyses of social behaviors using machine learning could provide greater insight into these behavioral phenomena, as such programs could allow for detection of more subtle changes in group dynamics. To control for cage-level social influence, animals were tested in mixed-treatment groups (oxycodone and saline) to maintain comparable environments and social conditions, but future studies could test animals housed in identical treatment groups to see how cage conditions impact social behavior.

Compared to the group social behavior task, the three-chamber task examines an individual mouse’s preference for a novel conspecific over an inanimate object. While the group social behavior task assessed social behavior within groups of familiar cagemates, this task examines social novelty. Meanwhile, in the group social behavior task, being together in a confined space forces interaction, and the other conspecifics can influence interactions. The group task allows us to measure both reciprocal (e.g., paired exploration) and non-reciprocal (e.g., anogenital sniffing). In this task, the animal can choose whether or not to interact with the novel animal or object, and preference for each (or neither) is assessed. Interestingly, there were no differences between oxycodone and control groups in time spent with a social partner versus a novel object. A reduced social preference may have been expected, given that mu-opioid receptors modulate social behavior, including social novelty preference (Smith et al., 2015). However, in support of our findings, it has been shown that morphine-abstinent rats show intact social preferences (Aghadoukht et al., 2025). Whereas studies reporting social behavior deficits, social behavior have focused on protracted abstinence (Goeldner et al., 2011; Lutz et al., 2014), the current study measured social behavior during early withdrawal. Therefore, our findings combined with past literature suggest that social preference differences may not emerge until later abstinence timepoints.

The discrepancy between the lack of findings in the three-chamber task compared to the social deficits seen in the group social behavior task highlights the importance of measuring more complex social behavior in a freely interacting group paradigm. Therefore, while the three-chamber task has frequently been used as an assessment of social preference, utilizing additional social behavioral paradigms can be advantageous in understanding more nuanced behaviors (Jabarin et al., 2022).

To examine anxiety- and compulsive-like behavior we used the marble bury task. Here, oxycodone-withdrawn animals buried significantly fewer marbles compared to control animals. This stands in contrast with a previous finding of increased marble burying during opioid withdrawal (Umathe et al., 2012). Differences in type of opioid, dosage, withdrawal timepoint, or mouse strain may underlie this discrepancy. Alternatively, the fact that we saw an increase in burying, rather than a null effect, likely reflects the presence of withdrawal-induced anhedonia, rather than anxiety. We have previously found a transition from anxiety- to dysphoric-like behavior during morphine withdrawal in rats, depending on whether behavior is assessed several or 24–48 hours after opioid exposure, as was the case here (Rothwell et al., 2009; Swain et al., 2020). Although the task was not designed to measure anhedonia or dysphoria, a recent study that quantified a range of behaviors found that marbles can become buried either through the direct act of burying and/or as an indirect consequence of digging/burrowing in the vicinity of the marbles (Coppola et al., 2026). Indeed, while the degree of marble burying differs substantially between 10 strains of mice, it correlates strongly across strains with the degree of digging behavior exhibited and not at all with other measures of anxiety-like behavior (Thomas et al., 2009). Digging itself is motivated behavior that can be depressed by pharmacological or behavioral stressors and rescued by the anti-depressant fluoxetine (Partridge & Hillhouse, 2024; Pond et al., 2021). Given the greater appreciation of the task’s complexity (Burgett et al., 2024; Pond et al., 2021), and that marble burying could be bidirectionally modulated by two different negative valences (anxiety and anhedonia), it would be valuable for future studies to employ detailed behavioral analysis to disambiguate anxiety- (i.e., targeted burying) versus anhedonia- (i.e., reduced burrowing) like behaviors (Burgett et al., 2024).

