
2023 Impact Factor
Epidemiological findings indicate that childhood maltreatment increases lifetime prevalence of major depression [1,2] and reduces therapeutic efficacy in major depression [3], which indicate the possibility that patients with major depression with the childhood maltreatment might have different pathophysiology from that of the patients without the childhood maltreatment. Although early disturbances in neurocognitive processes, such as threat processing, reward anticipation, and emotional regulation, mediated through altered brain region structure, function, or circuitry, are broadly viewed as increasing vulnerability to psychiatric disorders later in life, the precise mechanisms of this vulnerability remain elusive [4,5]. Cross-sectional human studies and retrospective self-reported studies suggest a causal relationship between childhood maltreatment and structural or functional alteration of the brain. However, several confounding factors in these studies (e.g., socioeconomic status, intelligence quotient, maltreatment type, duration, and age at the time of maltreatment) complicate the elucidation of the causal relationship [6,7]. On the other hand, animal studies under controlled maltreatment conditions are considered appropriate approaches to understanding the effects of maltreatment on the brain structure and function [8].
Among various types of childhood maltreatment, environmental deprivation, a central feature of child neglect and institutional rearing, has been reported to increase the risk of many forms of psychopathology, including anxiety, depression, attention-deficit/hyperactivity disorder, aggression, and substance abuse [9-11]. Notably, McLaughlin and her colleagues recently proposed that an environment of early deprivation constrains development of cognitive processes such as associative learning and implicit learning in infants, and subsequently may increase risk for psychopathology [12]. It is intriguing that the defect in associative and implicit learning due to environmental deprivation is also associated with pessimism. Pessimism, one of the well-known cognitive characteristics of depression, has been associated with the onset of depression [13] and severity of depressive symptoms [14], and the pessimistic mode of thinking (seeing the future as hopeless and giving up easily) can be interpreted as using inaccurate associative and implicit learning for coping with aversive stimuli or undesirable life events [15]. Based on these findings, it could be worthwhile to investigate the neural mechanism of pessimistic associative and implicit learning after environmental deprivation with the aim of developing novel therapeutic and preventive strategies for patients with major depression who have suffered environmental deprivation early in life.
Although the neural mechanism underlying associative and implicit learning is not fully understood, structures of the basal ganglia neurocircuitry such as the thalamus, striatum, and pallidum have been reported to be closely implicated in avoidance learning [16]. In fact, recent optogenetic and chemogenetic studies reveal the involvement of activation of vesicular glutamate transporter 2 (VGLUT2)-expressing glutamatergic ventral pallidum (VP) neurons [17] and inhibition of preproenkephalin (Penk)-expressing GABAergic VP neurons [18] in avoidance learning. With regard to the VP, Gehred and her colleagues reported that adults with adverse childhood experiences had a smaller VP [19], but they did not examine associative and implicit learning. Furthermore, the causal association of VP with disease states has not been verified [20] because the projection targets, anatomical subdivisions, and innervation patterns of the VP are very heterogeneous [21].
In the present study, to address the mechanism underlying environmental deprivation-induced maladaptive avoidance learning in childhood, we examine 1) whether neonatal isolation (NI), a stress experience in early life, could influence the incidence of major depression in adulthood using learned helplessness (LH) rats, an animal model of major depression mimicking pessimistic bias following exposure to uncontrollable stress [22], 2) whether NI could influence the number of VGLUT2-expressing VP cells and Penk-expressing VP cells by using immunohisto-chemistry, 3) the difference in the number of VGLUT2-expressing VP cells between LH and non-LH rats subjected to NI, and 4) the association between escape latency (seconds) in the LH test and the number of VGLUT2-expressing VP cells.
All experimental procedures are shown in Figure 1.
Sixty pregnant female Sprague–Dawley rats were purchased from The Jackson Laboratory Japan, Inc., and 222 male Sprague-Dawley rats were used in this study (n = 123 in Experiment 1, n = 38 in Experiment 2, and n = 61 in Experiment 3 and 4). The pregnant rats were housed individually in standard polycarbonate cages (38 × 23 × 20 cm; SEOBiT) with sawdust bedding. The day of birth was counted as postnatal day 1 (PN1). Dams were cohoused with their pups in home cages (38 × 23 × 20 cm clear plastic cages) until weaning (PN21). After weaning, male rats were housed 2−3 animals per cage, in a temperature-controlled room with a 12-hour light/dark cycle (lights on at 8:00 a.m.) and given ad libitum access to water and food. Only male pups were used for later experiments because of the reported bidirectional (facilitation and suppression) influence of the estrous cycle and gonadal hormone on learned fear [23]. All animal procedures were conducted in accordance with the Hiroshima Uni-versity Animal Care and Use Committee Guiding Prin-ciples on Animal Experimentation in Research Facilities for Laboratory Animal Science (approving number from the committee was #A20-72), and in accordance with the Guidelines for Proper Conduct of Animal Experiments of the Science Council of Japan.
