Serum- and Glucocorticoid-Inducible Kinase 1 Activity Reduces Dendritic Spines in Dorsal Hippocampus

Abstract

The hippocampus has a well-known role in mediating learning and memory, and its function can be directly regulated by both stress and glucocorticoid receptor activation. Hippocampal contributions to learning are thought to be dependent on changes in the plasticity of synapses within specific subregions, and these functional changes are accompanied by morphological changes in the number and shape of dendritic spines, the physical correlates of these glutamatergic synapses. Serum- and glucocorticoid-inducible kinase 1 (SGK1) regulates dendritic spine morphology in the prefrontal cortex, and modulation of SGK1 expression in mouse hippocampus regulates learning. However, the role of SGK1 in dendritic spine morphology within the CA1 and dentate gyrus regions of the hippocampus are unknown. Thus, herpes simplex viral vectors expressing GFP and various SGK1 constructs, including wild type SGK1, a catalytically inactive version of SGK1 (K127Q), and a phospho-defective version of SGK1 (S78A), were infused into the hippocampus of adult mice and confocal fluorescent microscopy was used to visualize dendritic spines. We show that increasing expression of SGK1 in the dentate gyrus increased the total number of spines, driven primarily by an increase in mushroom spines, while decreasing SGK1 activity (K127Q) in the CA1 region increased the total number of dendritic spines, driven by a significant increase in mushroom and stubby spines. The differential effects of SGK1 in these regions may be mediated by the interactions of SGK1 with multiple pathways required for spine formation and stability. As the formation of mature synapses is a crucial component of learning and memory, this indicates that SGK1 is a potential target in the pathway underlying stress-associated changes in cognition and memory.

Introduction:

The hippocampus is critical for learning and creating, consolidating, and recalling memories [12]. Alterations in hippocampal function are implicated in numerous diseases, including Alzheimer’s disease, depression, addiction, and cognitive impairment [2, 22]. The hippocampus is composed of subregions connected through a well-documented excitatory glutamatergic synaptic loop [12]. Information from the neocortex is relayed to the hippocampus via the entorhinal cortex (EC), which synapses onto the dentate gyrus (DG) granule cells via the perforant pathway. The DG neurons project onto the CA3 pyramidal neurons via the mossy fiber pathway, and CA3 neurons in turn send excitatory projections to CA1 pyramidal neurons via the Schaffer collaterals. Additional projections, including direct input from the EC to CA1, and from CA3 back to DG, are integrated into the feed-forward loop to fine-tune information being processed by the hippocampus.

Plasticity of hippocampal synapses is thought to be a biological correlate of learning and memory. Activity-dependent changes in synaptic strength, such as long-term potentiation (LTP) and depression (LTD), occur via distinct molecular mechanisms in the various hippocampal subregions. In the CA1, maturation of synaptic connections is largely dependent on increased insertion of postsynaptic α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors triggered by activation of N-methyl-D-aspartate (NMDA) receptors [32]. However, LTP at synapses between the medial perforant path and DG granule cells relies on brain-derived neurotrophic factor (BDNF) signaling, which triggers activation of ERK to initiate translation of proteins including Arc [4, 23]. LTP in the DG is uniquely contingent on Arc-dependent consolidation; Arc is increased for hours in the DG for after LTP, but for only minutes in other areas including the CA1[39]. Retrograde BDNF signaling and presynaptic alterations such as increased neurotransmitter vesicle docking are also important for plasticity in the DG [4], while presynaptic mechanisms are unlikely to be associated with LTP at CA1 synapses [32].

Glutamatergic presynaptic elements often synapse onto specialized compartments extending from postsynaptic dendrites called dendritic spines. In both the CA1 and DG, changes in plasticity and maturity of synapses are associated with morphological changes in these spines, and structural abnormalities in hippocampal spines are associated with diseases affecting memory and cognition, including Alzheimer’s, Fragile X syndrome, and chronic stress [7, 21, 29, 36]. Spines are highly dynamic, changing morphology, number, density, and motility in seconds, likely impacting neuronal signaling [29]. Spines with a low head-to-neck ratio, known as thin spines, are more mobile, transient, and are generally associated with weaker synapses, while those with a greater head-to-neck ratio, known as mushroom spines, are more stable and have larger post-synaptic densities with more AMPA receptors, and thus greater synaptic strength [29]. Indeed, induction of LTP and increased synaptic strength are associated with greater spine density and spine head diameter [16]. The molecular mechanisms governing hippocampal synaptic strength and spine structure are complex and vary with cell type and region, but signaling through changes in protein phosphorylation is central to these processes throughout the hippocampus.