In addition to assessing the behavioral features of opioid withdrawal, it is essential to investigate the underlying molecular changes associated with this state. Due to the white matter deficits seen in human patients with substance use disorders, we were interested in examining oligodendrocytes, the glial cells that form the myelin sheath around axons to produce white matter. Opioid receptors are expressed on developing oligodendrocytes (Knapp et al., 1998), and the endogenous opioid system is postulated to regulate oligodendrocyte development such that administration of exogenous opioids may disrupt this process (Velasco et al., 2021). Because alterations in oligodendrocyte expression affects neuronal communication, understanding how these glial cells may be associated with opioid use, especially in regions implicated in addiction, is an important avenue of research. For example, compromised myelin integrity could disrupt neural connectivity critical for processes such as decision-making and inhibitory control, resulting in disinhibited drug use.

In the current study, we found decreased expression of two genes in the mPFC in oxycodone-withdrawn animals: Tcf7l2 and Klk6. We have identified Tcf7l2 as an upstream regulator of gene network activity in the mPFC associated with anhedonia following morphine withdrawal (Liu et al., 2025). Oxycodone animals also showed decreased expression of Klk6 expression, a serine protease secreted by mature oligodendrocytes (Yoon et al., 2022). This suggests that specific stages of oligodendrogenesis are differentially impacted by repeated oxycodone administration and subsequent acute withdrawal.

Two of our genes of interest, Tc7fl2 and Klk6, were specifically altered in the oxycodone-withdrawn group while the other three genes (Mbp, Olig1, and Plp1) were not affected. Consistent with this finding, Zhang et al. (2025) found that Tcf7l2 deletion in differentiating oligodendrocytes impairs myelin sheath morphology without significantly affecting gene expression of myelin proteins. This suggests that Tcf7l2 may influence myelination through other pathways. Another study used a Klk6 knockdown, finding an increased number of mature oligodendrocytes in the spinal cord in juvenile mice, but this increase was not seen in adulthood (Murakami et al., 2013). Therefore, Klk6 is implicated more specifically during early developmental myelination and remyelination (Yoon et al., 2021). The decrease in Klk6 in the current study may thus reflect other mechanistic changes. For example, Klk6 may represent a separate myelination pathway like Tcf7l2 or may be related to another CNS function. Indeed, Klk6 is expressed in multiple cell types and is involved in other processes such as extracellular matrix remodeling and neuroinflammation. Taken altogether, past research suggests a role for Tcf7l2 and Klk6 in myelination-related processes. However, because the results of the current study did not find changes in myelin proteins, we can not say for certain that the changes in these genes represent white matter deficits. A future direction would be to examine myelin sheath morphology during opioid withdrawal.

This study provides us with greater insight into the neurobehavioral effects of oxycodone administration and withdrawal. First, we classified social behaviors during withdrawal from oxycodone with finer grained specificity than in previous studies, finding a decreased number of social behavior events overall and a decrease in the specific dyadic behavior of nose-to-body. Additionally, marble burying was reduced during withdrawal, potentially indicative of motivational anhedonia. Third, in terms of neural substrates, our findings add to a growing body of work across a range of addictive substances implicating a potential role for oligodendrocytes (Miguel-Hidalgo, 2009; Velasco et al., 2021). Future research will directly target the trajectory of oligodendrocyte differentiation and myelination to investigate whether these factors mitigate opioid-seeking behavior and the negative affective consequences of opioid withdrawal.

Supplementary Material

MMC2
MMC1

Fig. 1. Experimental Timeline.

Fig. 1.

Open in a new tab

In the pre-injection phase mice were briefly anesthetized for microchip implantation which were used for identification purposes. Additionally, baseline testing for open field behavior was conducted prior to injections. Daily injections occurred over 10 days of either 30 mg/kg/day of oxycodone or saline vehicle control. On Day 10, approximately 24 hours after the injection on day 9, the three-chamber task was conducted. This task was followed immediately by the final injection. On Day 11, ~24 hours after the final injection, the open field test was conducted again. At ~48 hours withdrawal, the marble burying task was conducted and brain tissue was collected.

Highlights.

