THE EFFECTS OF BUPRENORPHINE AND MORPHINE DURING PREGNANCY: IMPACT OF EXPOSURE LENGTH ON MATERNAL BRAIN, BEHAVIOR, AND OFFSPRING NEURODEVELOPMENT

Abstract

The escalating incidence of opioid-related issues among pregnant women in the United States underscores the critical necessity to understand the effects of opioid use and Medication for Opioid Use Disorders (MOUDs) during pregnancy. This research employed a translational rodent model to examine the impact of gestational exposure to buprenorphine (BUP) or morphine on maternal behaviors and offspring well-being. Female rats received BUP or morphine before conception, representing established use, with exposure continuing until postnatal day 2 or discontinued on gestational day 19 to mimic treatment cessation before birth. Maternal behaviors – including care, pup retrieval, and preference – as well as hunting behaviors and brain neurotransmitter levels were assessed. Offspring were evaluated for mortality, weight, length, milk bands, surface righting latency, withdrawal symptoms, and brain neurotransmitter levels. Our results reveal that regardless of exposure length (i.e., continued or discontinued), BUP resulted in reduced maternal care in contrast to morphine-exposed and control dams. Opioid exposure altered brain monoamine levels in the dams and offspring, and was associated with increased neonatal mortality, reduced offspring weight, and elevated withdrawal symptoms compared to controls. These findings underscore BUP’s potential disruption of maternal care, contributing to increased pup mortality and altered neurodevelopmental outcomes in the offspring. This study calls for more comprehensive research into prenatal BUP exposure effects on the maternal brain and infant development with the aim to mitigate adverse outcomes in humans exposed to opioids during pregnancy.

1. Introduction

Opioid use in the United States has rapidly increased over the last two decades and reached epidemic proportions (Krans & Patrick, 2016). Numbers are particularly concerning for pregnant women diagnosed with Opioid Use Disorder (OUD), as the prevalence has quadrupled from 1999-2014 (Haight et al., 2018; SAMHSA, 2021) with the rate of maternal opioid-related diagnosis at delivery estimated to be at 8.2 per 1000 delivery hospitalizations in 2017 (Hirai et al., 2021). Misuse of opioids during pregnancy and prenatal opioid exposure are associated with significant short- and long-term health issues for both mother and child such as preterm labor, stillbirth, Neonatal Opioid Withdrawal Syndrome (NOWS), and other complications (Hirai et al., 2021; SAMHSA, 2021). To prevent adverse effects of OUD, the American College of Obstetricians and Gynecologists (ACOG) recommends Medications for Opioid Use Disorder (MOUD) as the first line of treatment for pregnant women addicted to opioids (ACOG, 2017). This form of therapy has been shown to result in improved infant outcomes compared to infants born to mothers suffering from OUD who did not receive/use MOUDs (Binder & Vavrinkova, 2008; Kandall et al., 1976; Maas et al., 1990). While these treatments improve infant outcomes, the potential adverse effects on the physical and neurodevelopmental aspects of offspring necessitate further investigation (Byrnes & Vassoler, 2018; Lee et al., 2020).

The two most commonly prescribed MOUDs are the synthetic opioid methadone and semi-synthetic opioid buprenorphine (BUP). BUP, a partial mu-opioid receptor agonist and kappa-opioid receptor antagonist, was approved by the FDA in 2002 for the treatment of opioid dependence (Hudak & Tan, 2012) and has since replaced methadone as the gold standard for MOUD during pregnancy due to better infant health outcomes (Brogly et al., 2014; Gaalema et al., 2012; Hall et al., 2016; Jones et al., 2010; Lemon et al., 2018; Zedler et al., 2016). These improved outcomes may be due to BUP being less lipophilic and unable to readily cross placental barriers (Farid et al., 2008; Kraft & van den Anker, 2012). Despite these preferable findings, in utero exposure to BUP has been associated with potential neonatal harm (e.g., central nervous system stress, poor self-regulation, and other NOWS symptomology), indicating a need for a more comprehensive understanding of its effects (Jansson et al., 2017; Velez et al., 2018).

The current literature focuses predominantly on infant outcomes, leaving a gap in our understanding of BUP's effects on maternal behaviors and mother-infant interactions—factors crucial for offspring development (Wallin et al., 2021). A woman's brain is uniquely sensitive to the influences of exogenous opioids during her transition to motherhood due to the endogenous opioid regulation of the Maternal Brain Network (MBN) (Wallin et al., 2021). This network is largely responsible for the onset and maintenance of maternal care behavior. Therefore, using MOUDs during pregnancy could disrupt the endogenous opioid regulation of the MBN (Wallin et al., 2021). Specifically, BUP’s mechanisms of action may be responsible for the unique effect on the maternal brain, as the interference of BUP at the mu and kappa opioid receptors may disrupt the initiation of maternal care behaviors during the transition to motherhood.

Many existing preclinical models fail to replicate the chronic nature of opioid exposure seen in humans, as women often enter pregnancy already using opioids and continue their use. Our study addresses this gap with a translational animal model that commences BUP exposure before conception and continues postpartum, more closely mirroring human patterns. For decades, rodent studies have demonstrated that morphine exposure during gestation can significantly impair maternal behaviors (Bridges & Grimm, 1982; Chen et al., 2015; Slamberova et al., 2001; Sobor et al., 2010) but less is known about the impact of BUP on the maternal brain and subsequent maternal behavior. Our laboratory's previous findings indicate that BUP treatment is associated with diminished maternal caregiving and increased pup mortality (Wallin et al., 2019). In particular, our prior research found that female rats administered BUP at both high (1.0 mg/kg) and low (0.3 mg/kg) doses starting seven days before mating and continuing through pregnancy displayed reduced maternal care compared to control groups. Offspring from the low-dose group showed symptoms associated with NOWS, developmental delays, decreased birth weight, lower body temperature and length, and diminished pain sensitivity. Notably, offspring from the high-dose BUP group experienced a high mortality rate by postpartum day 2 (PND2), with maternal neglect being a significant contributing factor (Wallin et al., 2019). Further, many other studies have reported increased rates of pup mortality (Chen et al., 2015; Chiang et al., 2010; Hutchings et al., 1996; Robinson & Wallace, 2001; Wallin et al., 2019), which might be partially due to the reduced maternal care. In line with this, prior preclinical research where cross-fostering with drug-naïve dams took place, did not report such severe outcomes (Belcheva et al., 1994; Evans et al., 1988; Hung et al., 2013; Hutchings et al., 1995; Kongstorp et al., 2020).

The present study builds upon our previous work (Wallin et al., 2019) by examining the impact of the timing of opioid exposure—either continuous through parturition or discontinued during pregnancy—on maternal behavior and offspring viability. By introducing a comparative morphine condition, we aim to more closely simulate OUD in humans.

Opioid administration was discontinued on gestational day 19 for some groups. Although discontinuation prior to birth is not directly clinically translational, it allows us to examine the relationship between the presence of opioids at birth and subsequent adverse effects. We hypothesized that continuous exposure to BUP or morphine will result in greater maternal care deficits and more severe developmental challenges in offspring, including heightened NOWS symptoms, compared to offspring from dams where opioid treatment was discontinued prior to parturition.