For the evaluation of the effects of environmental deprivation early in life, animals were randomly assigned to 2 groups (sham or NI). The pups in the sham group were housed under normal conditions and left undisturbed until weaning, except for cage cleaning twice a week. NI was conducted as previously described [24,25]. Briefly, pups were isolated from their dam and siblings and placed in individual paper cups (7 cm diameter and 8 cm deep) without bedding in a temperature-controlled (30 ± 2°C, same as the nest temperature) and humidity-controlled (60%) chamber for 1 hour per day on PN days 2−9. Each paper cups was placed 5 cm apart from others. Daily isolation was performed between 09:00 a.m. and 12:00 a.m.
The LH test was performed as previously described [26,27] over two days (PN 54 and PN 55). Rats were subjected to an inescapable shock (IS) session on Day 1, subjected to an avoidance test (AT) session on Day 2, and habituated to the experimental room, which was dimly illuminated with indirect white light, for at least 20 minutes prior to each session. At the IS session, rats were placed in an experimental chamber (325 mm width, 280 mm height, 500 mm depth), and then exposed to 80 inescapable foot-shocks (single shock intensity: 0.8 mA, each duration: 15 seconds, without a light signal, interval: 10−20 seconds, total time: 40 minutes) generated by a shock generator-scrambler (SGS-003; Muromachi) through a stainless steel grid floor. After the completion of the IS-session, rats were returned to their home cage and kept undisturbed until the following day. At the AT session 24 hours after the IS session, rats were placed in the same chamber where the IS session was performed and exposed to 15 avoidable foot-shocks (single shock intensity: 0.8 mA, each duration: 60 seconds, inter-shock time: 24 seconds). During the AT session, a lever was mounted 5 cm above the grid floor on one wall of each chamber. Each foot-shock could be ended by depressing the lever, and a simultaneous signal of light placed above the lever was triggered by each foot-shock and served as a cue to facilitate detection of the lever and discriminate between the IS and AT sessions. Escape latencies, the time from the beginning to the end of foot-shock, were recorded automatically, and judged as follows: < 20 seconds as a ‘success’ and 20−60 seconds as a ‘failure’. Finally, LH status in each rat was classified as LH, intermediate LH, or non-LH according to the number of failures (LH: more than 11 failures; intermediate LH: 5−10 failures; non-LH: fewer than 4 failures). The rats classified as having ‘intermediate’ LH were excluded from subsequent experiments.
Chloral hydrate (300 mg/kg i.p.) was used to deeply anesthetize Sham and NI rats on Day 54, and LH rats in the Sham and NI groups immediately after the completion of the LH test (Fig. 1). Then rats were transcardially perfused with tris buffered saline (TBS) followed by 10% buffered formalin solution. Following perfusion, brains were quickly removed, post-fixed in the same fixative at 4°C for 48 hours, cryoprotected in 30% sucrose-TBS for 1 week, stored at −80°C, and sectioned (40 mm) using a cryostat.