  • Acute injection of oxycodone increased locomotion and thermal antinociception

  • Withdrawal from repeated oxycodone results in decreased marble burying

  • Social behavior events were reduced during oxycodone withdrawal

  • Withdrawal decreased expression of oligodendrocyte-related genes

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Aghadoukht A, Nasehi M, Ebrahimi-Ghiri M, Sadat-Shirazi M-S, Mahboubi S, & Zarrindast M-R (2025). Role of opioid tapering in rescuing anxiety and social novelty deficits in morphine-withdrawn male rats. Biochemical and Biophysical Research Communications, 778, 152353. 10.1016/j.bbrc.2025.152353 [DOI] [PubMed] [Google Scholar]
  2. Bravo IM, Luster BR, Flanigan ME, Perez PJ, Cogan ES, Schmidt KT, & McElligott ZA (2020). Divergent behavioral responses in protracted opioid withdrawal in male and female C57BL/6J mice. The European Journal of Neuroscience, 51(3), 742–754. 10.1111/ejn.14580 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Burgett LA, Alexander TB, Moya HA, & Kingsbury MA (2024). A revision to marble burying: Video analysis during the marble burying task is imperative to understanding rodent behavior. Lab Animal, 53(12), 387–389. 10.1038/s41684-024-01459-3 [DOI] [PubMed] [Google Scholar]
  4. CDC. (2025, July 12). Understanding the Opioid Overdose Epidemic. Overdose Prevention. https://www.cdc.gov/overdose-prevention/about/understanding-the-opioid-overdose-epidemic.html [Google Scholar]
  5. Christie NC (2021). The role of social isolation in opioid addiction. Social Cognitive and Affective Neuroscience, 16(7), 645–656. 10.1093/scan/nsab029 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Cicero TJ, Ellis MS, Surratt HL, & Kurtz SP (2013). Factors influencing the selection of hydrocodone and oxycodone as primary opioids in substance abusers seeking treatment in the United States. PAIN®, 154(12), 2639–2648. 10.1016/j.pain.2013.07.025 [DOI] [PubMed] [Google Scholar]
  7. Coppola I, De Magistris M, Macchioni G, Calamandrei G, Ricceri L, & Tartaglione AM (2026). Beyond marbles buried: A multidimensional analysis of the marble burying test in the C57BL/6J mouse strain. Behavioural Brain Research, 497, 115888. 10.1016/j.bbr.2025.115888 [DOI] [PubMed] [Google Scholar]
  8. Fan R, Schrott LM, Arnold T, Snelling S, Rao M, Graham D, Cornelius A, & Korneeva NL (2018). Chronic oxycodone induces axonal degeneration in rat brain. BMC Neuroscience, 19(1), 15. 10.1186/s12868-018-0417-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Friard O, & Gamba M (2016). BORIS: A free, versatile open-source event-logging software for video/audio coding and live observations. Methods in Ecology and Evolution, 7(11), 1325–1330. 10.1111/2041-210X.12584 [DOI] [Google Scholar]
  10. Fu H, Kesari S, & Cai J (2011). Tcf7l2 is Tightly Controlled During Myelin Formation. Cellular and Molecular Neurobiology, 32(3), 345–352. 10.1007/s10571-011-9778-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Goeldner C, Lutz P-E, Darcq E, Halter T, Clesse D, Ouagazzal A-M, & Kieffer BL (2011). Impaired Emotional-Like Behavior and Serotonergic Function During Protracted Abstinence from Chronic Morphine. Biological Psychiatry, Cognitive Flexibility and Neuroadaptations to Cocaine, 69(3), 236–244. 10.1016/j.biopsych.2010.08.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Guo F, & Wang Y (2022). TCF7l2, a nuclear marker that labels premyelinating oligodendrocytes and promotes oligodendroglial lineage progression. Glia, 71(2), 143. 