Conceived and designed the experiments: SG ME AA CM SS KD. Performed the experiments: ME CM SS KD. Analyzed the data: SG ME AA CM SS KD. Contributed reagents/materials/analysis tools: SG AA. Wrote the paper: SG ME AA CM SS.
The authors have declared that no competing interests exist.
Child neglect is the most common form of child maltreatment, yet the biological basis of maternal neglect is poorly understood and a rodent model is lacking.
The current study characterizes a population of mice (MaD1) which naturally exhibit maternal neglect (little or no care of offspring) at an average rate of 17% per generation. We identified a set of risk factors that can predict future neglect of offspring, including decreased self-grooming and elevated activity. At the time of neglect, neglectful mothers swam significantly more in a forced swim test relative to nurturing mothers. Cross-fostered offspring raised by neglectful mothers in turn exhibit increased expression of risk factors for maternal neglect and decreased maternal care as adults, suggestive of possible epigenetic contributions to neglect. Unexpectedly, offspring from neglectful mothers elicited maternal neglect from cross-fostered nurturing mothers, suggesting that factors regulating neglect are not solely within the mother. To identify a neurological pathway underlying maternal neglect, we examined brain activity in neglectful and nurturing mice. c-Fos expression was significantly elevated in neglectful relative to nurturing mothers in the CNS, particularly within dopamine associated areas, such as the zona incerta (ZI), ventral tegmental area (VTA), and nucleus accumbens. Phosphorylated tyrosine hydroxylase (a marker for dopamine production) was significantly elevated in ZI and higher in VTA (although not significantly) in neglectful mice. Tyrosine hydroxylase levels were unaltered, suggesting a dysregulation of dopamine activity rather than cell number. Phosphorylation of DARPP-32, a marker for dopamine D1-like receptor activation, was elevated within nucleus accumbens and caudate-putamen in neglectful versus nurturing dams.
These findings suggest that atypical dopamine activity within the maternal brain, especially within regions involved in reward, is involved in naturally occurring neglect and that MaD1 mice are a useful model for understanding the basis of naturally occurring neglect.
Child neglect is the most common form of child maltreatment and is highly debilitating. Although often considered together, child neglect and abuse are separable processes in humans and other primates
Dopamine is considered to be a key player in reward-related behaviors
In these studies, we describe a unique mouse model with consistent rates of naturally occurring maternal neglect within an otherwise nurturing population. The mice producing high levels of maternal neglect were originally one of four lines of mice selected for high wheel-running behavior from outbred (hsd:ICR) mice
When examined over 12 generations (∼80 lactating females evaluated per generation per group), MaD1 mice
Maternal neglect rate (neglect leads to death of all pups in a litter) was significantly higher in MaD1 relative to Outbred-S mice when examined over 12 generations (A). Birth rate (number of male-female pairings relative to number of litters born) is almost identical between MaD1 and Outbred-S mice when examined over 12 generations (B). When raising a second litter (examined in Generation 6), previously neglectful mice exhibited significantly higher levels of maternal neglect relative to previously nurturing MaD1 mice (C). Bars represent means±SE. *** = p<0.001.
In most generations, maternal behaviors were not recorded, but when they were recorded, they were based on undisturbed observations. In Generation 16, first time MaD1 mothers were examined three times per day to provide a profile of maternal neglect. During this generation, 9 mothers were neglectful. Within about 80% of the neglected litters, milk was apparent within the pups (
Descriptive Measure | Average | Standard Error |
Proportion of litters with milk bands in pups | 0.78 | ±0.14 |
Last day of nursing | 1.67 | ±0.50 |
Appearance of 1st dead pup (days) | 1.67 | ±0.57 |
All pups dead (days) | 2.89 | ±0.56 |
We examined general traits across different generations determine whether any of these varied with neglect. Using a 2-way ANOVA analysis that included generation number as a variable, we found that dam weight measured on postpartum Day 0 was significantly lower in neglectful mice relative to nurturing mice (
When examined for the first litter in on postpartum Day 0, neglectful MaD1 mice weighed significantly less relative to their nurturing sisters (A) and average pup weight was significantly lower for pups born to neglectful mothers (B). Litter size was smaller for neglectful mice, but levels did not reach significance(C). Prior to mating, decreases in self-grooming (D) and increases in activity (E) were seen in neglectful relative to nurturing mice (D). For A–C, data from Generations (G) 12 and 15 were combined and for D and E, data from G6, G12, and G15 were combined. Time in light in a light/dark box test, did not differ between mice that would become either neglectful or nurturing when examined in G15 (F). Mice were first examined for anxiety at age 50 when they were group housed and then again after being singly housed for one week. In G17, neglectful mice spent significantly less time floating in the forced swim test relative to nurturing mice (G). Bars represent means±SE. * = p<0.05; *** = p<0.001.