2. Methods

2.1. Subjects and Study Overview

Female (n=50) and male (n=10) Sprague-Dawley rats weighing 200-250g were purchased from Charles River Laboratory, Wilmington MA. Animals were kept in a 12-hour light/dark cycle (lights on at 6:00AM) with consistent humidity (~55%) and temperature (~22°C). Purina Rodent Chow and water were available ad libitum. Animals were habituated to the colony room for 4 days prior to the beginning of the study, followed by an additional 3 days of handling. All procedures were approved by the Wayne State University Institutional Animal Care and Use Committee (Protocol #20-09-2744) and were in accordance with the ARRIVE guidelines and the National Research Council's Guide for the Care and Use of Laboratory Animals.

Female rats were assigned pseudo-randomly across five experimental conditions. Treatments consisted of buprenorphine (BUP) at a concentration of 1.0 mg/kg (CASRN: 161772-95-8), morphine sulfate at concentrations ranging from 3.0 to 6.0 mg/kg (CASRN: 80573-75-7), or a saline control at 1.0 ml/kg. Initiation of treatment commenced seven days before mating, referred to as Preconception Day 1 (PC1). This schedule was designed to emulate typical human patterns of opioid consumption. Depending on the group, drug administration continued through to postnatal day 2 (PND2) or was discontinued and transitioned to saline on gestational day 19 (GD19) to examine effects stemming from the cessation of drug exposure prior to birth. Behavioral assessments and maternal care observations were systematically conducted for all dams and their litters from birth until PND2, with these procedures being detailed in subsequent sections 2.4 & 2.5. The study concluded with the humane euthanasia of the animals on PND2 (see Figure 1).

Fig 1. Experimental timeline from Preconception (PC) through Postnatal Day (PND) 2.

Female rats received either vehicle (VEH), buprenorphine (BC, BD) or morphine (MC, MD) until Gestational day (GD) 19 (discontinued (D)) or until PND 2 (continued (C)). Rats underwent a series of behavioral tests during preconception and the postpartum. Rx: drug treatment; ↑Rx: increase drug dose for morphine; Sac: sacrifice.

2.2. Drug Preparation and Administration

BUP and morphine were sourced from the NIDA drug supply program. We prepared liquid solutions of each drug at 1.0 mg/ml concentration through serial dilution, ensuring uniform injection volumes for all treatment groups. Injection volumes were adjusted for each dam according to individual weight changes every 4-5 days and after parturition. The solutions were administered subcutaneously (s.c.), alternating between the right and left sides above the hip on the dorsal posterior surface. Due to quick acquisition of tolerance, morphine dosage increased from 3.0 mg/kg to 6.0 mg/kg (final dose) by 1.0 mg/kg increments every 7-8 days (Chiang et al., 2010). Further, due to its short half-life, morphine was administered 2x/day (b.i.d.; 08:00 h - 09:00 h and 16:00 h - 17:00 h). To ensure all animals received the same handling experience, the BUP and vehicle groups received a saline injection in the evening. The BUP dose was chosen based on our previous work showing maternal care deficits with 1 mg/kg and to represent an approximate human dose equivalent (15mg/day (Martin et al., 2020)) which equals ~ 0.2mg/kg (based on 70kg body weight) multiplied by * 6 to account for increased metabolism in rats (Nair & Jacob, 2016)). Morphine dose was based on previous literature showing maternal care deficits at those doses (Chen et al., 2015; Chiang et al., 2010).

Adult female rats were assigned to one of the following five groups: saline (“VEH”, 1.0 ml/kg; b.i.d), BUP continuous (“BC”, 1.0 mg/kg; continued until PND2), BUP discontinued (“BD”, 1.0 mg/kg; switched to saline injections on GD19), morphine continuous (“MC”, 3-6.0 mg/kg; b.i.d; continued until PND2), or morphine discontinued (“MD”, 3-6.0 mg/kg; b.i.d; switched to saline injections on GD19) (see Figure 1).

2.3. Breeding Procedure

Seven days after the start of injections, female rats were mated with drug-naïve males until a vaginal lavage confirmed pregnancy. Successful conception was defined as GD0, and pregnant rats were housed with their cage mate until GD7 to mitigate stress from single housing. Rats were then single housed throughout the rest of gestation and postpartum to allow for unique litter identification.

2.4. Maternal and Offspring Measurements

Maternal characteristics included time to conceive, gestational length, and gestational weight gain (recorded on GD0, GD7, GD14, and then daily from GD19-PND2). Saphenous vein blood was collected from dams on GD7 to determine corticosterone (CORT) levels. Nest quality was scored on a scale from 0-4 from GD20-PND2 based upon the level of nesting proficiency of arranging bedding materials into a localized area of the cage and forming an enclosed nest edge (Brancato & Cannizzaro, 2018) (see Figure S1 in supplemental data for detailed scoring criteria and photographic examples).

On PND0, dams were monitored for signs of placentophagia and cannibalism. Maternal care behaviors were video recorded and subsequently observed by a trained experimenter in the colony room during the light cycle. These videos were scored in 1-min bins for 30 min each, 4x/day on PND0 and PND1, and 2x/day on PND2 with at least 60-min between observation sessions. All observations were performed at approximately the same time each day between 8:00AM and 3:00PM. The percentage of time a dam spent engaged in a certain behavior was calculated for each day. Time spent in key behaviors was evaluated regardless of multitasking (e.g., licking/grooming while active nursing was coded as "licking/grooming" and "active nursing"). Maternal care behaviors were divided into two categories for analysis: pup-directed behaviors (i.e., licking/grooming, active/passive nursing, time on nest) and non-pup directed behaviors (i.e., time off the nest, time sleeping off nest).

Litters were assessed for mortality and the presence of milk bands in each pup each morning from PND0-2. On PND1 litters were sexed, weighed, measured for body length, pseudo-randomly culled to 8 pups per litter (n=4 males, n=4 females if possible), and all remaining pups were assessed for NOWS symptoms (n=603 total before culling, n= 338 after culling). A total of 8 litters did not have enough of one sex to cull down to a 4:4 ratio and therefore had a different ratio of males to females after the culling procedure. On PND2, offspring were again weighed, their body length measured, and they underwent a surface righting task (see below).

2.5. Maternal and Offspring Behavioral Tests

2.5.1. Hunting (Predatory) Behavior

A hunting task was performed using live crickets (Fluker Farms, Louisiana: 2w old, 1/4 inch) on PC7 (virgin females; baseline) and repeated on PND2 (maternal hunting) to investigate behavioral selection (i.e., hunting vs. caregiving). Crickets were purchased online from a cricket supplier (Fluker Farms, Louisiana: 2w old, 1/4 inch). Rats were transported from the colony room to the laboratory and habituated for 30 min before testing began. Crickets (N=5) were placed into the home cage, and rats were individually observed for predatory insect-hunting for 5 min (Klein et al., 2014; Sukikara et al., 2006). On PND2, pups were kept in the cage during this task as opioids have been shown to shift behavior selection toward non-maternal behaviors (i.e., hunting) (Cruz Ade et al., 2010). Predatory hunting behavior was coded as (i) latency to capture each insect; (ii) frequency of insect consumption; and (iii) the number of insects captured. Any insects not captured by the rats at the end of the 5-min tests were removed from the home cage by an experimenter.

