Pruning is a process in brain growth periodization. which of the following is true?

2.1. Neuronal Morphology following Ethanol Administration

To investigate ethanol-induced alterations in MSNs, neuronal morphology was measured 24 h, 1 month, and 5 months after acute ethanol exposure during synaptogenesis. Representative Neurolucida tracings of MSNs from control and ethanol-treated mice are shown in Figure 1A,B, respectively. As depicted in Figure 2, acute high-dose ethanol administration during the early postnatal period in mice significantly increased MSN dendritic length after 24 h [p = 0.027], but this difference appears to be normalized by adulthood [5 months; p = 0.384]. Furthermore, ethanol-treated MSNs had significantly more branching nodes after 24 h [p = 0.018], but by the 1-month timepoint, there was no longer a significant difference in nodes [p = 0.868]. For soma size, the small sample and effect size provides insufficient statistical power to determine if there is an effect of ethanol at any timepoint examined [Figure 2J–L]. Despite the observed increases in dendritic complexity, ethanol treatment does not significantly increase the actual number of MSNs dendrites at 24 h [though it trends toward significance at p = 0.065]. At 1 month of age, ethanol does significantly increase the number of dendrites [p = 0.002]. Though the number of dendrites appeared to normalize at 5 months, the limited sample size for this study provides insufficient statistical power to detect ethanol effects on the number of dendrites at this time point.

2.2. Dendritic Spine Analysis following Ethanol Administration

To determine whether ethanol treatment affects dendritic spine number, length, or type, confocal microscopy was used to analyze Golgi-stained brain sections from ethanol and saline-treated mice. As shown in Figure 3, ethanol treatment significantly decreased the MSN dendritic spine number at 24 h, but increased the number of spines by 1 month of age. Though these spine differences appear to resolve by 5 months of age, there is insufficient statistical power to deduce ethanol’s impact in adulthood [Figure 3A,D,G]. We did not detect a significant difference in spine length, but there is insufficient statistical power to determine the relationship between treatment and spine length in all but the 1-month post exposure group [p = 0.944], which showed no difference [Figure 3B,E,H]. We also did not detect a significant difference in spine type between ethanol and saline treated groups, but there is insufficient statistical power to determine the impact of ethanol treatment on spine type [Figure 3C,F,I].

2.3. Protein Analysis

To investigate the mechanism by which ethanol is causing such an acute change in MSN dendritic morphology, simple Western blots were used to assess protein levels in striatal samples of saline and ethanol-treated animals 24 h after ethanol administration. Because microglia play an important role in synaptic pruning, levels of the microglia recruitment marker, fractalkine, were analyzed. Fractalkine is expressed by neurons and received by microglia and is known to play a role in the developmental pruning of neurons. Therefore, significant changes in fractalkine level could be responsible for the ethanol-induced changes in spine number. However, as shown in Figure 4A, no differences in fractalkine level were observed following ethanol treatment [p = 0.477].

As an alternative mechanism for the ethanol-induced dendritic branching, the levels of Hsp70 were examined. Hsp70 is known to facilitate the formation of cytoskeletal components, protect them during stress, and impact neurite outgrowth. As shown in Figure 4B, although ethanol did not produce a statistically significant difference in Hsp70 levels, there is a marginal trend towards being elevated 24 h after ethanol treatment [p = 0.084].

Because spine numbers are reduced 24 h after ethanol administration, and glutamatergic synapses are found on dendritic spines, the localization of postsynaptic density-95 [PSD-95] was assessed as a marker of active glutamatergic inputs. PSD-95 accumulates at synapses, thereby forming distinct puncta that are visible using immunofluorescence. As shown in Figure 4D–F, total levels of PSD-95 protein are increased 24 h following ethanol treatment in the caudate/putamen, both by simple and standard Western blot technique. Beta-tubulin levels were not affected by ethanol treatment [Figure 4C], and were therefore, used as a loading control to normalize the signals of other proteins. Ethanol treatment also increased the number of PSD-95 puncta in the striatum 24 h after administration [461.46 ± 58.38 saline, 984.17 ± 10.0 ethanol, p = 0.010, Figure 5]. For immunofluorescence experiments, FoxP1 was used as a marker to identify MSNs and verify that puncta were counted in the correct cell population. This indicates that the morphological differences identified at the 24 h time point correlate with increased levels of the synaptic marker, PSD-95, in the striatum.