Unexpectedly, pups born to previously neglectful (relative to nurturing) MaD1 mothers were significantly more likely to receive maternal neglect (F (1,50) = 11.3, p = 0.002); 2-way ANOVA). This effect was clearly seen when neglectful pups were being raised by either previously nurturing MaD1 (H (1,22) = 5.4, p = 0.02; ANOVA on Ranks) or Outbred-S mice (H (1,17) = 6.6, p = 0.01; ANOVA on Ranks) (
When raised by either previously nurturing MaD1 mice or Outbred-S mice, pups that were born to previously neglectful MaD1 mothers received significantly higher levels of maternal neglect relative to pups born to previously nurturing MaD1 mice (A). Similarly, the proportion of pups weaned was significantly lower for pups born to previously neglectful MaD1 mothers when raised by either previously nurturing MaD1 mothers or Outbred-S mothers (B). Previously neglectful (relative to nurturing) MaD1 mice weaned a lower proportion of pups when the pups were from nurturing mothers, but the differences did not reach significance (p = 0.071). Bars represent means±SE. * = p<0.05; ** = p<0.01.
When comparing proportion of nurturing MaD1 pups weaned between neglectful and nurturing MaD1 mothers, heightened neglect was seen in the neglectful mothers but this did not reach significance (p = 0.07) (
Although a main goal of the cross-fostering study was to examine how offspring would fare as adults when given different rearing environments, during the pup rearing phase we conducted one maternal behavior observation for 1 hour on postpartum Day 3. No significant differences were observed between previously neglectful and nurturing MaD1 mothers, although a consistent trend towards lower maternal care was seen among neglectful mothers (data not shown). Negative effects of neglectful mothers on offspring were seen when offspring were adults (see below). So it is possible that with prolonged maternal care observations, significant differences between groups could be identified.
Previously neglectful (relative to nurturing) MaD1 mothers imparted deficits on offspring they were raising when offspring were examined as adults. When using a 2-way ANOVA to examine the overall effect of neglectful versus nurturing mothers when raising pups from all three pup groups, adult weight of female offspring raised by neglectful mothers was significantly lower relative to those raised by nurturing dams (
When results from all offspring were combined, previously neglectful MaD1 mothers negatively impacted offspring adult performance in terms of body weight pre-mating (A), flipping rate pre-mating (B), body weight on postpartum Day 0 (C), and number of maternal defense attacks (D). Additionally, when just offspring from previously nurturing MaD1 mothers were examined, they exhibited significantly lower levels self-grooming pre-mating when raised by a neglectful mother (E). When just offspring from Outbred-S mothers were examined as adults, deficits in terms of proportion of pups that survive to postpartum Day 10 (a marker of maternal neglect) (F), and the average weight of pups on postpartum Day 10 (G) were observed when these mice were raised by neglectful MaD1 mothers. Bars represent means±SE. * = p<0.05; ** = p<0.01.