2.5.2. Olfaction Preference Test

Dams were subjected to a two-choice olfaction test on PND1 to evaluate preference or aversion for pup odors (i.e., pup-soiled bedding). To begin, dams were placed into the center division of a two-choice apparatus with a clear plastic top. One side contained pup-soiled bedding from their home cage, while the other contained new clean bedding. The bedding type was counterbalanced for each side of the apparatus between dams (Kinsley & Bridges, 1990). Dams were habituated to the apparatus for 5-min during late gestation to prevent novelty stress. Time spent on each side and the number of crossings were recorded for a total of 5-min (300sec) and a preference was calculated by setting 150sec (i.e. 50% of the test time) as the “no preference” (i.e., chance) point and calculating how much more time each dam spent on either the home (soiled) or clean side.

2.5.3. Pup Retrieval

Pup retrieval has been established as a valid marker of maternal care and interaction evaluation (Gonzalez, 2001; Pardon et al., 2000). On PND2, the dam was briefly removed while the pups were spread across the cage at approximately equal distances outside the nest. A researcher counted the number of pups a dam returned to her nest and the latency to return each pup during a 15-min period. Any pups not retrieved within the 15-min were carefully placed back in the nest by the researcher.

2.5.4. Neonatal Opioid Withdrawal Syndrome

On PND1, each litter was observed for NOWS symptoms. Pups were placed into holding bins with separated compartments, with one pup per compartment. Observations of withdrawal behaviors were done in increments of 30-sec for 15-min, alternating between males and females. Behaviors included hyperactivity, tremors, vocalization, mouth movements, stretching, and “face-washing” (Robinson & Wallace, 2001) using a tally-system. The whole group (females or males) were given a score of “yes” or “no” if any one of the four pups exhibited the withdrawal behavior within the 30-sec window. Individual withdrawal behaviors were then aggregated into a simple summative composite score for all withdrawal behaviors for each litter and analyzed as percent different from VEH scores.

2.5.5. Surface Righting

The latency (sec) for the pup to return itself to a prone (i.e., upright on paws) position after being placed in a supine position (i.e., on its back) was measured. Any pups unable to right themselves after 60-sec were gently turned upright on their paws.

2.5.6. Corticosterone Assay

Blood collected from the saphenous vein (GD7), or the trunk (sacrifice) was centrifuged, and plasma was extracted and frozen at −20°C until analysis. Serum samples were analyzed for corticosterone levels (CORT) using enzyme-linked immunosorbent assay (ELISA) assays (K014-H5) from Arbor Assays according to the manufacturer’s instructions. All samples were run in duplicate, and the inter-assay coefficient was less than 10%.

2.5.7. High-Pressure Liquid Chromatography (HPLC)

Immediately after the pup retrieval assessment on PND2, both dams and pups were euthanized via rapid live decapitation. Whole brains were removed and placed on dry ice. The dam brain was sliced into 3mm coronal slices (pups: 2mm) followed by use of a 2mm (pups: 1mm) hole punch to isolate specific brain regions (Dams: Ventral Tegmental Area “VTA”, Nucleus Accumbens “NAc”, Medial Preoptic Area “mPOA”, and Periaqueductal Gray “PAG”; Pups: Hippocampus, Hypothalamus, Prefrontal Cortex “PFC”). Each tissue punch was weighed and sonicated in 50μl (pups: 25μl) of 0.2N perchloric acid for 3-5 seconds and centrifuged. The resulting supernatant was collected for analysis (as previously described in (Davidson et al., 2022). The samples were analyzed using a Dionex Ultimate 3000 HPLC system's autosampler at 5°C (Thermo Fisher Scientific, Waltham, MA). Standards for dopamine (DA), serotonin (5-HT), and norepinephrine (NE), as well as metabolites 3,4-dihydroxyphenylacetic acid (DOPAC), homovanillic acid (HVA), 3-methoxytyramine (3-MT), and 5-hydroxyindoleacetic acid (5-HIAA), were prepared using serial dilution, and run before and after each tissue analysis. A detection threshold of 2 times baseline was set, and standard curve equations were created for all monoamines and metabolites. Tissue levels of monoamines and metabolites were expressed as ng (neurochemical)/mg (wet tissue weight). DOPAC turnover was calculated by dividing DOPAC/DA and HVA/DA, which was used to determine the rate at which DA was metabolized in the brain, assuming no other pathways were involved.

2.6. Data Analysis

2.6.1. Maternal Data

Maternal body weight, behaviors, latency, preference scores, litter size, and litter sex ratio were analyzed using either a one-way, repeated measure (RM) or covariate ANOVA (ANCOVA) followed up with Tukey’s post-hoc as appropriate (i.e., if main or interaction effects were significant). LSD post-hoc analysis was used for repeated analyses with covariates. Frequency counts or ordinal scoring (i.e., cricket hunting, nesting scores, pups grouped, etc.) were analyzed using Chi-Square and followed up by Kruskal-Wallis Test for non-parametric data. Total monoamine and metabolite concentrations (ng/mg) for NE, 5HT, DA, DOPAC, 3MT, HVA, and HIAA were analyzed using separate one-way ANOVAs for each brain region of interest (mPOA, PAG, VTA, NAc).

Treatment Groups (BC, BD, MC, MD, and VEH) were used as initial between-subject factors for all analyses after discontinuation (GD19). For all analyses prior to GD19, Opioid Type (BUP vs. morphine vs. VEH) was used as the initial between-subject factor because BC/BD and MC/MD groups had the same treatment regimen up until GD19. A significant main effect of Treatment Group or Opioid Type (prior to GD19) was reported for all analyses when identified. If there was no significant main effect in the initial analysis, up to 3 follow-up analyses were conducted due to our a priori hypotheses that (a) BUP, (b) continuously treated, or (c) dams exposed to any opioid would show deficits in maternal care, behavior, and monoamine concentration relative to VEH dams. Due to these a priori hypotheses, if a significant result was not identified by Treatment Group, we evaluated the same test by Opioid Type (i.e., BUP vs. morphine vs. VEH; after GD19) as a follow-up. If there was no Opioid Type effect, we evaluated by Treatment Length (continued vs. discontinued vs. VEH), and if there was no effect of Treatment Length, we analyzed by Drug (i.e., any Opioid exposure vs. VEH). In brief:

Initial Analysis: by Treatment Group (BC vs. BD vs. MC vs. MD vs. VEH) *

• Follow-up 1: by Opioid Type (BUP vs. morphine vs. VEH) between-subject factors

• Follow-up 2: by Treatment Length (Continuous vs. Discontinued vs. VEH)

• Follow-up 3: by Drug (i.e., Opioid exposure vs. VEH)

*For measures before GD19 (discontinuation day): analysis started at “Opioid Type”.

2.6.2. Offspring Data

Data was collected from each individual pup. To avoid litter effects, data was averaged per sex for each litter (i.e., one mean score for males and one mean score for females for each measure). Further, PND1 and PND2 data were analyzed separately since PND1 data included all pups (whole litter before culling) while PND2 data only included n=8 pups per litter (after culling).

Pup mortality was analyzed using a Mann-Whitney U Test (for comparison between 2 groups) or Kruskal Wallis Test (for comparison between 3+ groups) due to non-normality of the data). Offspring weight, length, and milk bands were analyzed using ANCOVA while controlling for litter size. NOWS symptoms and surface righting data were analyzed using a 2-way ANOVA with Treatment Group and Sex as between-subject factors. If there was no significant main effect at the level of Treatment Group, follow-up analyses and post-hoc tests were conducted as described above. All data was analyzed with SPSS Statistics (version 26) and all figures were graphed with Graph Pad Version 8. Data are presented as Mean +\− SEM and results were considered significant if p < 0.05.