In the hippocampus, serum- and glucocorticoid-inducible kinase 1 (SGK1) a member of the AGC (PKA-, PKG-, PKC-related) family of serine/threonine kinases, is important for learning and memory consolidation. Like other AGC family members, SGK1 phosphorylation at two canonical sites (threonine 256, T256 and serine 422, S422) increases its catalytic activity [15, 37] though it also can be phosphorylated at a site in the N-teminus (serine 78, S78) [19, 27]. SGK1 phosphorylation and protein expression is increased in the hippocampus following fear conditioning, spatial learning and environmental enrichment [26–28, 33] and hippocampal SGK1 overexpression improves spatial learning [25, 28, 41, 42], while hippocampal transfection of phospho-defective SGK1 mutants leads to impaired spatial learning, fear conditioning, and novel object recognition [26–28, 41]. However, these effects are dependent on phosphorylation site, as SGK1 phospho-mutants directly linked to impaired catalytic activity (S422A and T256A) are sufficient to impair spatial learning while interference with N-terminal phosphorylation (S78A) is not [27, 41]. In contrast, impaired N-terminal (S78A) and canonical (S422A) site phosphorylation both impact fear conditioning [26, 28]. SGK1 activation increases PSD-95 expression in the hippocampus and inhibition impairs expression (but not induction) of LTP, consistent with increased SGK1 phosphorylation (S422) after tetanus protocols that induce LTP [33]. Thus, SGK1 appears to be a critical player in functional synaptic plasticity in the hippocampus.

In the prefrontal cortex (PFC), decreasing SGK1 activity reduces spine density and synaptic strength [30] and stress-induced increased SGK1 activity leads to increased surface AMPAR and NMDAR expression and potentiates glutamatergic transmission [44]. Moreover, transfection of constitutively active SGK1 in cultured spinal cord neurons promotes increased primary dendrites and branches [11]. However, it is unclear whether SGK1 effects the morphology of dendritic spines in hippocampus. Thus, we overexpressed various SGK1 constructs to evaluate the role of SGK1 in the density and morphology of dendritic spines in DG and CA1 of the dorsal hippocampus. We found that SGK1 differentially affects spine number and morphology in CA1 pyramidal neurons compared to granule cells of the dentate gyrus, indicating a complex role for SGK1 in the synaptic connectivity of the hippocampus.

Methods:

Animals and surgery

All experimental procedures were approved by the Institutional Animal Care and Use Committee at Michigan State University and performed in accordance with AAALAC and NIH guidelines. Adult mice (male, C57BL/6J, 7–8 weeks old, 20–30 g) were purchased from Jackson Labs. Mice were anesthetized with a mixture of ketamine (100 mg/kg) and xylazine (10 mg/kg). Thirty gauge needles (Hamilton Company) were bilaterally placed using two sets of stereotaxic coordinates that both target the CA1 and Dentate Gyrus (DG) of the dorsal hippocampus (Fig 1A).

Figure 1. Injection sites and viral vector validation.

A. Micrograph of hippocampus from mouse injected with herpes simplex viral vector showing GFP expressing CA1 and dentate gyrus neurons (left) and a single CA1 pyramidal neuron expressing GFP with dendritic arbor clearly visible (right). B. Micrograph showing groups of virally transduced hippocampal CA1 neurons expressing GFP (green) and SGK1 (red) demonstrating specific SGK1 overexpression with novel mutant vectors K127Q and S78A. C. Illustration depicting the three main classes of dendritic spines: thin, stubby, and mushroom. Mature mushroom spines contain more glutamate receptors (blue) while immature spines contain fewer.

10° angle from the midline −2.2 mm anteroposterior, +/− 2.5 mm mediolateral. Purified high-tilter virus (0.6 μL) was infused separately (0.3 μL/infusion) over 5 min periods at two sites: −2.1 mm dorsoventral (DV) and −1.9 mm DV.

7° angle from the midline −2.2 mm anteroposterior, +/− 2.0 mm mediolateral. Purified high-tilter virus (0.5 μL) was infused separately over 5 min periods at two sites: 0.2 μL at −2.0 mm DV and 0.3 μL at −1.8 mm DV.