Interestingly, planned comparisons revealed negative impacts of being reared by neglectful MaD1 mothers that were specific to offspring genotype. For example, decreased self-grooming rate as adults was observed if the offspring were from previously nurturing MaD1 mothers (
Neglectful females show region specific differences in c-Fos, phosphorylated tyrosine hydroxylase (pTH), and phosphorylated DARPP-32 (pDARPP-32) expression compared to their nurturing counterparts. In particular, c-Fos expression was significantly higher in neglectful (N = 8) versus nurturing mice (N = 9) in dopamine releasing and responding regions (
Heightened c-Fos expression in dopamine releasing and responding regions in neglectful (N = 8) relative to nurturing (N = 9) mice. Significantly higher levels of c-Fos are found in ZI, VTA, and substantia nigra (SN), all of which are involved in dopamine production. Both nucleus accumbens shell (AcS) and core (AcC), which respond to dopamine signaling, also show increased c-Fos in neglectful mice. Other regions examined for c-Fos are shown in
Example of heightened pTH immunoreactivity in ZI in neglectful relative to nurturing mice is shown in (A). Significant elevations of pTH-ir area are found in ZI, but not other dopamine producing regions, in neglectful mice (N = 7) in comparison to nurturing dams (8) (B). TH-ir area does not differ between neglectful (N = 8) and nurturing (N = 8) mice. ZI = zona incerta, VTA = ventral tegmental area, A14 POA = A14 region of preoptic area. Bars represent means±SE. * = p<0.05.
Previously neglectful dams (N = 5) have significantly darker optical density of pDARPP-32 expression in Ac and CP when compared to previously nurturing dams (N = 6). Ac = nucleus accumbens, LS = lateral septum, CP = caudate-putamen, BST = bed nucleus of stria terminalis, dorsal, CeA = central amygdala. Bars represent means±SE. * = p<0.05.
Brain Region | Neglectful | Nurturing | p-value |
LSV | 31.3±6.0 | 6.6±2.6 | p<0.001 |
CG | 119.2±26.3 | 7.7±2.1 | p<0.001 # |
MeA | 102.3±24.5 | 16.8±5.6 | p = 0.004 # |
CeA | 57.0±25.6 | 12.0±2.7 | p = 0.030 # |
LH | 23.8±7.0 | 5.6±2.3 | p = 0.012 # |
Pir | 265.0±62.1 | 24.6±5.7 | p = 0.005 # |
PVN | 36.0±7.8 | 12.7±3.4 | p = 0.013 |
cPAG | 32.1±5.0 | 13.5±2.3 | p = 0.004 |
MPOM | 77.5±18.6 | 38.6±15.9 | p = 0.083 # |
MPA | 45.5±18.5 | 34.2±4.5 | p = 0.665 |
BNSTv | 19.6±3.1 | 8.0±2.2 | p = 0.008 |
AHA | 22.1±6.0 | 9.5±2.3 | p = 0.136 # |
SCN | 71.5±18.9 | 90.0±18.4 | p = 0.248 # |
cPAG1 | 31.0±9.5 | 17.7±3.7 | p = 0.210 # |
See
Abbreviations: LSV = lateral septum ventral; CG = cingulate cortex; MeA = medial amygdala; CeA = central amygdala; LH = lateral hypothalamus; Pir = piriform cortex; PVN = paraventricular nucleus; cPAG = caudal periaqueductal gray; cPAG1 = more caudal aspect of cPAG; MPOM = medial preoptic nucleus; MPA = medial preoptic area; AHA = anterior hypothalamic area; SCN = suprachiasmatic nucleus. #- data non-normal, analyzed with ANOVA on Ranks.