3. Results

3.1. Litter Exclusion and Reassignment

Of the fifty female rats received, seven were excluded from analysis at some point due to unexpected events including one unexpected death (n =1; VEH, prior to mating), high rates of pup mortality (n=1 BC), or failure to give birth (n=1 BD, n=1 MD, n=3 MC). Rats that failed to give birth had visible sperm in their vaginal lavage samples, suggesting that they did not carry their pregnancies to term (i.e., may have experienced an early spontaneous abortion). To make up for the reduced sample size in the MC group, one dam that was originally assigned to the MD group was reassigned to the MC group before drug discontinuation. The total number of litters born and included in the analysis per experimental group are as follows: BC (N = 9), BD (N = 9), MC (N = 8), MD (N = 8), and VEH (N = 9).

3.2. Maternal Data

3.2.1. Pre-conception Characteristics and Behavior

Weight gain during preconception was first analyzed using RM ANOVA on PC0, PC4, and PC7 (with PC0 as a covariate) with a between-subject factor of Opioid Type. There were no main effects or interactions between PC Day and Opioid Type (all p’s > 0.05) and there were no significant effects for the follow-up analyses (all p’s > 0.05; see Figure 2). Days to conceive were not significant at the level of Treatment Group (p’ > 0.05). However, there was a significant main effect at the level of Drug [F(1, 42) = 4.125, p = 0.049], with opioid-exposed dams taking longer to conceive compared to controls (see Figure 3A).

Fig 2: Overview of body weight changes across all groups from preconception (PC) to the postpartum (PND).

For a more detailed graph of each time period with all significant differences, please see Supplementary Figure S2. Only significant differences between each opioid group (i.e. BC(*), BD(#), MC($), MD (@) and VEH are indicated in this graph. There was no significant difference between treatment exposures before conception. During gestation, BC and BD dams weighed less than morphine dams on GD0, 7 and 14 and BUP dams weighed less than VEH dams on GD14 and 19. During late gestation (i.e. GD 20 and 21) BC, BD, and MD dams weighed less than MC and VEH dams. During the postpartum period (i.e. PND0-2) BC and BD weighed less than VEH dams on all days, while MC weighed less than VEH dams on PND1-2, MD weighed less than VEH dams on PND1, and BD weighed less than MD on PND1-2 (For more details see supplemental Fig S2D). Data are shown as mean ± SEM. *(BC), #(BD), $(MC), @(MD) <0.05 vs VEH, respectively. BC: buprenorphine continued, BD: buprenorphine discontinued, MC: morphine continued, MD: morphine discontinued, VEH: vehicle (saline), PC: Preconception, GD: Gestational Day, PND: Postnatal Day.

Figure 3: Maternal behaviors and physiology.

A: Days to conceive. Shown are the effects of BUP, morphine and vehicle on days to conceive in milliparous female rats (reflected by the number of days between first mating exposure and positive spenn identification). There was a significant effect of Opioid (collapsed across treatment groups) demonstrating opioid exposure delayed time to conception.

B: Serum levels of corticosterone (CORE, ng/ml) were significantly decreased in BUP dams on gestational day (GD) 7 compared to VEH dams (*p < 0.05) and morphine dams (#p < 0.05).

C: Placentophagia was assessed by the total number of placentas the dam did not consume. BD dams consumed fewer placentas after birth than VEH dams (*p < 0.05).

D: Pup directed maternal behavior (presented as percent of time the dam spent in the specified behavior) on postnatal day (PND) 0, 1 and 2. On PND1, BUP (BC, BD) dams spent significantly less time in pup-directed behaviors compared to VEH dams (*p < 0.05). There were no significant differences on PND0 or PND2.

E: Nesting behavior was coded ordinally (See Suppl. Fig S1) every day from GD20-PND2. On GD20. BC & BD dam nest scores were lower compared to VEH (*), MC (#), and MD ($) scores. Similarly, MC & MD were also lower compared to VEH dams(*). BC and BD continued to display less evolved nests compared to VEH (*) until PND2 and compared to MC and MD at least until PND 0 (#,$) (*,#,$ p < 0.05).

All data are shown as mean ± SEM and as individual data points. BUP: buprenorphine, BC: buprenorphine continued, BD: buprenorphine discontinued, MC: morphine continued, MD: morphine discontinued, VEH: vehicle (saline), GD: Gestational Day, PND: Postnatal Day.

Chi-Square analysis revealed a significant effect of Opioid Type on the number of crickets caught in preconception (baseline; max: 5 crickets) [X²(8, N = 49) = 16.666, p = 0.034] with morphine rats capturing more crickets in a 5-min period as compared to BUP and VEH rats (Mann-Whitney: p = 0.041 and 0.014, respectively) (see Table 1). Hunting latency was not significant (all p’s > 0.05).

Table 1:

Dam and Offspring Measurements

Measurement	VEH (n=9)	BC (n=9)	BD (n=8)	MC (n=8)	MD (n=9)
Gestational length	22.5 ± 0.11	22.6 ± 0.1	22.44 ± 0.19	22.38 ± 0.13	22.44 ± 0.11
Litter Size	14.00 ± 0.94	14.00 ± 0.65	14.11 ± 0.61	13.50 ± 1.41	12.88 ± 1.34
Number males born	8.22 ± 1.01	6.80 ± 0.57	6.11 ± 0.84	7.88 ± 0.81	6.50 ± 0.91
Number females born	5.78 ± 0.92	6.50 ± 0.79	7.78 ± 0.91	5.63 ± 0.78	6.38 ± 1.00
PND0 mortality	0.11 ± 0.11	1.80 ± 0.92	0.89 ± 0.39	0.75 ± .42	0.25 ± 0.25
PND0 male mortality	0.00 ± 0.00	0.70 ± 0.34	0.67 ± 0.37	0.13 ± 0.13	0.13 ± 0.13
PND0 female mortality*	0.11 ± 0.11	1.00 ± 0.52	0.11 ± 0.11	0.63 ± 0.38	0.13 ± 0.13
Total mortality	0.22 ± 0.15	3.10 ± 1.44	2.33 ± 0.80	0.88 ± 0.40	1.13 ± 0.58
PND0 milk bands	13.00 ± 1.25	12.25 ± .62	10.00 ± 1.65	13.00 ± 1.35	11.50 ± 1.85
PND1 milk bands	13.44 ± 1.00	12.22 ± 0.74	11.78 ± 0.43	12.25 ± 1.08	11.50 ± 1.21

Note: All data is presented as Mean ± SEM. N is based on litter number per group. *BUP treatment resulted in higher female mortality as compared to VEH (p = 0.041).

3.2.2. Gestational Characteristics

Weight gain during gestation was analyzed using RM ANCOVA while controlling for litter size. For GD0-19 weight, there was a significant main effect of Opioid Type [F(2, 40) = 3.339, p = 0.046] and a significant interaction of Day of Gestation x Opioid Type [F(6, 78) = 3.309, p = 0.006], with BUP dams weighing less than morphine dams on GD0, 7, and 14 and less than VEH dams on GD14 and 19 (LSD post hoc: p's < 0.05). For GD20 and GD21 weight, there was a significant main effect of Treatment Group [F(4, 37) = 4.344, p = 0.006] such that BC, BD, and MD dams weighed less than MC and VEH dams on both days (Tukey post-hoc: p's < 0.05) (see Figure 2 and Supplemental Figure S2).