3.1. Dendrogenesis and HSP-70

MSNs are remarkably plastic, even in adult animals, especially in the morphology of their dendrites [12]. However, the mechanism for how ethanol causes such an acute change in MSN dendritic morphology is not known. One explanation for our arborization findings could be that ethanol induces the branching within the dendrites. Neurite outgrowth and spine morphogenesis are mediated by reorganization of actin components and the microtubule cytoskeleton. As such, we examined levels of Hsp70, which is a heat shock protein [70 kDa] able to facilitate the formation of cytoskeletal components and protect them during stress. In the postnatal period, heat shock proteins continue to protect the developing CNS from unfavorable conditions and are able to impact neurite outgrowth [13].

Although we found no statistically significant difference in HSP70 levels, there is a marginal trend towards being elevated 24 h after ethanol treatment. Normally, HSP70 is present at the synapse, associated with the post synaptic density, and can colocalize with PSD95 [14], but Hsp70 also localizes to the synapse upon stress [15]. Acute administration of 5 g/kg ethanol results in an increased level of Hsp70 in various brain regions in rats, including the striatum [16]. Although the specific function of the PSD-associated HSP-70 is not known, it could possibly facilitate local synthesis of the new proteins required for neural remodeling and neurite outgrowth or stabilize synapses during periods of structural plasticity, including that seen immediately after ethanol administration. Changes here likely involve actin/microtubule regulator molecules, although the mechanisms by which immediate changes occur after ethanol administration are still under investigation, and further research is warranted on heat shock family proteins.

3.2. Fractalkine Signaling and Neuronal Pruning

Another possible explanation for the observed branching differences is that ethanol could inhibit a natural pruning process that takes place during neurodevelopment. To investigate a potential change in the pruning of neurons, we examined a marker of microglia recruitment. The microglia play a key role in a number of important processes in the developing brain, including dendritic spine pruning and the clearing of apoptotic cells during development [17,18]. However, it has been established in the literature that ethanol treatment activates microglia [19,20], so instead, we assessed levels of fractalkine, a molecule expressed by neurons and received by microglia, which plays a normal role in the developmental pruning of neurons and can recruit microglia to a particular location. Not only is fractalkine signaling a mediator of microglia–neuron communications during synaptic plasticity, but fractalkine is also released after neuronal stress and has a role in the response to ethanol injury [21].

During the postnatal period of development, fractalkine is responsible for circuitry refinement via synaptic engulfment and timing of microglia recruitment, via glutamatergic synapse formation and delayed microglia-synapse interaction during development [22,23,24]. Although ethanol can suppress the expression of the fractalkine ligand CX3CL1 in microglia in a neonatal mouse model of FASD [25], our striatal protein quantification found no significant differences in fractalkine levels. Our results align with a similar previous study that found no changes in fractalkine mRNA expression in a similar postnatal ethanol paradigm [21].

3.3. PSD-95 as a Neuroprotector

We next investigated the role of PSD-95 as both a neuroprotective molecule and as a potential marker for active new synapses on these dendritic branches. Proteins such as PSD-95 that are primarily involved in synaptic functions can also play developmental roles in shaping branch patterns. PSD-95 is a postsynaptic scaffolding protein that acts as an organizer for signaling components at the synapse, but also promotes dendritic spine maturation [26,27] and regulates synaptic transmission [28].

Our results show that ethanol increases PSD-95 levels in the striatum after 24 h. This is consistent with previous research, which found that spine reduction may actually result in an enhancement of the amount of PSD-95 present on the dendritic shaft [29]. Our results are also in line with a similar study that found an immediate early rise in PSD-95 levels after a single acute ethanol exposure [30].