Needles remained at the injection site for 5 min to allow diffusion of virus following injection. Proper targeting of CA1 and DG was confirmed post hoc by slicing brains and confirming location of GFP expression (see below). Previously validated [13, 20] viral vectors included the following: herpes simplex viral (HSV) vectors expressing GFP alone (HSV-GFP), HSV expressing GFP and wild type SGK1 (HSV-SGK1). Novel vectors validated here (Fig 1B) include HSV vectors expressing GFP and a catalytically dead, dominant-negative SGK1 mutant, K127Q [6, 24] (HSV-K127Q), or a N-terminal phospho-defective SGK1 variant[19, 27] (HSV-S78A). All SGK1 constructs contain an N-terminal HA-tag [20, 37], allowing validation of SGK1 overexpression (Fig 1B). Vectors were prepared by the Massachusetts General Hospital Viral Vector Core.

Immunohistochemistry

Adult C57Bl6J mice were perfused transcardially with ice-cold PBS followed by 10% formalin. Brains were post-fixed 24 hours in 10% formalin, cryopreserved in 30% sucrose, and then sliced into 35 μm sections. Immunohistochemistry was performed using a rabbit anti-HA antibody to detect virally overexpressed SGK1 (Cell Signaling C29F4, Lot 9) and a goat anti-GFP antibody (Abcam Ab5450) and secondary antibodies: Alexa-Flour 488 anti-goat IgG (Jackson; 705-545-147 1:200) and Cy3 anti-rabbit IgG (Jackson; 711-165-152 1:200).

Spine Analysis

72 hours after stereotaxic surgery, mice were perfused transcardially with ice-cold PBS followed by 10% formalin. Brains were post-fixed 24 hours in 10% formalin, and then sliced into 100 μm sections on a vibrating microtome (Leica). Sections were mounted on slides using Vectashield Hard Set mounting medium (Vector Laboratories, Inc.). Microscopy was performed at the Center for Advanced Microscopy at Michigan State University. GFP fluorescence was visualized using an Olympus FluoView 1000 filter-based laser scanning confocal microscope. Z stacks were taken at 0.5 microns per slice, laser voltage was 600–900V, and offset (contrast) was set at 12%.

Spines were analyzed essentially as described previously [7, 13]. Morphological analysis was performed using NeuronStudio with the rayburst algorithm, and investigator was blind to conditions. Dendritic segments 50–150 μm away from the soma were chosen randomly from HSV-transduced cells expressing GFP and satisfying the following requirements: (1) the segment had to be completely filled (all endings were excluded), (2) the segment must be at least 50 μm from the soma, and (3) the segment could not be overlapping with other dendritic branches[38]. An average of 3 slices per animal per region were photographed, depending on the availability of high-quality neurons (n=23 mice; n= 9 mice for the GFP group, n= 4–5 mice per all other groups). An average of 4 dendrites per slice were analyzed in the DG, and an average of 6 dendrites per slice were analyzed in the CA1. Thus, ~9000 dendritic spines per experimental group were analyzed for DG studies, and ~18,000 dendritic spines per CA1 experimental group were analyzed, totaling over 125,000 spines analyzed. NeuronStudio classifies spines as thin, mushroom, or stubby (see Fig 1C, Fig 2A, and Fig 3A for examples) based on the following values: (1) aspect ratio; (2) head/neck ratio; and (3) head diameter.

Figure 2. SGK1 regulates hippocampal CA1 spine formation.

A. Example images showing: Top – GFP CA1 dendritic spines. Upper Middle – SGK1 CA1 dendritic spines. Lower Middle – S87A CA1 dendritic spines. Bottom – K127Q CA1 dendritic spines. Brown indicates mushroom spines, yellow indicates thin spines, pink indicates stubby spines. B. Quantification of total spines in CA1 neurons shows K127Q had significantly more total spines than all other groups: GFP alone, SGK1, or S78A. The K127Q group also demonstrated significantly greater expression of stubby spines as compared to the GFP control group and the SGK1 overexpression group as well as significantly greater expression of thin spines as compared to the S78A group. The K127Q group also demonstrated a significantly large increase in mushroom-shaped spines as compared to all other groups. Error bars indicate mean ± SEM; * = p<0.05, ** = p<0.01, **** = p<0.0001.

Figure 3. SGK1 regulates hippocampal DG spine formation.

A. Example images showing: Top – GFP DG dendritic spines. Upper Middle – SGK1 DG dendritic spines. Lower Middle – S78A DG dendritic spines. Bottom – K127Q DG dendritic spines. Brown indicates mushroom spines, yellow indicates thin spines, pink indicates stubby spines. B. Quantification in DG neurons shows no difference in total or stubby spines between groups. The wt-SGK1 overexpression group had significantly less thin spines than the S78A group but significantly more mushroom spines than the GFP control. Error bars indicate mean ± SEM; * = p<0.05, ** = p<0.01.