We suggest MaD1 mice can be utilized as a powerful tool for examining the biological basis of naturally occurring neglect. This line of mice can consistently produce a subset of neglectful mothers (∼17% each generation) (
Interestingly, this study indicated dysregulated dopamine signaling in neglectful MaD1 mice. This may prove very pertinent because in humans, abnormalities in parental appetitive motivation (lack of hedonic reward from children) rather than an inability to perform a particular parental behavior is implicated in some cases of child neglect
By examining behavior both prior to mating, during pregnancy, and at the time of birth, we were able to determine predictors, or risk factors, that were associated with maternal neglect (
Low self-grooming (
It is unclear why pups born to previously neglectful mice were more likely to be neglected by two of the three maternal groups in the cross-fostering study (
Together, we see a deleterious environmental effect of being reared by previously neglectful mothers that manifests itself in terms of lower adult body weight, heightened flipping, decreased dam weight, and decreased number of maternal defense attacks (
How the rearing environment provided by previously neglectful MaD1 mothers negatively affects offspring performance is not known, but recent work suggests epigenetic contributions to elevated activity in rats
There are region specific differences in the brain of neglectful versus nurturing mice. We found altered c-Fos activity (an indirect marker for neuronal activity) in many brain regions that are important for dopamine signaling, maternal behavior, or both. Elevated c-Fos in nucleus accumbens shell region (AcS) and core region (AcC) is interesting because these are targets of dopamine neurotransmission from VTA
We observed a key difference in the activational state of TH in neglectful mice. The phosphorylation of TH is involved in the short term regulation of this enzyme
As we find no difference in overall levels of TH, these findings suggest a difference in the activity of dopamine neurons rather than in neuronal number within neglectful MaD1 mice. Elevated pTH activity in ZI does not simply reflect underlying differences in enzyme expression and instead, might be due to neglectful mice having a heightened input to ZI. In future work it will be important to examine in neglectful mice possible differences in dopamine receptors and inputs to dopamine systems.
Another useful approach to examine whether dopamine activity is heightened in neglectful mice is to examine the postsynaptic phosphorylational state of DARPP-32. Phosphorylation of DARPP-32 is increased by D1-like, but not D2-like, dopamine receptors and is specifically phosphorylated on threonine34 (Thr34) in response to dopamine acting through a cAMP-dependent protein kinase (PKA)
Elevated pDARPP-32 levels with neglect are also of interest because progesterone can affect
The altered c-Fos, pTH, and pDARPP-32 activity in neglectful mice is consistent with altered dopamine signaling. Elevated pTH in VTA (a key dopamine releasing region) in neglectful mice did not reach significance, but c-Fos did. It is possible that altered dopamine production occurs in VTA and that another approach will be needed to evaluate this. The differences seen in c-Fos expression compared to pTH and pDARPP-32 expression also indicate that dopamine may not be the only player influencing neglect.
Our findings implicating dopamine dysregulation in the production of naturally occurring neglect are consistent with a number of prior studies that suggest dopamine involvement in neglect. For example, disruption of D1, D2, or D4 dopamine receptors in medial preoptic area impairs differing aspects of maternal care in rats
Normal maternal care involves, in part, linking response to offspring with natural reward systems. For example, lactating females will bar press for pups
The current study characterizes a population of mice (MaD1) in which an average of 17% naturally neglect their offspring per generation. We have identified a set of behavioral risk factors, such as decreased self-grooming and increased activity, which are associated with maternal neglect. As elucidated by the cross-fostering study, there are possible epigenetic contributions to neglect. That is, offspring raised by neglectful mothers will then exhibit decreased maternal care as adults, suggesting a mother-to-offspring transference of neglect. We also report that offspring born to previously neglectful dams can elicit heightened maternal neglect when cross-fostered to normally nurturing mothers, suggesting that factors regulating maternal neglect can be triggered by the offspring. While the biological basis for maternal neglect is not known, we found that atypical dopamine activity might be one factor regulating maternal neglect of offspring. In conclusion, these data suggest that MaD1 mice can be utilized as a powerful model for examining the biological basis of naturally occurring neglect, as well as the mother-offspring relationships that regulate neglect.
MaD1, Outbred-S, and outbred hsd:ICR (Harlan, Madison, WI) mice were used. MaD1 mice have been maintained in our lab for over 17 generations using selection for high maternal defense. Outbred-S mice were derived by us from outbred hsd:ICR mice also using selection for high maternal defense
Prior to pairing with a male, MaD1 females were examined for general behaviors for Generations 6, 12, and 15. All observations were made of mice in their home cage in the home room and took place in the morning for 1 hour between 0900 and 1000. Observations included self-grooming and activity (locomotion, flipping or cage top climbing). In Generation 12, general behavioral observations of the females were also conducted during the second week of pregnancy. For all behavioral observations here and below, mice were coded and recordings were made by individuals blind to experimental conditions. Maternal neglect rate was determined by examining the number of live litters on postpartum Day 5 relative to the number of live litters born. Birth rate was determined as the number of live births relative to the number of females paired with males for breeding. Litter size was determined by counting live pups on postpartum Day 0. In Generation 16, first time mothers were examined 3 times a day (∼8–10 am; 4–6 pm; and 10–12 pm) from postpartum Day 0–7. Maternal features, such as nursing, and pup features, such as milk in stomach and pup death, were recorded. In Generation 17, pup retrieval was examined on postpartum Day 0 on mothers with their second litter.