As a synaptic protein, PSD-95 can balance the interplay between the dopamine and glutamate systems in synapses by regulating the localization, trafficking, and stabilization of both glutamatergic NMDA receptors and dopamine receptors [31]. Dopamine alters the sensitivity of other neurons to neurotransmitters, particularly glutamate [32]. Overactivity of dopamine D1 receptors and glutamatergic NMDA receptors can lead to NMDAR-mediated excitotoxicity, causing cell death and neurological phenotypes. PSD-95 can act as a neuroprotector in the striatum that can limit D1 activity and inhibit the crosstalk between D1 and NMDA receptors to limit the effects of excitotoxicity.

PSD-95 knock-out mice show striatal NMDA receptor hyperactivity and MSN cell death [33], as well as increased sensitivity to the sedative/hypnotic effects of ethanol [34]. Therefore, it is possible that the increased PSD-95 seen in our model is an immediate neuronal compensatory response to the ethanol, possibility elevated to protect against excitotoxicity effects in the striatum.

3.4. PSD-95 in Development

In its developmental role, PSD-95 can shape the dendritic arbor and spines via microtubule-dependent mechanisms [35] in a manner specific to immature neurons [36]. Research suggests that as the neuron matures, increasing levels of PSD-95 provide an environment that is more friendly to the development of new spines than new branches, and dramatic neuronal morphology changes become more infrequent [37]. In addition, overexpression of PSD-95 increases spine maturation [26] which collectively supports a role for PSD-95 during multiple phases of neuronal development, all of which could be affected by ethanol-induced expression changes, depending on the exposure window [38].

In light of this, it is likely that PSD-95 plays a role in the changes in spines that we see, but is unlikely to be the mechanism causing the increased immediate neuronal branching identified after 24 h. In fact, high PSD95 expression levels decrease branching, so the mechanism underlying the immediate branching phenotype must be robust enough to counter the PSD95 role in decreasing neurite outgrowth. An important next research step will be to isolate ethanol’s effect on the developmental role of PSD-95 from PSD-95′s more immediate role in synaptic function. In addition, an examination of later timepoints could provide information about how long the PSD-95 elevation persists.

3.5. Adolescence and Ethanol Withdrawal Effects

At 1 month of age [3 weeks of withdrawal], we observed an increase in MSN dendrite number and spine number in ethanol treated animals. Previous research has demonstrated a pattern of increased dendritic modifications after ethanol withdrawal using models of chronic ethanol exposure. Researchers found that after 5 months of chronic ethanol administration and 2 months of withdrawal, dendrite segments increased in hippocampal CA1 neurons as the cells compensated for the ethanol-induced developmental retardation [39]. A separate study found an immediate decrease in spines in the nucleus accumbens following one day of alcohol withdrawal after chronic administration of ethanol in rats, followed by increased dendritic length in the context of unaltered spine density in nucleus accumbens core MSNs at 7 days of withdrawal. Spine loss may represent ethanol-dependent pruning of spines, where most active synapses are retained following ethanol exposure [40].

The processes occurring during ethanol intoxication are different from those during the withdrawal period. Even though acute doses of ethanol inhibit NMDA receptors in adults, neurons can adapt by increasing the sensitivity or number of NMDA receptors. Once the ethanol is metabolized, this receptor increase can result in excitotoxicity. Early postnatal ethanol exposure in rodents has been shown to increase markers for glutamatergic activity [NMDA1 receptor] in the prefrontal cortex, dorsal striatum, somatosensory cortex, and motor cortex during acute withdrawal [22 or 26 h post ethanol administration] [41]. During normal development, there is a sharp increase in vulnerability to NMDA receptor-mediated excitotoxicity in the third trimester equivalent [P6–P7 in rats], and the associated behavioral deficits can be ameliorated with the administration of an NMDA antagonist 21–33 h after ethanol administration. This explains how ethanol-induced cell death can be excitotoxic, even though ethanol itself inhibits the excitatory activity of the NMDA receptor.

3.6. The Importance of Arborization within the Striatum

Our results offer a new perspective on the neural response to ethanol, as we show that neurons exhibit an immediate response to the ethanol intoxication where spines decrease, but dendrites are more arborized. These findings open multiple avenues for further research to determine exactly how and why the neurons respond in this way. We know that the brain is most sensitive to injury during periods of intensive growth [42], and so it is also important to consider the effect that these structural changes will have on cell and circuit functionality.