Postpartum dam weight (PND0-2) revealed a main effect of Treatment Group [F(4, 37) = 4.998, p = 0.003] with BC and BD weighing less than VEH, MC weighing less than VEH, MD weighing less than VEH only on PND1, and BD weighing less than MD on PND1-2 (Tukey post hoc: all p's < 0.05) (see Figure 2). No significant results were found for Treatment Group or other follow-up analyses for Gestational Length (all p's > 0.05).

Blood serum samples collected on GD7 were analyzed for CORT levels (ng/ml) using a one-way ANOVA. There was a significant effect of Opioid Type [F(2,27) = 5.923, p = 0.007] with BUP dams exhibiting lower levels of CORT than morphine and VEH dams (Tukey post-hoc: p's < 0.05, see Figure 3B).

3.2.3. Peri- and Postpartum Behavior

Though a one-way ANOVA did not reveal a main effect of Treatment Group on the number of placentas unconsumed on PND0 [F(4, 39) = 2.590, p = 0.051], due to a priori hypothesis we conducted Tukey’s post hoc analysis which revealed that BD dams left more placentas unconsumed after birth than MD dams (Tukey’s post hoc p = 0.048; see Figure 3C).

Percent maternal care behaviors were analyzed using one-way ANOVAs following arcsine transformations. PND1 pup directed and PND1 non-pup directed behaviors trended towards significance [F(4,39) = 2.534, p = 0.055]. Due to the a priori hypothesis that the BUP dams would partake in less maternal care behaviors, Tukey’s post hoc analysis was conducted which revealed that BC and BD dams partook in less pup directed behaviors and more non-pup directed behaviors than VEH on PND1 (BC vs. VEH: p = 0.029); BD vs. VEH: p = 0.016) (see Figure 3D). No significant differences were found on PND0 or PND2.

Chi-Square analyses revealed a significant effect of Treatment Group on nesting scores for GD20 [X²(16, N = 44) = 59.714, p < 0.001], GD21 [X²(16, N = 44) = 45.398, p < 0.001], PND0 [X²(12, N = 44) = 37.198, p < 0.001], PND1 [X²(12, N = 44) = 35.030, p < 0.001] and for PND2 [X²(12, N = 43) = 25.064, p = 0.015]. Kruskal Wallis post hoc tests showed that BC and BD dams had significantly flatter and non-localized nest quality than MC, MD, and VEH from GD20-PND0 (p's < 0.05). On PND1, BD dams had flatter, less localized nests than MC, MD, and VEH dams, while BC dams had flatter nests compared to VEH dams (all p's < 0.05). On PND2, BC & BD dams had flatter and less localized nests than VEH (but not MC or MD) dams (all p's < 0.05). For morphine groups, MC & MD dams had less proficient nests as compared to VEH dams on GD20, but not GD21 or beyond (p's < 0.05, see Figure 3E).

Pup retrieval was analyzed with a one-way ANOVA which revealed a significant effect of Treatment Group on the latency to the first pup [F(4,38) = 3.838, p = 0.010] and latency to complete - defined as the time to retrieve the first 5 pups, as one litter had only 5 pups on PND2 [F(4,38) = 3.025, p = 0.029] (see Figure 4). Tukey’s post hoc test showed that BD dams took longer than VEH dams to retrieve the first pup and complete pup retrieval (p's = 0.023 and 0.026, respectively) with no difference between any other groups (p's > 0.05). For percent of pups returned to the nest, there was a significant effect of Opioid Type (but not Treatment Group) [F(2,39) = 4.169, p = 0.023] with BUP dams retrieving fewer pups than VEH dams (Tukey’s post hoc: p = 0.037, see Figure 4C) with no difference between any other groups (Tukey’s post hoc: p's > 0.05).

Figure 4: Pup retrieval.

Pup Retrieval Latency was scored as the time (in seconds) it took to retrieve the first pup (A) and the last pup (cut off time: 900-sec represented as the dotted line; B) back to the nest. BD dams took longer than VEH dams to retrieve the first and last pup (*p < 0.05). C: Pup Retrieval success was scored as a percent success rate (number of pups returned to the nest / total of pups included). There was a significant effect of drug with BUP dams retrieving fewer pups than VEH dams (*p < 0.05). Data are shown as mean ± SEM and each dam’s individual score. BLIP: buprenorphine, BC: buprenorphine continued, BD: buprenorphine discontinued, MC: morphine continued, MD: morphine discontinued, VEH: vehicle (saline).

For the olfaction test, one-way ANOVAs revealed no main effects of time spent on clean vs. pup-soiled side or the number of crosses (all p’s > 0.05, see Supplemental Figure S3). On PND2, Chi-Square revealed no significant effects on the number of crickets captured, and a one-way ANOVA revealed no significant effects for hunting latency (all p's > 0.05; see Supplemental Figure S4).

3.2.4. Maternal Brain

Monoamine and metabolite (NE, 5HT, DA, DOPAC, HIAA, HVA) concentrations were analyzed with separate one-way ANOVAs for each brain region of interest (PAG, VTA, NAc, and mPOA). Some monoamines and/or metabolites were omitted by region due to not meeting thresholds for detection.

For the PAG, there was a significant effect of Opioid Type [F(2,11) = 4.568, p = 0.036] on DOPAC turnover [F(2,10) = 25.262, p < 0.001] and HIAA [F(2,31) = 3.433, p = 0.045]. BUP and morphine had lower DOPAC and DOPAC turnover than VEH, and morphine had reduced levels of HIAA as compared to VEH (Tukey’s post hoc: p's < 0.05, see Figure 5A). The components of PAG and HVA could not be determined because sample levels did not meet the threshold set for detection.

Figure 5: Monoamine and their metabolites concentrations in the PAG, VTA and NAc.

A: In the PAG, Morphine and BUP treatment were associated with reduced DOPAC and DOPAC turnover as compared to VEH dams (*). Morphine treatment was also associated with reduced levels of HIAA. B: In the VTA, MC treatment was associated with reduced levels of NE as compared to BD treatment (# p < 0.05). Morphine treatment was also associated with reduced levels of NE and HIAA as compared to BUP and VEH dams respectively. C: In the NAc, BD treatment was associated with reduced levels of HIAA as compared to VEII treatment (*p < 0.05). There were no significant effects of treatment in the medial preoptic area (data not shown). Data are shown as mean ± SEM plus each dam’s individual data point. PAG: periaqueductal gray, VTA: ventral tegmental area, NAcc: Nucleus Accumbens. DA: dopamine, 5-HT: serotonin, NE: norepinephrine, DOPAC: 3,4-Dihydroxyphenylacetic Acid, HVA: Homovanillic Acid, 3-MT: 3-Methoxytyramme, HIAA: 5-Hydroxymdoleacetic Acid, BC: buprenorphine continued, BD: buprenorphine discontinued, MC: morphine continued, MD: morphine discontinued, VEH: vehicle (saline).

For the VTA, there was a significant difference in NE concentration between Treatment Group [F(4,23) = 3.395, p = 0.025]. Post hoc results showed BD NE levels exceeded MC NE levels (Tukey’s post hoc: p < 0.05). There was a significant difference in Opioid Type on VTA HIAA [F(2,25) = 3.734, p = 0.038] with morphine exhibiting lower levels of HIAA than VEH dams (Tukey’s post hoc: p < 0.05) (see Figure 5B). Surprisingly, there was no significant difference in DA in the VTA. However, these null effects may be due to relatively few samples meeting the threshold for detection, and components of VTA and 3-MT could not be detected.