10.1002/glia.24249 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Hampton WH, Hanik IM, & Olson IR (2019). Substance Abuse and White Matter: Findings, Limitations, and Future of Diffusion Tensor Imaging Research. Drug and Alcohol Dependence, 197, 288–298. 10.1016/j.drugalcdep.2019.02.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Huhn AS, Meyer RE, Harris JD, Ayaz H, Deneke E, Stankoski DM, & Bunce SC (2016). Evidence of anhedonia and differential reward processing in prefrontal cortex among post-withdrawal patients with prescription opiate dependence. Brain Research Bulletin, Addiction: From Mouse to Man, 123, 102–109. 10.1016/j.brainresbull.2015.12.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Jabarin R, Netser S, & Wagner S (2022). Beyond the three-chamber test: Toward a multimodal and objective assessment of social behavior in rodents. Molecular Autism, 13(1), 1–29. 10.1186/s13229-022-00521-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Jo AY, Xie Y, Rodrigues A, Sandoval Ortega RA, Creasy KT, Beier KT, Blendy JA, & Corder G (2024). VTA μ-opioidergic neurons facilitate low sociability in protracted opioid withdrawal. bioRxiv: The Preprint Server for Biology, 2024.07.08.602522. 10.1101/2024.07.08.602522 [DOI] [Google Scholar]
  17. Katz N, Fernandez K, Chang A, Benoit C, & Butler SF (2008). Internet-based Survey of Nonmedical Prescription Opioid Use in the United States. The Clinical Journal of Pain, 24(6), 528. 10.1097/AJP.0b013e318167a087 [DOI] [PubMed] [Google Scholar]
  18. Klibaite U, Kislin M, Verpeut JL, Bergeler S, Sun X, Shaevitz JW, & Wang SS-H (2022). Deep phenotyping reveals movement phenotypes in mouse neurodevelopmental models. Molecular Autism, 13(1), 12. 10.1186/s13229-022-00492-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Knapp PE, Maderspach K, & Hauser KF (1998). Endogenous opioid system in developing normal and jimpy oligodendrocytes: μ and κ opioid receptors mediate differential mitogenic and growth responses. Glia, 22(2), 189–201. 10.1002/(SICI)1098-1136(199802)22:2<189::AID-GLIA10>3.0.CO;2-U [DOI] [PubMed] [Google Scholar]
  20. Liu SX, Muelken P, Maxim ZL, Ramakrishnan A, Estill MS, LeSage MG, Smethells JR, Shen L, Tran PV, Harris AC, & Gewirtz JC (2025). Transcriptional signatures in the rat medial prefrontal cortex associated with vulnerability and resilience across distinct phases of opioid use disorder (p. 2024.02.29.582799). bioRxiv. 10.1101/2024.02.29.582799 [DOI] [Google Scholar]
  21. Lutz P-E, Ayranci G, Chu-Sin-Chung P, Matifas A, Koebel P, Filliol D, Befort K, Ouagazzal A-M, & Kieffer BL (2014). Distinct Mu, Delta, and Kappa Opioid Receptor Mechanisms Underlie Low Sociability and Depressive-Like Behaviors During Heroin Abstinence. Neuropsychopharmacology, 39(11), 2694–2705. 10.1038/npp.2014.126 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Mathis A, Mamidanna P, Cury KM, Abe T, Murthy VN, Mathis MW, & Bethge M (2018). DeepLabCut: Markerless pose estimation of user-defined body parts with deep learning. Nature Neuroscience, 21(9), 1281–1289. 10.1038/s41593-018-0209-y [DOI] [PubMed] [Google Scholar]
  23. Mella C, Figueroa CD, Otth C, & Ehrenfeld P (2020). Involvement of Kallikrein-Related Peptidases in Nervous System Disorders. Frontiers in Cellular Neuroscience, 14, 166. 10.3389/fncel.2020.00166 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Miguel-Hidalgo JJ (2009). The Role of Glial Cells in Drug Abuse. Current Drug Abuse Reviews, 2(1), 76–82. 10.2174/1874473710902010076 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Murakami K, Jiang Y-P, Tanaka T, Bando Y, Mitrovic B, & Yoshida S (2013). In vivo analysis of kallikrein-related peptidase 6 (KLK6) function in oligodendrocyte development and the expression of myelin proteins. Neuroscience, 236, 1–11. 10.1016/j.neuroscience.2012.12.073 [DOI] [PubMed] [Google Scholar]