Although MSN morphology appears to be largely corrected in our paradigm by adulthood, even a transient window of dysregulation in neuroarchitecture could have enormous impacts on the animal, particularly in the striatum, at this developmental timepoint. The striatum is an important signaling hub of the brain, collecting information and distributing it to a number of targets. It receives extensive glutamatergic inputs from the cortex, thalamus, hippocampus, and amygdala. It integrates this information and sends it to the globus pallidus, subthalamic nucleus, substantia nigra, and ventral tegmental area [VTA] [reviewed in [43,44]]. The majority of alcohol research within the striatum focuses on the ventral striatum because of its key role in addiction, but we chose to examine the caudate and putamen within the dorsal striatum. While the ventral striatum is involved with cued outcome-mediated behaviors, the dorsal striatum makes contributions to goal-directed and habitual reward-seeking behavior [45].. Acute ethanol can change GABAergic transmission in this area [46], with a net effect of increasing output from the sensorimotor striatum, possibility initiating habit formation and having additional implications for addictive behaviors [8].

Before the dendritic arbor stabilizes in the mature animal, large-scale morphological changes occur in neurons during the first weeks of postnatal development, characterized first by growth of dendritic branches and followed by the pruning of excessive or mistargeted branches [36]. This branched morphology of the neuronal dendritic arbor allows for the induction of synaptic plasticity and enables information to enter, be integrated, and be sent to the rest of the neural circuit. Proper regulation of these processes is crucial for neuronal function and synaptic transmission, and as such, abnormal development of the dendritic arbor contributes to a host of brain disorders [47]. We also know that spines can be differentially regulated by neuronal activity; specifically, dendritic shaft filopodia elongate in response to focal glutamate application [48], which is particularly relevant for the striatum. Alterations in the number and shape of spines contribute to aberrant synaptic and neural signaling, since the distribution pattern of synapses and the properties of each synapse will determine the connections in and out of each MSN [38,49].

In many neurons, the location of a synaptic conductance within the dendritic tree has the ability to profoundly influence the timing and amplitude of the somatic EPSP [50]. Therefore, even in the absence of gross morphology, deficits may be present in high-level sensory and motor processing, as well as complex cognitive thought and behavior.

3.7. The Importance of the MSN Early Postnatal Period

Our chosen developmental timepoint is particularly important for immature MSNs, as many MSN electrophysiological characteristics change abruptly between mouse postnatal days 6 and 10. Research indicates the presence of a developmental program that switches MSN activity from an immature low threshold activation state to a high threshold state in the adult during resting conditions that coincides with the emergence of locomotion. This “adult” state is also mirrored in the development of the MSN arborization/spines, an increase in glutamatergic synapses, and the solidifying of nigrostriatal dopaminergic inputs [51]. Of note, these patterns are similar to those seen in the developing cortex, so it is likely that other brain areas have comparable mature patterns taking over at timepoints appropriate to support each brain area’s main function.

Just as alcohol impacts the developmental events that occur at the time of administration, learning and memory processes are also limited by the state of the neurons during the time that acquisition is happening. Among other things, the dorsal striatum is important for the consolidation of memory learning that accrues over many trials [52]. Our lab has investigated both social behavior and learning/memory in mice treated with this acute ethanol developmental paradigm. We found no overt social differences, but the animals exhibit memory deficits on the Barnes maze when learning acquisition occurs during adolescence and retention is tested in adulthood [7,53] This acquisition deficit fits well with our finding of PSD-95 abnormalities, as PSD-95 also appears to be important in coupling the NMDA receptor to pathways controlling synaptic plasticity and learning [28]..

The dendrogenesis and spinogenesis observed immediately after an intoxication event indicate an immediate response to ethanol, but morphological changes also persist into adolescence. Therefore, brief acute exposure to a high ethanol dose during the third trimester equivalent of human pregnancy may have a permanent negative impact on not just the morphology of neurons, but also the behavior and neural functioning of the offspring in adulthood.