Lastly, there was a significant difference in NAc HIAA levels by Treatment Group [F(4,28) = 3.082, p = 0.032] with BD dams having less HIAA than VEH dams (Tukey post hoc: p < 0.05, see Figure 5C).

Although the threshold for detection was met for NE, 5HT, DA, DOPAC, HIAA, and DA/DOPAC turnover, none demonstrated significant differences between opioid type or treatment group (all p’s > 0.05). The components of 3MT, HVA, and HVA turnover could not be determined in the mPOA because sample levels did not meet the threshold for detection.

3.3. Offspring Data

3.3.1. Litter characteristics

Univariate analyses revealed no significant differences for litter size, number of males born, or number of females born (all p’s > 0.05) (see Table 1).

3.3.2. Mortality

Results from the Kruskal Wallis Test did not reach significance at the level of Treatment Group for PND0 mortality, PND0 female mortality, PND0 male mortality, or total mortality (see Table 1). Because male offspring typically have higher mortality than female offspring (Vatten & Skjaerven, 2004), a follow-up analyses for PND0 female mortality and PND0 male mortality was conducted at the level of Opioid Type. Surprisingly, there was a significant main effect for PND0 female mortality [H(2) = 6.399, p = 0.041], with BUP resulting in higher female mortality as compared to VEH. Follow-up analysis for total mortality at the level of Opioid Type was significant [H(2) = 8.032, p = 0.018] with BUP exhibiting higher rates of total mortality as compared to VEH (p = 0.006) (see Figure 6A).

Figure 6: Offspring Mortality and physiology.

A: Mean number of deceased pups from PND0-PND2, Kruskal Wallis Test revealed a significant effect of Opioid Type on Total Mortality, with * BUP pups having a higher rate of morality than VEH (*p < 0.05). B: Presence of milk bands on PND2. BD pups had significantly less milk bands than VEH (*p < 0.05).C: Body weight on PND1 and PND2: BC, BD and MC pups weighed significantly less than VEH (*) and MD (#) pups on PND1 and 2 (*,# p < 0.05). D: Body length on PND1 and PND2. BD body length was significantly smaller than VEH on PND2 (*p < 0.05).

Data are shown as mean ± SEM for each litter’s individual data point.

PND: Postnatal Day, BUP: buprenorphine, BC: buprenorphine continued, BD: buprenorphine discontinued, MC: morphine continued, MD: morphine discontinued, VEH: vehicle (saline).

3.3.3. Milk Bands

Univariate analyses showed a significant difference between Treatment Group for PND2 milk bands [F(4,37) = 3.112, p = 0.026] with Tukey's post-hoc analyses revealing that BD offspring had less milk bands than VEH offspring (p = 0.042) (see Figure 6B). There were no significant effects found for PND0 or PND1 milk bands (all p’s > 0.05; Table 1).

3.3.4. Offspring Weight and Body Length

A 2-way ANCOVA revealed a significant difference between Treatment Group on PND1 weight [F(4,75) = 8.778, p < 0.001] and PND2 weight [F(4,75) = 9.282, p < 0.001]. There was also a main effect of Sex on PND1 weight [F(1,75) = 10.507, p = 0.002] with females weighing significantly less than males. This sex effect was not seen on PND2 (p > 0.05). Post hoc tests revealed that on PND1 and PND2, BC, BD, and MC weighed significantly less than VEH and MD (all p’s < .05; see Figure 6C).

A 2-way ANOVA indicated a main effect of Treatment Group on PND1 body length [F(4,76) = 2.726, p = 0.035]. Tukey’s post hoc analysis revealed that BD’s smaller body length compared to VEH, while trending, did not reach significance (p = 0.051). There was also a main effect of Sex in that females had smaller PND1 body length than males [F(1,76) = 5.005, p = 0.028]. For PND2 body length, there was a main effect of Treatment Group [F(4,76) = 3.392, p = 0.013] with Tukey’s post hoc analysis revealing that BD was smaller than VEH (p = 0.008; see Figure 6D).

3.3.5. Offspring Surface Righting

A 2-way ANOVA revealed no main effect of Treatment Group on surface righting (Figure 7A), but a significant main effect of Sex on righting latency [F(1,76) = 8.531, p = 0.005] with females taking significantly longer to right themselves than males across all groups (Figure 7B). Follow-up analyses also revealed a main effect of Discontinuation [F(1,64) = 5.538, p = 0.022] with the discontinuous group taking significantly longer to right themselves as compared to the continuous group (Figure 7C).

Figure 7: Offspring Behavioral Tests.

A: Surface Righting Reflex. There was no treatment effect on the time it took for pups to right themselves, hut females took significantly longer to right themselves than males (B; *p < 0.05). Further, pups bom to dams that discontinued opioid treatment took significantly longer to right themselves as compared to pups from dams that continued their opioid treatment (C; *p < 0.05). D: Neonatal Opioid Withdrawal Symptoms (NOWS) scores. BC, BD, and MC exhibited significantly more withdrawal behaviors than VEII (*p < 0.05). There was also a significant sex effect (E), with males showing more withdrawal symptoms than females (*p < 0.05). Data are shown as mean ± SEM plus each pup’s or litter’s individual data point. BC: buprenorphine continued, BD: buprenorphine discontinued, MC: morphine continued, MD: morphine discontinued, VEH: vehicle (saline).

3.3.6. Neonatal Opioid Withdrawal Syndrome

A 2-way ANOVA revealed a significant main effect of Treatment Group on NOWS scores [F(4,76) = 6.440, p < 0.001] with Tukey’s post hoc analyses showing that BC, BD, and MC had higher NOWS scores as compared to VEH (all p’s < 0.05; Figure 7D). There was also a main effect of Sex in that males exhibited more withdrawal behaviors than females [F(1,76) = 8.754, p = 0.004; Figure 7E].

3.3.7. High-Pressure Liquid Chromatography: Offspring Neurotransmitters

Univariate ANOVA revealed a significant effect of DOPAC concentration within the PFC at the level of Treatment Group [F(4,74) = 4.062, p = 0.005], with BC (p = 0.007), BD (p = 0.019) and MC (p = 0.028) offspring exhibiting a lower DOPAC concentration in the PFC on PND2 than VEH offspring (see Figure 8). All other neurotransmitter analyses were non-significant.

Figure 8: DOPAC levels in the Prefrontal cortex (PFC) of neonatal offspring.

BC, BD and MC offspring exhibited lower DOPAC levels in the PFC on postnatal day (PND) 2 compared to VEH offspring (*p < 0.05). All other neurotransmitters and metabolites were not significantly different. BC: buprenorphine continued, BD: buprenorphine discontinued, MC: morphine continued, MD: morphine discontinued, VEH: vehicle (saline).

4. Discussion

This study systematically explored the consequences of continuous versus discontinuous buprenorphine (BUP) or morphine administration throughout pregnancy to examine the resulting maternal care behavior and offspring outcomes in a translational model. Our findings revealed that BUP-exposed dams showed impairments in maternal behaviors, nest building, reduced gestational weight gain, decreased placentophagia and compromised pup retrieval behavior compared to controls—factors likely contributing to elevated mortality among BUP-exposed pups. Interestingly, discontinuation of BUP in late gestation was associated with greater maternal impairment compared to continued BUP exposure. Comparatively, discontinuation of morphine treatment had a possible "rescue effect" on maternal behavior and nest building. These findings underscore the complex interplay between MOUD use and maternal-neonatal health and highlights the need to raise awareness of the impact of MOUDs use on mothers, their care behavior, and their subsequent mother-infant bonds.