Statistics

All statistical analyses were performed using Prism software (GraphPad). Relative spine densities were analyzed using a between-subjects one-way ANOVA, followed by Tukey post hoc comparisons. Because these experiments examine the effects of SGK1 manipulation on the dendritic spines on individual cells, and in keeping with previous publications from our group and many others, we performed statistical analyses using the individual dendrite (one measured per cell) as our (n) unit of observation (Table 1), and our interpretations are based on these data. Reported P values reflect Tukey’s post hoc tests adjusted for multiple comparisons. We also performed statistical analyses using the average spine density for individual animals as the(n) unit of observation (Table 2).

Table 1: Statistical analyses of spine number and type.

Spine data from CA1 and DG were analyzed separately using four one-way ANOVAs for each region: one for total spines, and one for each spine subtype. Post hoc multiple comparisons of all groups were performed using Tukey test. Significant differences are highlighted in bold text.

Figure	Comparison	Cells (n)	F (DFn, DFd)	ANOVA P	Tukey’s Adjusted P
2B - CA1 Total Spines	Overall	337	17.19 (3, 333)	<0.0001
	GFP vs SGK1				0.8801
	GFP vs S78A				0.9995
	GFP vs K127Q				<0.0001
	SGK1 vs S78A				0.9251
	SGK1 vs K127Q				<0.0001
	S78A vs K127Q				<0.0001
2B - CA1 Stubby Spines	Overall	337	4.444 (3, 333)	0.0044
	GFP vs SGK1				0.9991
	GFP vs S78A				0.9117
	GFP vs K127Q				0.0067
	SGK1 vs S78A				0.9643
	SGK1 vs K127Q				0.0235
	S78A vs K127Q				0.0614
2B - CA1 Thin Spines	Overall	337	3.776 (3, 333)	0.0109
	GFP vs SGK1				0.9921
	GFP vs S78A				0.7760
	GFP vs K127Q				0.1069
	SGK1 vs S78A				0.9312
	SGK1 vs K127Q				0.0861
	S78A vs K127Q				0.0096
2C - CA1 Mushroom Spines	Overall	337	30.52 (3,333)	<0.0001
	GFP vs SGK1				0.4728
	GFP vs S78A				0.9462
	GFP vs K27Q				<0.0001
	SGK1 vs S78A				0.2245
	SGK1 vs K127Q				<0.0001
	S78A vs K127Q				<0.0001
3B - DG Total Spines	Overall	242	0.468 (3, 238)	0.7047
	GFP vs SGK1				>0.9999
	GFP vs S78A				0.9492
	GFP vs K127Q				0.7174
	SGK1 vs S78A				0.9507
	SGK1 vs K127Q				0.7506
	S78A vs K127Q				0.9891
3B - DG st.oby Spines	Overall	242	2.610 (3, 238)	0.0521
	GFP vs SGK1				0.0791
	GFP vs S78A				0.2863
	GFP vs K127Q				0.9988
	SGK1 vs S78A				0.9965
	SGK1 vs K127Q				0.4576
	S78A vs K127Q				0.5396
3B - DG Thin Spines	Overall	242	2.962 (3, 238)	0.0329
	GFP vs SGK1				0.2761
	GFP vs S78A				0.3625
	GFP vs K127Q				0.7590
	SGK1 vs S78A				0.0487
	SGK1 vs K127Q				0.1560
	S78A vs K127Q				0.9246
3B - DG Mushroom Spines	Overall	242	4.960 (3, 238)	0.0023
	GFP vs SGK1				0.0014
	GFP vs S78A				0.9970
	GFP vs K127Q				0.8237
	SGK1 vs S78A				0.0752
	SGK1 vs K127Q				0.3233
	S78A vs K127Q				0.8626

Table 2: Statistical analyses of spines by animal.

Spine data from CA1 and DG were analyzed separately using four one-way ANOVAs for each region: one for total spines, and one for each spine subtype. Post hoc multiple comparisons were performed using Tukey test only when ANOVA revealed differences. Significant differences are highlighted in bold text.