4.1. Animal Care

Protocols for animal use were approved by the Institutional Animal Care and Use Committee at Hampden-Sydney College. All mice were offspring from timed pregnant C57BL6/J females [Jackson Laboratories], received on the same gestational day and subsequently housed in a temperature-controlled room under 12-h light/dark cycles [lights off at 6 p.m.], with ad libitum access to food and water. A total of 6 C57/Bl6 timed pregnant females were ordered from Jackson labs. Each pup within a litter was randomly assigned to either the saline or ethanol treatment. Cumulatively for all experiments, n = 19 saline pups and n = 20 ethanol pups were assessed.

4.2. Ethanol Treatment Paradigm

C57BL6/J mice were exposed to a high acute ethanol dose on postnatal day 6 [P6] by dorsal subcutaneous injection of a 20% ethanol solution [2.5 g/kg dose] or saline control, followed by a second injection two hours later [7]. Pups were individually randomly assigned a treatment condition and were returned to their mothers between and after injections.

When a 2.5 g/kg dose is repeated subcutaneously after 2 h, as in our paradigm, C57BL/6 pups achieve a mean blood ethanol content [BEC] 1 h after the second injection of 472 [±6] mg/dL, with an alcohol clearance rate of 283 mg/dL/h. This blood ethanol level is within the range that a human fetus might be exposed to after maternal ingestion of a moderate to heavy dose of ethanol [3].

Mice were anesthetized deeply with isoflurane, followed by cervical dislocation, at either 24 h, 1 month, or 5 months following exposure. Brains were harvested and halved at the longitudinal fissure. One entire hemisphere was flash frozen using isopentane and frozen at −80 °C until use, and the striatum was dissected out from the other hemisphere before freezing. The 24-h time point animals came from 4 litters of timed-pregnant females injected at P6 and euthanized at P7. The one- and five-month time point animals came from 4 separate litters injected at P6 and euthanized at P30 and 5 months, respectively.

4.3. Golgi Staining and Neuronal Morphology

Twenty-four-hour brains were coronally sectioned using a sliding microtome at 60 µm and Golgi processed as described in Brunjes and Kenerson [54]. One-month and 5-month brains were sectioned using a cryostat at 60 µm and Golgi processed using the FD GolgiStain kit [FD Neurotechnologies, Inc., Columbia, MD, USA].

Camera Neurolucida [2015] tracings were prepared using the 40× lens. For analysis, all selected neurons had to meet rigid criteria: the neuron had to be well-impregnated, the branches could not be obscured by other neurons, or other precipitates, there could be no visibly broken processes on the neuron, it had to be located within the caudate/putamen region of the striatum [between Bregma +1.54 and −1.94 mm], and it had to visually resemble a typical medium spiny neuron.

For neuronal morphology, between 6 and 8 medium spiny neurons in the caudate-putamen region of the striatum were traced in 3–5 animals at each treatment level [n = 4 saline and n = 4 ethanol at 24 h; n = 3 saline and 3 ethanol at 1 month; n = 3 saline and n = 5 ethanol at 5 months].

Following the tracing, the NeuroExplorer program was used to quantify the morphological data. Four parameters were assessed, including the number of dendrites, number of nodes, total length of dendrites per neuron, and cell body size. The number of dendrites was defined as the number of projections that were extending directly from the soma. The number of nodes was the count of the branch points on those dendrites. Dendritic length was the total length of all projections outside the soma.

4.4. Dendritic Spine Analysis

Image stacks were taken on a Zeiss confocal scope using the 100× objective lens. For each animal, a section of apical dendrite was measured on first order branches [located immediately between the soma and the first dendrite branch point] on 4 different neurons. The average number of spines were counted [n = 5 saline and 4 ethanol at 24 h; n = 3 saline and n = 3 ethanol at 1 month; n = 3 saline and n = 4 ethanol at 5 months]. All visible spine heads protruding from the apical dendrite were counted beginning at 40 µms from the soma and ending at 80 µms from the soma to obtain a spine density. The length of individual spines was measured from the distal end of the spine to the point where the spine attached to the dendrite and an average spine length was calculated for each neuron. Spines were simply classified as [1] filipodia >2 µm or [2] non-filopodial spines