Evaluation of maternal neglect rate and birth rate between MaD1 and Outbred-S mice was made using a one-way ANOVA with data from each of 12 generations for each group. For each generation, approximately 80 lactating females were examined per group. For examination of maternal neglect rate on a second litter a one-way ANOVA was used. For examining risk factors with neglect, if data were collected from multiple generations, then generation itself was included as a variable and a 2-way ANOVA was used. In cases here and below where the data were not normally distributed, either transformations were used to achieve normality before running the ANOVA or non-parametric tests were used.
Mice were placed in the dark portion of the light/dark box to initiate the 5 min test session. Time spent in the light and dark portions of the box were recorded with time in light portion of the box defined as entry of all four paws into this region. Mice were examined prior to mating both while they were group-housed and also following one week of individual housing. All behaviors were recorded on videotape and subsequently analyzed off-line. The light/dark box was used as a tool for examining levels of anxiety
In Generation 17, at first sign of neglect (in almost all cases Day 1), neglectful and nurturing (stage matched) mice were placed in a glass cylinder (30 cm tall, 15 cm diameter) that was half filled with room temperature water and tested for 5 min. Time spent swimming and floating were recorded. Additionally, number of fecal boli produced while in the water were counted at the end of the session. All behaviors were recorded on videotape and subsequently analyzed off-line. Results were analyzed using a one-way ANOVA between neglectful and nurturing MaD1 mice.
In order to examine the effects of maternal rearing environment and genotype on maternal neglect and offspring performance as adults, a cross-fostering study was conducted using mice on their second litter. Neglectful and nurturing MaD1 mice were identified based on performance with their first litter. Outbred-S mice that successfully raised pups for their first litter were also used. All mice were age matched and bred with mice of the same strain at the same time. At birth, all litters were cross-fostered among the three groups. Thus, 9 groups were created with ∼10 lactating mice per group. Previously neglectful MaD1 dams cared for pups from either a) previously neglectful MaD1 mice; b) previously nurturing MaD1 mice; or c) previously nurturing Outbred-S mice. Previously nurturing MaD1 and Outbred-S mothers also cared for these three groups. In all cases, the cross-fostering was completed whereby no dam raised her own pups. A maximum of 11 and minimum of 9 pups were cross-fostered per dam. On postpartum Day 3 of cross-fostering, behaviors of dams were examined in the home room for one hour. Behaviors were recorded every 30 sec using handheld Palm Pilots (Palm, Inc, Sunnyvale, CA) and information was downloaded to a computer for analysis. At age 21 days, pups were weaned and the percent pup survival was recorded. When cross-fostered pups were adults, behavioral measures were examined both prior to mating and during lactation. Cross-fostered female mice were mated as adults with an outbred (hsd:ICR strain) breeder male and examined for maternal behaviors. Maternal aggression and pup retrieval were examined using techniques previously described
For analysis, a 2-way ANOVAs (using maternal rearing group and pup group) were used. Also, planned comparison one-way ANOVAs were used. For example, for examining the effect of rearing environment on offspring adult behavior, comparisons were made between common offspring that were raised by either neglectful versus nurturing MaD1 mothers.
Beginning on postpartum Day 0 in Generation 16, all MaD1 mice were monitored for maternal behaviors twice per day. At first sign of maternal neglect, brains from neglectful mice were immediately collected. All brains were collected on Days 1 or 2 (the majority collected on Day 1), so this variable was not used as a covariate. Control brains from maternally normal mice were collected at the same time and the timing from birth to brain collection was identical for both groups. In Generation 17, previously neglectful or nurturing dams were sacrificed on postpartum day 0 of their second litter and processed for pDARPP-32.