BUP or morphine exposure prior to conception did not appear to influence body weight. During pregnancy, BUP dams gained less weight until GD19 compared to morphine (4.0-5.0 mg/kg, s.c.) and VEH dams, consistent with previous findings (Chiang et al., 2010; Wallin et al., 2019). Intriguingly, after discontinuation on GD19, the morphine discontinued group showed lower weight compared to VEH dams on GD20-21 but caught up to the MC group by PND0 and matched VEH by PND2, likely due to morphine withdrawal. This is in line with work by Chiang and colleagues who also showed reduced dam weight when morphine (2.0-4.0 mg/kg, s.c.) was discontinued on GD20 (Chiang et al., 2010). Importantly, BUP-discontinued dams did not catch up, weighing nearly 40g less than VEH dams after giving birth (PND0). Given that we controlled for litter size in our analyses and that litter size was not different between any of the groups, smaller litters cannot explain this phenomenon. However, others have reported that BUP is associated with reduced food intake, which may account for the reduced body weight (Chiang et al., 2010; Hutchings et al., 1995; Mori et al., 1982; Robinson & Wallace, 2001), though we did not measure food intake in the current study.

Interestingly, we saw reduced CORT levels in BUP dams on GD7 compared to VEH and morphine dams. Similar reduced levels of CORT have been seen in male rats 1-hr after BUP administration (0.05 or 0.1 mg/kg, s.c., (Goldkuhl et al., 2010)), further suggesting BUP’s influence over baseline hypothalamic-pituitary-adrenal (HPA) axis output. Generally, lower stress hormone levels during pregnancy are seen as protective and thus BUP’s ability to lower CORT levels may contribute to the beneficial outcomes seen in human infants after BUP medication compared to continued misuse of opioids or other MOUDs. However, more research is needed to elucidate the exact impact of BUP on HPA axis function during pregnancy, parturition, and the postpartum period.

We did not observe any strong effects of preconceptional or gestational opioid exposure on dam hunting behavior. Morphine rats caught more crickets and at faster rates compared to VEH or BUP rats before conception, but not in the postpartum period. This is somewhat in line with previous studies that have shown that low doses of morphine (1.0-3.0 mg/kg, s.c.) in dams are associated with an increase in preference for hunting over time spent nursing (Sukikara et al., 2007). However, our rats received a higher morphine dose (6.0 mg/kg, s.c.) in the postpartum period and we did not quantify maternal behavior or nursing during the hunting task which could explain the discrepancies.

Results from the current study revealed that preconceptional opioid exposure delayed time to conception which is in line with a recent human study that reported opioid use delayed the time to conceive (Flannagan et al., 2020). Further, despite not reaching statistical significance, we observed several pregnancy “losses” in opioid exposed dams (1 BUP and 4 morphine-exposed) that had clear sperm on their lavage slides. This suggests that morphine may have influenced the ability for the females to carry to term and resulted in possible spontaneous abortions (i.e., 20% of morphine dams may have lost pregnancies). This aligns with human risk factors of stillbirth, growth restrictions, or preterm birth that have been reported with opioid misuse during pregnancy (CDC, 2023). We also saw reduced body weight and length in BUP and morphine-exposed pups compared to controls and compared to pups from dams that had morphine discontinued before parturition, which is in line with the observed growth restriction seen in opioid-exposed human babies (Smith et al., 2015) and other rodent studies (Robinson & Wallace, 2001; Wu et al., 2014). However, preclinical studies that used shorter exposure times or started exposure during pregnancy often do not report lower body weight at birth after BUP or morphine exposure (Chen et al., 2015; Chiang et al., 2010; Hutchings et al., 1996), suggesting that this effect may be partially attributable to differences in length of exposure time between studies. Importantly, the increased pup mortality in BUP-exposed litters, which is in line with our and other previously conducted studies (Chen et al., 2015; Chiang et al., 2010; Hutchings et al., 1996; Mori et al., 1982; Wallin et al., 2019), is highly concerning given that BUP is meant to alleviate adverse effects of opioid abuse during pregnancy. Studies using very high doses of morphine (up to 150 mg/kg liquid diet and up 40 mg/kg s.c.) have reported increased pup mortality (Eriksson & Ronnback, 1989; Sobrian, 1977) which could not be rescued by cross-fostering the morphine pups to drug-naive dams (Sobrian, 1977). This suggests that morphine-associated mortality is dose-dependent and due to the direct exposure to drug prenatally. In contrast, cross-fostering BUP-exposed pups to drug-naive dams can rescue the mortality (unpublished data from our lab, (Hutchings et al., 1995; Robinson & Wallace, 2001), suggesting that BUP-associated mortality may, at least partially, be due to the maternal behavior deficits.

Both our BUP groups and our continuous morphine group exhibited more NOWS symptoms than our controls, while discontinuing morphine rescued withdrawal symptoms, as expected. Given that the elimination half-life of BUP in the (non-pregnant) rat is 2.8 - 5.3 hrs (Gopal et al., 2002; Ohtani et al., 1994), the majority of BUP should have been eliminated from the system of the BUP discontinued dams at the time of parturition, suggesting that the observed “withdrawal symptoms” in the BUP discontinued pups may not have been solely drug-related. Pups were separated from their dam for 15 minutes for this assessment, so some behaviors may reflect irritability or general discomfort due to this maternal separation or to maternal neglect (and possible lack of nutrition) (Kuhn & Schanberg, 1998).

In line with the idea that offspring outcomes may partially depend on the opioid effects on the dams, we also report decreased placentophagia in BUP-discontinued dams in the present study, similar to our previous work and other reports (Mayer, 1985; Wallin et al., 2019), suggesting that decreased placentophagia may be an early indicator of reduced maternal care in this preclinical model. Moreover, pups from these dams had less visible milk bands compared to control pups on PND2, suggesting possible increases in maternal care behavior deficits over time in this group (e.g., less nursing). Unfortunately, we were unable to assess if the reduced milk bands were due to an inability of dams to nurse (i.e., lack of milk production) or due to their behavior selection (i.e., choosing not to nurse). Given that even a low dose of BUP (0.15 mg/kg, i.p.) caused a significant inhibition of oxytocin release (milk-ejection reflex) in lactating rats (Clarke & Wright, 1984) and that kappa opioid receptor (κOR) antagonists inhibit prolactin release in the prepartum (Andrews & Grattan, 2003), it is feasible that BUP may interfere with milk production and let-down in the rat.

Further in line with the idea that BUP is impacting offspring health primarily through deficits in maternal care and physiology is the fact that BUP exposure delayed the onset of nest building behavior. While morphine-exposed dams had less proficient nest building compared to VEH dams on GD20, they caught up to VEH dams by GD21. However, BUP exposed dams (continued and discontinued) barely touched the provided nesting materials until after parturition and most failed to build proper nests even during the postpartum period. Few studies have reported effects of gestational opioid exposure on nesting quality, but the observed scores are in line with our previous report that also found nests of low quality (i.e., flat, and sparse) for BUP-exposed dams (Wallin et al., 2019). Intriguingly, previous studies have shown that κOR agonists and stress can suppress nest building which can be rescued by κOR antagonists in both male and female (non-pregnant) rodents (Jacobson et al., 2020). This suggests that κOR may play an important role for nest building behaviors. However, given that BUP is a κOR antagonist, and that BUP lowered CORT levels in early pregnancy, one may have expected BUP to improve rather than diminish nest quality, More research is needed to better understand the impact of BUP's unique pharmacodynamics on pregnant rodents and their nests (Wallin et al., 2019).

