Sex and the Estrous-Cycle Phase Influence the Expression of G Protein-Coupled Estrogen Receptor 1 (GPER) in Schizophrenia: Translational Evidence for a New Target

Schizophrenia is a mental disorder with sex bias in disease onset and symptom severity. Recently, it was observed that females present more severe symptoms in the perimenstrual phase of the menstrual cycle. The administration of estrogen also alleviates schizophrenia symptoms. Despite this, little is known about symptom fluctuation over the menstrual cycle and the underlying mechanisms. To address this issue, we worked with the two-hit schizophrenia animal model induced by neonatal exposure to a virus-like particle, Poly I:C, associated with peripubertal unpredictable stress exposure. Prepulse inhibition of the startle reflex (PPI) in male and female mice was considered analogous to human schizophrenia-like behavior. Female mice were studied in the proestrus (high-estrogen estrous cycle phase) and diestrus (low-estrogen phase). Additionally, we evaluated the hippocampal mRNA expression of estrogen synthesis proteins; TSPO and aromatase; and estrogen receptors ERα, ERβ, and GPER. We also collected peripheral blood mononuclear cells (PBMCs) from male and female patients with schizophrenia and converted them to induced microglia-like cells (iMGs) to evaluate the expression of GPER. We observed raised hippocampal expression of GPER in two-hit female mice at the proestrus phase without PPI deficits and higher levels of proteins related to estrogen synthesis, TSPO, and aromatase. In contrast, two-hit adult males with PPI deficits presented lower hippocampal mRNA expression of TSPO, aromatase, and GPER. iMGs from male and female patients with schizophrenia showed lower mRNA expression of GPER than controls. Therefore, our results suggest that GPER alterations constitute an underlying mechanism for sex influence in schizophrenia.


Introduction
Men with schizophrenia, in contrast to women, present an earlier age of onset, have a higher propensity to show negative symptoms, and generally lower social functioning [1]. On the other hand, women, besides demonstrating a significantly less severe course of illness before menopause [1,2], also show a second peak of incidence after age 45, which may be causally associated with the advent of menopause [3,4].
Clinical observations such as these have led researchers to suggest that ovarian hormones, especially 17β-estradiol (E2), play a neuroprotective role in women, the so-called estrogen hypothesis [5,6]. Reinforcing this hypothesis, a recent systematic review and meta-analysis showed a strong correlation between hospital admission for psychotic disorders and the perimenstrual phase of the menstrual cycle [7]. However, the exact mechanisms underlying this correlation remain undetermined.
Estrogen counteracts prepulse inhibition (PPI) deficits, a neurophysiological endophenotype of schizophrenia [8], both in patients [9] and in animal models [10,11]. The ventral hippocampus regulates PPI [12]. Therefore, we theorize that knowing the hippocampal expression of receptors, steroidogenic regulatory proteins, and enzymes is a first step toward understanding E2 pathway regulation across the estrous cycle and its relationship with psychotic symptoms and behaviors. In addition, this knowledge may guide the discovery of new targets for schizophrenia treatment.
E2 is synthesized by the cytochrome P450 enzyme aromatase, with androgens as substrates. Hence, it is suggested that a reduction in aromatase activity/expression in mice signifies a loss of neuroprotection due to a reduction in estrogen levels [13]. Furthermore, microglial cells express the 18 kDa translocator specific protein (TSPO), a five transmembrane domain protein that is a rate-limiting step for neurosteroid synthesis. TSPO translocates cholesterol from the outer to the inner mitochondrial membrane [14], also presenting a wide range of vital cellular functions, including regulation of cell proliferation, programmed cell death, and heme synthesis [15]. Recently, alterations in TSPO levels have been found in several psychiatric disorders, including schizophrenia [16,17].
E2 receptors, ERα and ERβ, can initiate genomic and non-genomic E2 actions [18]. In addition, the novel estrogen receptor, G protein-coupled estrogen receptor 1 (GPER), appears to have neuroprotective properties [19] and facilitate social cognition, learning, and memory [20]. The densities of E2 receptors differ in the two sexes and are found in multiple brain areas, including the hippocampus [21]. Moreover, the hippocampal function is sexually dimorphic, as evidenced by male/female differences in cognitive function in schizophrenia [22].
Recent evidence points to alterations in GPER serum levels in schizophrenia patients [23]. The first study addressing this issue [23] showed higher serum levels of GPER in male schizophrenia patients compared with control subjects, while females had no apparent alterations. The authors argued that, despite bringing preliminary results due to the small sample size, this receptor seems to have a high diagnostic value in this mental disorder.
Microglia are brain-resident immune cells implicated in schizophrenia pathophysiology. Indeed, schizophrenia is characterized by increased microglial activation and serum concentration of several pro-inflammatory cytokines, the socalled microglia hypothesis of schizophrenia [24,25]. Estradiol can modulate microglial behavior through its receptors, mainly ERβ and GPER, leading to anti-inflammatory/proresolving phenotype. For this reason, in the present study, we evaluated GPER mRNA expression in induced microglialike cells (iMGs) from schizophrenia patients and controls.
The literature on sex and estrous cycle influences in schizophrenia neurobiology is limited. Furthermore, no previous study has analyzed the hippocampal expression of steroidogenic regulatory proteins and enzymes (TSPO and aromatase) and estrogen receptors (ERα, ERβ, and GPER) at different phases of the estrous cycle in the context of a neurodevelopmental model of schizophrenia. Thus, we hypothesized that PPI deficits triggered by the two-hit animal model of schizophrenia are influenced by sex and the estrous cycle phase and are related to a distinctive expression pattern of steroidogenic factors and E2 receptors. Additionally, since we detected significant alterations in GPER expression in the hippocampus of adult mice subjected to the two-hit model of schizophrenia (which is related to PPI alterations and influenced by sex), we decided to evaluate the expression of this receptor in iMGs obtained from male and female schizophrenia patients. These results may add to the evidence for GPER as a target in schizophrenia, as recently proposed by our research group [26].