For the detection of VGLUT2-expressing cells and Penk-expressing cells, double immunohistochemical staining was performed using the ImmPRESSⓇ Duet Double Staining Polymer Kit (MP-7714; Vector Laboratories). The brain sections were incubated first in TBS for 1 hour, then in 4 % paraformaldehyde for 10 minutes to ensure fixation, in 0.3% H2O2 solution at room temperature for 10 minutes to quench endogenous peroxidase activity, 0.6% TritonX-100 for 15 minutes, and rinsed three times, each time for 5 minutes with TBS at room temperature. For blocking of nonspecific secondary antibodies, each section was placed in 2.5% normal horse serum for 1 hour, followed by a mixture of 1% bovine serum albumin with 3% skim milk for 1 hour, and then incubated overnight with primary antibody (rabbit, anti-VGLUT2, 1:1,000; Abcam, ab216463 or rabbit, anti-Penk, 1:800; Abcam, RA14124) and anti-substance P antibody (mouse anti-substance P, 1:500; Abcam, ab14184) at 4°C. For the identification of the boundary of the VP, anti-substance P antibody staining was performed because this region of the basal forebrain is heavily innervated by GABAergic Substance P [17]. After three wash steps in TBS for 5 minutes each, sections were incubated with secondary antibodies (ImmPRESS Duet Reagent, containing HRP-conjugated horse anti-rabbit IgG and alkaline phosphatase [AP]-conjugated horse anti-mouse IgG) for 1 hour at room temperature, washed three times in TBS with Tween 20 (TBS-T) for 10 minutes each, and incubated with the labeling substrates for each antibody included in the kit (ImPACT DAB EqV substrate for 7 minutes and ImPACT Vector Red AP substrate for 20 minutes) to chromogenically detect the secondary antibodies. Finally, sections were dehydrated and mounted with DPX (Sigma-Aldrich).
Bright field images of dually stained tissues by immunohistochemistry were captured using a Keyence BZ-X9000 fluorescence microscope (Keyence Corporation). At first, we identified the substance-P positive VP region in 4 slices (bregma + 0.32 mm, + 0.24 mm, + 0.16 mm, + 0.08 mm) at × 4 magnifications, then in 2 images at × 40 magnifications. The camera exposure and gain settings were held constant between sections.
The color deconvolution plugin of ImageJ Fiji software (Version 1.53t, National Institutes of Health) was used to distinguish DAB-stained cells. The color thresholds for the detection of stained cells were held constant between sections. The number of DAB-positive cells was counted automatically, and the mean number of positive cells per 0.1 mm2 of four coronal sections in each hemisphere of each animal was calculated.
In the analysis of the number of escape failures from foot-shocks in the LH test in Sham and NI rats, the data are expressed as actual numbers and were analyzed by the Kolmogorov–Smirnov two-sample test (Fig. 2A).
In the comparison between Sham rats and NI rats, the mean number of failures in the LH test (Experiment 1) and data of the number of VGLUT2-expressing cells and Penk-expressing cells (Experiment 2) are expressed as the mean ± SEM and were analyzed by independent t tests (Figs. 2B, 3B, 3D).
In Experiment 3, data are expressed as mean ± SEM and were analyzed by two-way ANOVA (factors: NI, LH) (Fig. 4).
In Experiment 4, data are expressed as the actual numbers and the correlation between the number of VGLUT2-expressing cells and reaction times in the LH test was evaluated using the Pearson’s correlation coefficient test.
Results were considered statistically significant at p < 0.05.
Depression-like behavior in response to stress was compared between 67 Sham and 56 NI rats by measuring the number of escape failures from foot-shocks in the LH test (Fig. 1, Experiment 1).
Non-LH, intermediate LH, and LH were identified in 40 rats (59.7%), 18 rats (26.9%), and 9 rats (13.4%), respec-tively, of the Sham group and 8 rats (14.3%), 13 rats (23.2%), and 35 rats (62.5%) of the NI group. Statistical analysis revealed a significant difference in the LH responses (incidences of non-LH, intermediate, LH) to stress between the Sham and NI groups (D = 0.538, p < 0.001). (Fig. 2A) and a significantly higher numbers of escape failures in the NI group than the Sham group (t = −7.14, df = 121, p < 0.01) (Fig. 2B).
These results indicate that NI could increase susceptibility to LH in adulthood.
To address the mechanisms underlying increased susceptibility to LH in adulthood in NI rats (found in Experiment 1), the number of VGLUT2-expressing VP cells and the number of Penk-expressing VP cells were compared between NI and Sham rats because of reports showing the relevance of the activity of these neurons to avoidance learning [17,18].
Statistical analysis revealed that the number of VGLUT2-expressing cells was significantly decreased in the NI group (t = 2.88, df = 20, p < 0.01). (Fig. 3A, B) but the number of Penk-expressing cells was not significantly different between the Sham and NI groups (t = 0.21, df = 14, p = 0.83) (Fig. 3C, D).
These results indicate that NI might decrease the number of VGLUT2-expressing VP cells in adulthood.