Both BUP groups (continued and discontinued) were less likely to complete pup retrieval resulting in more ungrouped pups by the end of the test than VEH dams, replicating and extending results from our previous study (Wallin et al., 2019). Importantly, the reduced retrieval cannot be explained by a deficit in olfaction, as we did not see significant differences between the groups in the odor preference test. Aversion to pup odors/pups in general is typically seen in virgin rats or during early pregnancy, but late pregnancy usually induces a “switch” with dams, exhibiting a preference for pup orders and shorter latencies to retrieve pups (Brunton & Russell, 2008; Orpen & Fleming, 1987). Thus, we would have expected all dams to show a clear preference for the side with the pup odor (i.e., soiled bedding). Therefore, our findings suggest that BUP may delay or impair this “switch” in the maternal brain towards pup preference.

Previous studies have shown that an acute challenge of morphine (5.0 mg/kg, s.c.) during lactation decreases preference for pup odors when tested 60-min later (Kinsley & Bridges, 1990). Interestingly, the same study demonstrated that the same morphine challenge in virgin rats increased the preference for pup odors, which typically is adverse in virgins. This highlights the critical role that opioids have in regulating olfaction preferences and how this can change with reproductive status (Kinsley et al., 1995). The lack of an effect for morphine in our current study might be due to our test timing or set-up as we used soiled bedding from the nest area of the home cage of each dam. Given the critical role of olfaction in establishing a healthy mother-infant dyad, future studies should investigate the role of the various opioid receptors for adaptations in the maternal brain that may promote maternal care.

It is important to note the variability within the BUP groups in the behavioral tests. Maternal behavior impairments may manifest differently across individual animals, possibly accounting for the within group variability. For instance, a rat that performed poorly on the retrieval task might have scored well on the nesting behavior chart or the maternal care observations, and vice versa. This finding is in line with our previous studies that similarly had some BUP dams acting relatively normal on some behavioral tasks but displaying maternal impairments in other tests (Wallin et al., 2019). This variability requires further investigations.

In line with the idea that opioid exposure can alter maternal brain circuits, previous studies have shown that chronic opioid use (in non-pregnant animals) can alter the mesolimbic DA system (Koob, 2009; Sorge & Stewart, 2006) and lead to dysregulation of reward circuits (Vassoler et al., 2016), which share many common features and areas with the MBN such as the NAc, VTA and PAG (for review, see Wallin et al., 2021). We found that opioid treatment altered HIAA (the main metabolite for 5HT) in the NAc, VTA, and PAG, as well as DOPAC levels and DA/DOPAC turnover rate in the PAG, and NE in the VTA. The PAG is partially responsible for behavior selection and the organization of motivated responses, including the execution of defensive and reproductive behaviors (Sukikara et al., 2010). Changes in PAG activation may thus influence the dam's behavior selection. Our findings of low levels of HIAA and reduced DOPAC in the PAG may suggest that BUP and morphine impact DA synthesis and 5HT turnover, which in turn may impact PAG activation or function. This is in line with our behavioral results that showed that dams treated with BUP or morphine were less likely to engage in adaptive responses (i.e., retrieving pups back to the nest) than VEH dams. Further, 5HT agonists directly injected in the VTA have been shown to disrupt pup retrieval (De Almeida & Lucion, 1997), suggesting that opioid-induced alterations in the serotonergic system can impair maternal behavior (Numan et al., 2009). More research is needed to confirm the exact role of the various neurotransmitter systems in regulating maternal behavior and how opioids may alter these fine-tuned functions.

Interestingly, we also saw lower DOPAC concentration in the PFC of opioid-exposed (BC, BD, and MC) pups on PND2 compared to VEH offspring, suggesting that gestational or postnatal opioid exposure can impact the offspring’s dopaminergic system. We cannot rule out that opioids may affect other neurotransmitter systems or areas as well, but our values were often below the detection limit likely due to the small size of our areas of interest in a neonatal pup brain. Our findings are in line with a recent study by Elam and colleagues (2022) who found that gestational BUP exposure led to altered DA neuron activity in adult offspring and pre-pulse inhibition deficits compared to saline-exposed rats. More research is needed to illuminate the exact outcome of BUP exposure on offspring neurochemistry and whether a proportion of the deficits could be mitigated with improved maternal care behavior.

5.0. Conclusion

Taken together, the current study demonstrated that the administration of morphine or BUP from preconception through postpartum can deleteriously affect maternal behavior and offspring. Additionally, discontinuation of BUP in late gestation (GD19) was related to greater maternal impairment relative to the continuation of BUP through PND2, while discontinuing morphine in late gestation (GD19) had a 'rescue effect' on maternal behavior relative to the disrupted maternal care behavior seen with the continuation of morphine through PND2.

Translational implications from animal models are not always easily mapped onto the human experience, even when considering patterns of use and relevant exposure times (i.e., throughout pregnancy and the postpartum period). Humans are far more complex than rodents in terms of maternal care and maternal brain adaptations. There is less research on opioid-induced changes during parturition that could help guide our understanding of how exogenous opioids would interfere with the neuropeptide systems and the onset of maternal behaviors. However, clinical studies indicate that early breastfeeding efforts and rates of breastfeeding are lower in MOUD moms than in non-opioid-using mothers (Chard, 1989; Jansson, Choo, Velez, Harrow, et al., 2008; Jansson, Choo, Velez, Lowe, et al., 2008; Klaman et al., 2017). Additionally, there is evidence that misuse of opioids in pregnancy is associated with signs of impaired caregiving and reduced responses to infant cues (Conradt et al., 2019; Durjava, 2018; Johnson & Jones, 2018; Wang et al., 2018; Wells, 2009). In particular, Swain et al. (2019) suggested impairment in the reciprocal inhibition between PAG and the mPOA (hypothalamus) in human mothers treated with BUP for OUD. Despite our concerning preclinical and some clinical findings suggesting that gestational BUP exposure can interfere with maternal care behavior (which subsequently contributed to pup mortality), we want to stress that it is still recommended that pregnant women who are addicted to opioids continue on or transition to a MOUD such as BUP during pregnancy. Clinical findings still suggest that BUP exposure during pregnancy is less detrimental to both mother and baby as opposed to continued opioid use (e.g., morphine). However, our results do warrant further investigation into BUP’s effects on the maternal brain as well as neurodevelopmental effects to better understand the impact on mother and baby in humans. Understanding these mechanisms can help in developing safer treatment strategies and mobilize more support for opioid-dependent mothers and improve maternal brain health and well-being, which will ultimately benefit maternal attachment, bonding, and infant outcomes.

Highlights:

• Gestational buprenorphine (BUP) exposure resulted in reduced maternal care.

• Gestational BUP exposure resulted in high pup mortality.

• Both BUP and morphine decreased offspring weights and elevated withdrawal symptoms

• Increased pup mortality and morbidity may be due to maternal neglect.

• Discontinuing BUP exposure before parturition did not rescue adverse effects.

Data Availability Statement:

The data that support the findings of this study are available from the corresponding author, [SB], upon reasonable request.