Animal Protocol
Thirty adult Swiss mice (male:female = 1:2) were used for breeding. Pregnant females were monitored for the parturition day, which was counted as postnatal day 0 (PN0). All animals were obtained from the Drug Research and Development Center animal facility. The animals were housed in micro isolator cages under a standard 12 h dark/ light cycle (lights on at 7:00 AM) at a controlled temperature (22 ± 1 °C) with ad libitum access to food and water. All experimental procedures followed the NIH Guide for the Care and Use of Laboratory Animals [27]. The local research ethics committee approved the experimental protocol under the number 08/15. We followed The ARRIVE guidelines for writing the paper [28].

Experimental Protocol
The study followed guidelines to improve the maternal immune activation model's rigor, reproducibility, and transparency [29]. We used neonatal mice, 32 males and 64 females. These mice were randomly assigned to the control and poly(I:C) groups. On postnatal days (PNs), 5-7 mice were challenged with intraperitoneal (IP) injection of sterile saline or poly(I:C) 2 mg/kg. Poly(I:C) or saline was administered at 0.1 ml/10 g body weight. The pups were maintained in individual cages with their respective dams until weaning on PN21. During periadolescence, we randomly selected the animals for stress (S + ) exposure or left undisturbed (S − ). The protocol of peripubertal unpredictable stress (PUS) was implemented to simulate stressful life events, as previously proposed [30].
Concerning hit exposure, the animals were divided into the following groups: (1) Saline + S − , zero-hit, or control group (mice neonatally challenged with saline and left undisturbed during periadolescence); (2) Saline + S + or one-hit-stress group (mice that received saline in neonatal life and were exposed to PUS); (3) poly(I:C) + S − or one-hit-poly(I:C) group (mice neonatally challenged with poly(I:C) and left undisturbed during periadolescence); and (4) poly(I:C) + S + or two-hit group (mice neonatally challenged with poly(I:C) and exposed to PUS). To evaluate sex and estrous cycle alterations, we used males and females in two distinct estrous phases: proestrus (high estrogen levels) and diestrus (low estrogen levels). The estrous cycle phase was determined every morning between 7:00 and 8:00 AM from PN70 to PN78 based on the cell types observed in the vaginal smear [31]. Thus, based on sex and estrous cycle, the groups were divided into males, females-in-diestrus, and females-in-proestrus.
The PPI test was conducted on PN70-78 (adulthood) immediately after verifying the estrous cycle in female mice [32].
Six animals per group were assigned for the experiments. To minimize potential confounds associated with litter effects, offspring from a given dam were subjected to a neonatal treatment with saline or Poly(I:C), weaned on PN21, and divided by sex for PUS exposure or non-exposure. Furthermore, each experimental group consisted of pups from at least six litters. No mortality was observed in offspring treated with Poly(I:C). Immediately after PPI determination, hippocampi were removed and later stored in RNA at -80 °C for qPCR assays.