To investigate whether the number of VGLUT2-expressing VP cells could affect the avoidance learning in both Sham and NI rats, differences in the number of VGLUT2-expressing VP cells were compared between the LH and non-LH groups of Sham and NI rats.
Two-way ANOVA revealed a significant main effect of NI (F(1, 60) = 4.967, p = 0.0298), of LH (F(1, 60) = 4.251, p = 0.0438), and an interaction between NI and LH (F(1, 60) = 4.472, p = 0.0388) (Fig. 4).
These results indicate that the number of VGLUT2-expressing VP cells might affect the avoidance learning only in NI, but not in Sham rats.
The influence of decreased VGLUT2-expressing VP cells on the maladaptive avoidance learning of NI rats was examined by evaluating the correlation between the number of VGLUT2-expressing VP cells and escape latency in the LH test.
The number of expressing cells and escape latency were found to be significantly correlated (r = −0.37, p = 0.029) (Fig. 5).
These results indicate that NI might increase the susceptibility to LH in adulthood via decreasing the number of VGLUT2-expressing VP cells.
The higher incidence and treatment resistance of major depression in patients having a history of childhood maltreatment suggest that childhood maltreatment increases susceptibility to major depression. Based on the pessimistic bias of patients with major depression and the clinical findings in victims of early environmental deprivation, such as associative and implicit learning impairments and smaller VP volume, it is hypothesized that the lack of opportunity to learn through the interaction with early environment may induce a pathophysiological cascade that leads to functional and morphological changes in the VP. The results of the present study showed an association of NI-induced decrease in the number of VGLUT2-expressing VP cells with impaired associative and implicit learning for dealing with aversive stimuli in adulthood, suggesting that NI increases susceptibility to LH in adulthood. Our findings support the involvement of VP maldevelopment in the pathophysiology of major depression in patients with a history of early environmental deprivation.
Associative and implicit learning, that is to say learning associations between goal representations and relevant environmental stimuli, are fundamental processes involved in mastering complex forms of cognition and learning. While the early development of associative and implicit learning processes depends on rich sensory and social inputs, it has recently been proposed that the lack of such inputs confers long-term difficulties in multiple domains of executive functioning [12,28,29]. NI is a means of depriving of animals (both the dam and littermates) of social contact in the early postnatal period. Thus, the present study suggests that NI likely induces inaccurate associative and implicit learning needed for coping with aversive stimuli in adulthood.
The neurobiological consequences of early postnatal deprivation such as the dysfunction of the hypothalamic-pituitary-adrenal axis in response to stress [29], alteration of gene expression through epigenetic mechanisms (e.g., DNA methylation) [30], and altered neurotransmission mediated by GABA [31] and glutamate [32] have been reported. In the present study, immunohistochemical analyses revealed that NI significantly decreased the number of VGLUT2-expressing VP cells, but not the number of Penk-expressing VP cells. The neurological development in the basal ganglia of rats was reported to be vulnerable to early adversities during the postnatal period [33,34]. Recently, Martin and Cork reported that the maturation of neuronal assemblies in the basal ganglia occurs postnatally via alteration in afferent innervation and gene expression to establish neurotransmitter/receptor phenotypes and form specific patterns of neuronal connectivity [35]. In addition, the glutamatergic innervation within the rat globus pallidus increases rapidly after birth and then decreases to adult levels over a period of weeks [36]. Taken together, these observations suggest that NI could cause a decrease in the number of the VGLUT2-expressing VP cells and this loss could persist until adulthood.
In addition, in NI rats, the number of VGLUT2-expressing VP cells was significantly decreased in the LH group compared with the non-LH group and significantly correlated with escape latency in the LH test. Although the VP has critical roles in motivated behavior [37] and the processing both rewards [38-42] and aversive stimuli [43-45], the considerable heterogeneity of neuronal cell types, projection patterns, and subregional circuits in the VP makes it difficult to associate the role of VP neurons with ultimate behaviors [20,21]. Recently, it was reported that selective activation of glutamatergic VP neurons projecting to the lateral habenula (LHb) using optogenetics induced behavioral avoidance in the real time place preference/avoidance test [17], and glutamate VP neurons increased the firing activity in the LHb, rostromedial tegmental nucleus neurons, and GABAergic and dopaminergic neurons in the ventral tegmental area (VTA) [46]. On the other hand, Macpherson and his colleagues found that Penk was expressed in GABAergic VP neurons rather than glutamatergic VP neurons, and that chemogenetic stimulation of Penk-expressing VP neurons disrupted the acquisition of inhibitory avoidance learning possibly through the downstream LHb-VTA circuit [18]. This evidence suggests the importance of fine tuning of the VP inhibitory–excitatory balance needed for avoidance learning, and in this study, the aberrant VP neuronal activity induced by decreased number of VGLUT2-expressing VP cells in NI rats may be, at least in part, associated with the increase in the rate of failure to escape traumatic shock in the LH test.