  26. Ozdemir D, Meyer J, Kieffer BL, & Darcq E (2024). Model of negative affect induced by withdrawal from acute and chronic morphine administration in male mice. Scientific Reports, 14(1), 9767. 10.1038/s41598-024-60759-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Partridge KJ, & Hillhouse TM (2024). Differential effects of repeated fluoxetine and ketamine administration on behavioral and pharmacological stressor-induced depression of digging behavior in mice. Experimental and Clinical Psychopharmacology, 32(5), 506–517. 10.1037/pha0000711 [DOI] [PubMed] [Google Scholar]
  28. Pereira TD, Aldarondo DE, Willmore L, Kislin M, Wang SS-H, Murthy M, & Shaevitz JW (2019). Fast animal pose estimation using deep neural networks. Nature Methods, 16(1), 117–125. 10.1038/s41592-018-0234-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Pereira TD, Tabris N, Matsliah A, Turner DM, Li J, Ravindranath S, Papadoyannis ES, Normand E, Deutsch DS, Wang ZY, McKenzie-Smith GC, Mitelut CC, Castro MD, D’Uva J, Kislin M, Sanes DH, Kocher SD, Wang SS-H, Falkner AL, … Murthy M (2022). SLEAP: A deep learning system for multi-animal pose tracking. Nature Methods, 19(4), 486–495. 10.1038/s41592-022-01426-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Pergolizzi JV Jr, Raffa RB, & Rosenblatt MH (2020). Opioid withdrawal symptoms, a consequence of chronic opioid use and opioid use disorder: Current understanding and approaches to management. Journal of Clinical Pharmacy and Therapeutics, 45(5), 892–903. 10.1111/jcpt.13114 [DOI] [PubMed] [Google Scholar]
  31. Pomrenze MB, Pinto DFC, Neumann PA, Llorach P, Tucciarone JM, Morishita W, Eshel N, Heifets BD, & Malenka RC (2022). Modulation of 5-HT release by dynorphin mediates social deficits during opioid withdrawal. Neuron, 110(24), 4125–4143.e6. 10.1016/j.neuron.2022.09.024 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Pond HL, Heller AT, Gural BM, McKissick OP, Wilkinson MK, & Manzini MC (2021). Digging behavior discrimination test to probe burrowing and exploratory digging in male and female mice. Journal of Neuroscience Research, 99(9), 2046–2058. 10.1002/jnr.24857 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Rokach A (2005). Drug Withdrawal and Coping With Loneliness. Social Indicators Research, 73(1), 71–85. 10.1007/s11205-004-2008-y [DOI] [Google Scholar]
  34. Rothwell PE, Thomas MJ, & Gewirtz JC (2009). Distinct Profiles of Anxiety and Dysphoria during Spontaneous Withdrawal from Acute Morphine Exposure. Neuropsychopharmacology : Official Publication of the American College of Neuropsychopharmacology, 34(10), 2285–2295. 10.1038/npp.2009.56 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Smith CJW, Wilkins KB, Mogavero JN, & Veenema AH (2015). Social Novelty Investigation in the Juvenile Rat: Modulation by the μ-Opioid System. Journal of Neuroendocrinology, 27(10), 752–764. 10.1111/jne.12301 [DOI] [PubMed] [Google Scholar]
  36. Substance Abuse and Mental Health Services Administration. (2024). Key Substance Use and Mental Health Indicators in the United States: Results from the 2023 National Survey on Drug Use and Health. [Google Scholar]