	Comparison	Animals (n)	F (DFn, DFd)	ANOVA P	Tukey’s Adjusted P
CA1 Total Spines	Overall	20	2.998 (3, 16)	0.0616
CA1 Stubby Spines	Overall	20	1.489 (3,16)	0.2554
CA1 Thin Spines	Overall	20	1.488 (3, 16)	0.2556
CA1 Mushroom Spines	Overall	20	4.941 (3, 16)	0.0129
	GFP vs SGK1				0.9854
	GFP vs S78A				>0.9999
	GFP vs K127Q				0.0266
	SGK1 vs S78A				0.9.813
	SGK1 vs K127Q				0.0255
	S78A vs K127Q				0.0382
DG Total Spines	Overall	18	0.368 (3, 14)	0.7775
DG Stubby Spines	Overall	18	1.102 (3, 14)	0.3811
DG Thin Spines	Overall	18	1.152 (3, 14)	0.3629
DG Mushroom Spines	Overall	18	1.315 (3, 14)	0.3087

Results:

Previous studies have shown that alterations in SGK1 levels affect the density of dendritic spines in layers II/III of the prefrontal cortex, but it is unknown how SGK1 activity affects dendritic spine morphology in the hippocampus. Therefore, we used HSV-mediated overexpression of GFP-SGK1, GFP-S78A, or GFP-K127Q in both CA1 and DG (Fig 1A and B) to assess the effects of SGK1 overexpression, N-terminal phosphorylation, and catalytic activity, respectively, on hippocampal spine density and morphology. The novel GFP-S78A and GFP-K127Q clearly expressed SGK1 specifically in transduced neurons, as indicated by overlap of SGK1 staining with GFP staining in dorsal hippocampus (Fig 1C).

Catalytically Inactive SGK1 in CA1 drives increased spine density

Results are presented numerically in Table 3. We found that expression of K127Q catalytically inactive SGK1 significantly increased the total number of dendritic spines in CA1 hippocampal neurons compared to all other groups (Fig 2B). Furthermore, this effect was driven primarily by an increase in mushroom-shaped spines in the K127Q group as compared to all other groups (Fig. 2B). The K127Q group also had an increase in thin spines as compared to the S78A group (Fig 2B), and greater expression of stubby spines compared to the GFP control and the SGK1 overexpression groups (Fig. 2B). As mushroom spines are considered mature and indicate a strengthened response to glutamatergic input, these data suggest that that in CA1, K127Q-mediated inhibition of endogenous SGK1 activity increases spine maturation.

Table 3: Results of spine analyses.

Averages and standard error from the mean are presented for each group.

CA1 Total Spines	Average per μm	SEM
GFP	1.624	0.034
WT SGK1	1.579	0.040
S78A SGK1	1.618	0.048
K127Q SGK1	1.939	0.042
CA1 Stubby Spines	Average per μm	SEM
GFP	0.387	0.009
WT SGK1	0.389	0.011
S78A SGK1	0.397	0.012
K127Q SGK1	0.433	0.001
CA1 Thin Spines	Average per μm	SEM
GFP	0.641	0.023
WT SGK1	0.631	0.025
S78A SGK1	0.608	0.025
K127Q SGK1	0.717	0.026
CA1 Mushroom Spines	Average per μm	SEM
GFP	0.599	0.015
WT SGK1	0.559	0.017
S78A SGK1	0.613	0.021
K127Q SGK1	0.789	0.020
DG Total Spines	Average per μm	SEM
GFP	1.453	0.035
WT SGK1	1.450	0.047
S78A SGK1	1.506	0.099
K127Q SGK1	1.543	0.084
DG Stubby Spines	Average per μm	SEM
GFP	0.404	0.010
WT SGK1	0.362	0.012
S78A SGK1	0.356	0.025
K127Q SGK1	0.400	0.026
DG Thin Spines	Ave age per μm	SEM
GFP	0.616	0.022
WT SGK1	0.545	0.025
S78A SGK1	0.716	0.076
K127Q SGK1	0.669	0.060
DG Mushroom Spines	Average per μm	SEM
GFP	0.442	0.016
WT SGK1	0.543	0.023
S78A SGK1	0.434	0.031
K127Q SGK1	0.474	0.032

Overexpression of SGK1 in DG drives increased mushroom spine density

Results are presented numerically in Table 3. Overexpression of wild-type SGK1 in DG hippocampal neurons significantly increased the number of mushroom spines compared to GFP controls (Fig. 3B), with no significant effect on total spines or stubby spines. Additionally, expression of S78A in DG neurons significantly increased the number of thin spines as compared to wt-SGK1 overexpression (Fig. 3B). This suggests that in the DG, increasing SGK1 activity promotes the maturation of dendritic spines, in contrast to the effects observed in CA1.