Noteworthily, there was no significant difference in the number of VGLUT2-expressing VP cells between the LH and non-LH subgroup of Sham rats unlike NI rats. Originally, the concept of LH referred to learning that responses are independent of their outcomes, moreover, separating the responses from their outcomes undermines trying to escape [47]. Previous studies on LH proposed several behavioral changes corresponding to functional changes in the neural circuitry including 1) passivity in response to prolonged aversive events – mediated by the serotonergic activity of the dorsal raphe nucleus; 2) learning to sense stressor controllability – mediated by activity of circuits in the ventromedial prefrontal cortex and the dorsal medial striatum; and 3) acting to overcome passivity and expecting the shock to be controllable in new aversive situations – mediated by activity in the ventromedial prefrontal cortex leading to dorsal raphe nucleus inhibition [48]. How the ventromedial prefrontal cortex - dorsomedial striatum circuit mediates learning to sense stressor controllability is not fully understood. The striatum is the part of the basal ganglia circuit that receives most of the cortical input. In addition, the striatum projects to other basal ganglia structures including the pallidum, and influences the executive functions (e.g., planning for movement and cognitive behavior) [49,50]. The striatal projections to the substantia nigra (the direct pathway) and pallidum (the indirect pathway) are especially relevant [51,52]. The indirect pathway is considered indispensable to aversive forms of learning [53]. Although the precise mechanism by which Sham rats exhibit diverse susceptibility to LH without manifesting a difference in the number of VGLUT2-expressing VP cells is not known, from the above evidence, it is reasonable to conclude that the diversities in other neurocircuitry may be, at least in part, involved in LH.
Several limitations warrant mention in the present study. First, the VP neuronal activity was not examined. Additional studies that include electrophysiological analysis are needed to strengthen the findings of the present study. Second, it is unclear when NI could decrease the number of VGLUT2-expressing VP cells. Further studies could reveal the appropriate period for the prevention of future pathophysiology induced by NI. Lastly, we evaluated only LH paradigm. The relationship between NI and major depression would be more elaborated by examining other depression-related behavioral tests (e.g., forced swim test, tail suspension test, sucrose preference test, social defeat paradigm).
In summary, the decrease in the number of VGLUT2-expressing VP cells in adulthood in response to NI is associated with inaccurate associative and implicit learning for processing aversive stimuli, and subsequently increases susceptibility to LH. It was supposed that the decrease in VGLUT2-expressing VP cells persists from the neonatal period. The occurrence of LH without the decreased VGLUT2-expressing VP cells in Sham rats indicated the involvement of different behavioral and neurocircuitry deficits in LH. Based on these findings, we postulate that the aberrant VP neuronal activity due to environmental deprivation early in life leads to pessimistic associative and implicit learning. Future studies examining the relationship between the aberrant VP activity in patients with major depression having a history of environmental deprivation and pessimistic features are needed to develop novel therapeutic and preventive strategies for patients with specific pathophysiology.
This study was supported by Japan Society for the Promotion of Science (JSPS) KAKENHI (a grant-in aid for Scientific Research, C) Grant Number JP23K07014. The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
No potential conflict of interest relevant to this article was reported.
Methodology: Hironori Kobayashi, Tatsuhiro Miyagi, Yasumasa Okamoto. Investigation: Hironori Kobayashi, Kenichi Oga, Tatsuhiro Miyagi, Sho Fujita, Satoshi Fujita, Satoshi Okada. Visualization: Hironori Kobayashi, Tatsuhiro Miyagi. Conceptualization: Manabu Fuchikami, Shigeru Morinobu. Funding acquisition: Manabu Fuchikami. Project administration: Manabu Fuchikami. Supervision: Yasumasa Okamoto. Writing—Original Draft: Hironori Kobayashi. Writing—Review & Editing: Manabu Fuchikami, Shigeru Morinobu.
![]() |
![]() |