  37. Swain Y, Muelken P, Skansberg A, Lanzdorf D, Haave Z, LeSage MG, Gewirtz JC, & Harris AC (2020). Higher anhedonia during withdrawal from initial opioid exposure is protective against subsequent opioid self-administration in rats. Psychopharmacology, 237(8), 2279–2291. 10.1007/S00213-020-05532-W [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Thomas A, Burant A, Bui N, Graham D, Yuva-Paylor LA, & Paylor R (2009). Marble burying reflects a repetitive and perseverative behavior more than novelty-induced anxiety. Psychopharmacology, 204(2), 361–373. 10.1007/s00213-009-1466-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Truong V, Moore JE, Ricoy UM, & Verpeut JL (2024). Low-Cost Approaches in Neuroscience to Teach Machine Learning Using a Cockroach Model. Eneuro, 11(12), ENEURO.0173–24.2024. 10.1523/ENEURO.0173-24.2024 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Umathe SN, Mundhada YR, & Bhutada PS (2012). Differential effects of acute morphine, and chronic morphine-withdrawal on obsessive–compulsive behavior: Inhibitory influence of CRF receptor antagonists on chronic morphine-withdrawal. Neuropeptides, 46(5), 217–221. 10.1016/j.npep.2012.06.002 [DOI] [PubMed] [Google Scholar]
  41. Vanderschuren LJ, De Vries TJ, Wardeh G, Hogenboom FA, & Schoffelmeer AN (2001). A single exposure to morphine induces long-lasting behavioural and neurochemical sensitization in rats. The European Journal of Neuroscience, 14(9), 1533–1538. 10.1046/j.0953-816x.2001.01775.x [DOI] [PubMed] [Google Scholar]
  42. Velasco B, Mohamed E, & Sato-Bigbee C (2021). Endogenous and exogenous opioid effects on oligodendrocyte biology and developmental brain myelination. Neurotoxicology and Teratology, 86, 107002. 10.1016/j.ntt.2021.107002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Verpeut JL, Bergeler S, Kislin M, William Townes F, Klibaite U, Dhanerawala ZM, Hoag A, Janarthanan S, Jung C, Lee J, Pisano TJ, Seagraves KM, Shaevitz JW, & Wang SS-H (2023). Cerebellar contributions to a brainwide network for flexible behavior in mice. Communications Biology, 6(1), 1–17. 10.1038/s42003-023-04920-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Volkow ND, Baler RD, & Goldstein RZ (2011). Addiction: Pulling at the Neural Threads of Social Behaviors. Neuron, 69(4), 599–602. 10.1016/j.neuron.2011.01.027 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Wesson DW (2013). Sniffing Behavior Communicates Social Hierarchy. Current Biology, 23(7), 575–580. 10.1016/j.cub.2013.02.012 [DOI] [PubMed] [Google Scholar]
  46. Yalçın B, Pomrenze MB, Malacon K, Drexler R, Rogers AE, Shamardani K, Chau IJ, Taylor KR, Ni L, Contreras-Esquivel D, Malenka RC, & Monje M (2024). Myelin plasticity in the ventral tegmental area is required for opioid reward. Nature, 630(8017), 677–685. 10.1038/s41586-024-07525-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Yoon H, Triplet EM, Simon WL, Choi C-I, Kleppe LS, De Vita E, Miller AK, & Scarisbrick IA (2022). Blocking Kallikrein 6 Promotes Developmental Myelination. Glia, 70(3), 430–450. 10.1002/glia.24100 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Zhang S, Zhu M, Lan Z, & Guo F (2025). Transcription factor 7-like 2 (TCF7l2) regulates CNS myelination separating from its role in upstream oligodendrocyte differentiation. Journal of Neurochemistry, 169(1), e16208. 10.1111/jnc.16208 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

MMC2
NIHMS2189816-supplement-MMC2.xlsx (61KB, xlsx)
MMC1
NIHMS2189816-supplement-MMC1.docx (3.3MB, docx)

RESOURCES