For all brain collections, mice were decapitated following light isoflurane anesthesia and the brains removed. Brains were post-fixed overnight in 5% acrolein (Sigma) in phosphate buffered saline (PBS) and cryoprotected in 30% sucrose in PBS for two days. Brains were frozen on a platform and cut into 40 micron thick coronal sections using a sliding microtome (Leica, Microsystems, Heidelberg, Germany) and stored in a cryoprotectant solution at −20 degrees C until processing for immunohistochemistry. For each antibody used, immunohistochemistry was run for all mice in both groups in one batch.
For c-Fos, sections were incubated with 0.5% sodium borohydride for 30 min, washed in PBS in the presence of 0.2% Triton-X-100 (PBS-X), and blocked in 5% NGS for 1 hr. Sections were then incubated for two days at 4 degrees C in rabbit anti-c-Fos (1∶15,000; Calbiochem, La Jolla, CA; catalog # PC38). After washes in PBS-X, the sections were incubated for 90 min at room temperature with anti-rabbit secondary antibodies (1∶500, Vector), washed in PBS-X, exposed to an avidin-biotin complex (Vector) for 1 hr, washed again in PBS-X, and stained using diaminobenzidine (Sigma) enhanced with 0.008% nickel chloride. The sections were then mounted, dehydrated in a series of ethyl alcohols and xylenes, and coverslipped.
For pTH, TH and pDARPP-32, sections were washed three times in 0.1 M tris-buffered saline (TBS, pH 7.4) and then in TBS containing 0.1% sodium borohydride for 15 minutes. Following three more washes, sections were placed for 1 hour in TBS containing 20% normal goat serum (NGS) and 3% hydrogen peroxide. Tissue sections were then incubated overnight at room temperature in rabbit anti- ser40 pTH (1∶2,000; GeneTex, San Antonio, TX; catalog # GTX16557), mouse anti-TH (1∶5,000; GeneTex; catalog # GTX30172), or rabbit anti-pDARPP-32 (1∶1,000; Zymed, San Francisco, CA; catalog # 38-7500) in TBS containing 0.3% Triton X-100 (TBST), 2% NGS, and 0.5% gelatin. Following primary incubation, sections were washed three times in TBST and incubated in TBST containing the appropriate secondary antibody (1∶500, Vector Laboratories, Burlingame, CA) and 0.5% gelatin for ninety minutes. Following further washes, sections were exposed to an avidin-biotin complex (1∶400, Vector) for 1 hr, washed three times in TBST, and visualized using Vector SG (1∶167, Vector). Developed sections were mounted on gelatin-coated slides and coverslipped. Because the anti-TH antibodies were made in mouse, it was possible that the anti-mouse secondary would react with mouse brain tissue. However, when running control sections with just secondary anti-mouse antibodies, no staining occurred indicating that staining with primary antibodies was specific.
Bright field microscopy was used for counting c-Fos-positive cells. The images of brain sections were projected from an Axioskop Zeiss light microscope using a 10× objective (Zeiss, Gottingen, Germany) through an Axiocam Zeiss high resolution digital camera attached to the microscope and interfaced with a computer. Counting from specified brain regions was based on a previously used paradigm
Bilateral analyses of pTH, TH, and pDARPP-32 for one brain section per region were conducted using an Olympus BX61 microscope fitted with an Olympus FV II digital camera, connected to a PC compatible computer. The software used for analysis was Olympus MicroSuite (Soft Imaging System Corp., Lakewood, CO). Thresholds to detect foreground were set independently for each measurement to account for possible variability in background staining. The threshold was determined automatically by the imaging software, and was approximately 3× the standard deviation greater than the gray value mean of the background staining. Staining with a gray value greater than the threshold was detected and the total area covered by positive staining was obtained using the detection setting within the Olympus Microsuite software program. We have used this method of detection threshold of around 3X the standard deviation of mean background routinely with great success
The authors wish to thank Derek Powell, Dan Goldman, and Michael Hlavenka for technical assistance, and Kate Skogen and Jeff Alexander for animal care.