Discussion:

The current study suggests that endogenous SGK1 may prevent spine maturation in CA1 pyramidal cells while promoting spine maturation in DG granule cells, and there are multiple pathways that may mediate these effects. Activation of the MAPK/ERK pathway is important for both the initiation of translation that promotes the induction of LTP in DG[4, 23] and SGK1 expression [35]. In the DG, LTP induction requires increased ERK phosphorylation, so it is possible that upregulation of SGK1 can promote these same LTP drivers in the DG to drive the maturation of spines [4, 43]. Of particular note, activated (phosphorylated) ERK phosphorylates CREB, which increases protein synthesis to mediate morphological changes associated with LTP[9]. In addition, SGK1 phosphorylates CREB at S133, a site that promotes CREB-dependent transcription [10]. Thus, it is possible that modulation of SGK1 results in differing degrees of CREB regulation in the CA1 and DG, resulting in differential morphological alterations of dendritic spines.

SGK1 phosphorylation of guanosine nucleotide dissociation inhibitor (GDI) promotes the binding of GDI to Rab4, a GTPase which controls redistribution of AMPA receptors from early endosomes into the plasma membrane. Thus, SGK1-induced GDI-Rab4 complex formation promotes the recycling of AMPA receptors into the synaptic membrane in pyramidal neurons in the PFC [31]. Rab4 activity is critical for maintaining spine size, and downregulating Rab4 reduces the spine to dendrite ratio in hippocampal slices [5]. Thus, it is possible that in the DG, overexpression of SGK1 facilitates increased AMPA receptor recycling, promoting maturation of the spines.

Alterations in SGK1 expression are relevant to various neurological diseases. The expression of SGK1 in the PFC of posttraumatic stress disorder patients is reduced by over 80 percent [30], while expression of peripheral SGK1 is increased depressed patients [1]. Moreover, treatment with glucocorticoids increases AMPA and NMDA trafficking and function in PFC mediated directly by SGK1 [44]. In addition, SGK1 upregulates CreaT in Xenopus oocytes; in people, defective CreaT can lead to mental retardation [18, 40]. SGK1 also phosphorylates tau, and hyperphosphorylated tau is associated with Alzheimer’s Disease [8, 14].

The current findings are in line with previous studies demonstrating that environmental and behavioral stimuli can have differential effects on dendritic spines in different hippocampal subregions. For example, chronic stress affects spine shape in the CA1 but not the CA3 [34]. A few hours of intense stress decreases spine density in the CA3, while 30 minutes of acute stress increases the density of spines on neurons in the CA1[16]. Further, chronic mild stress in rats decreases the number of total spines in CA3 neurons, while having no effect on the spines of granule cells in the DG [3]. In the same animals, mushroom spines decreased in DG of rats exposed to chronic mild stress, but no morphological changes were observed in the pyramidal neurons of the CA3 [3]. Additionally, mild but chronic stress decreases BDNF and CREB in the DG but not in CA1 [17], indicating that signaling pathways potentially regulating SGK1 activity and phosphorylation are differentially regulated by the same stimuli in DG vs CA1.

Here, we demonstrate that SGK1 modulates the morphology of dendritic spines in a subregion-dependent manner in the mouse hippocampus. It will be critical to determine whether the function and plasticity of these synapses is similarly altered by SGK1 activity using slice electrophysiology. Due to the intensive nature of spine measurements and the inclusion of three separate viral manipulations of SGK1, we were forced to limit this initial study to male animals and only two subregions of the hippocampus (DG and CA1). Future studies will include determine the effects of SGK1 manipulation on spine morphology in female mice and in other brain regions, including hippocampal CA3, pyramidal neurons of the prefrontal cortex, and medium spiny neurons of the nucleus accumbens. Although studies exist demonstrating a role for hippocampal SGK1 in learning and memory [25, 28, 41, 42], the tools provided here will allow future studies to uncover a potential role for SGK1 serine 78 phosphorylation in the hippocampus in learning. Overall, the current study suggests that SGK1 may be a critical player in diseases characterized by altered hippocampal function, from Alzheimer’s disease to mood disorders, and uncovering the mechanisms of its function in hippocampus may potentially provide novel therapeutic inroads in the treatment of these diseases.

Highlights.

• Novel viral tools for the investigation of SGK1 catalytic activity and phosphorylation

• A novel role for SGK1 in hippocampal synaptic structure

• The function of SGK1 in the hippocampus is specific to the type of synapse