Prepulse Inhibition of the Startle Reflex (PPI)
This test was used to determine deficits in sensorimotor gating, an endophenotype of schizophrenia [33]. As described previously, mice were tested in a startle chamber (Insight, São Paulo, Brazil) [34]. The test was initiated with a 5-min acclimatization to the startle chamber in the presence of 65 dB background noise. Next, the animals received nine 120 dB pulses without pre-pulse (startle amplitude) and eighteen pulses preceded by 100 ms by a pre-pulse (PP) of 70-, 75-, or 80-dB intensity. PPI was calculated as a percentage score for each acoustic pre-pulse trial type: %PPI = 100 − startle amplitude with PP × 100 startle amplitude of pulse alone. The results were expressed as mean % PPI.

Participants
Participants included individuals with schizophrenia from the schizophrenia outpatient department at the Hospital das Clinicas of the Universidade Federal de Minas Gerais (UFMG), Brazil. Entry criteria included (1) a DSM-5 diagnosis of schizophrenia and (2) no psychotic symptoms (clinically stable) and treatment for at least three months. After signed consent, six males and six females with schizophrenia and 12 healthy controls matched by sex and age participated in the study, aged between 18 and 70. In addition, healthy controls (neurotypical volunteers), matched for age and sex, were selected among students of postgraduate courses at the UFMG and ambulatory staff after screening via the Mini International Neuropsychiatric Interview (MINI) by a trained psychiatrist. This project was conducted according to the Helsinki Declaration and the Brazilian National Health Council Resolution 466/12 and approved by the UFMG Institutional Review Board (COEP-UFMG) (protocol number 90424518.3.1001.5149). Figure 1 shows an overview of the experimental design.

Isolation of Peripheral Blood Mononuclear Cells (PBMCs)
A 40 mL of whole blood from patients and controls was drawn into heparin tubes and processed up to 4 h after collection [35]. Blood was layered onto Histopaque 1077 density gradient (Sigma Aldrich, 10771) and centrifuged at room temperature, 900 g, for 25 min. The PBMC fraction was collected, and the cell suspension was diluted in PBS and centrifuged twice at 400 g for 8 min. Then, the cell Fig. 1 Overview of the experimental design of the study. A For the animal protocol, mice on postnatal days 5-7 were challenged with intraperitoneal injection of sterile saline or poly(I:C) as a first hit. The pups were maintained in individual cages with their respective dams until weaning on PN21. During periadolescence, the animals were exposed to five sessions of unpredictable stress (S +) or left undisturbed (S-). At adulthood (PNs 70-78), male, female-in-diestrus, and female-in-proestrus mice were euthanized for hippocampal collection for qPCR analysis of proteins required for estrogen synthesis and estrogen receptors. B The human protocol consisted of 6 male and 6 female patients with schizophrenia and 12 healthy controls who had their blood collected for peripheral blood mononuclear cells (PBMCs) isolation and conversion to microglia-like cells (iMGs). The iMGs were challenged with lipopolysaccharide (LPS) or PBS and evaluated by qPCR for G protein-coupled estrogen receptor 1 (GPER) expression. Abbreviations: PN, postnatal days; i.p., intraperitoneal; Poly (I:C), polyinosinic-polycytidylic acid; Sal, saline; GM-CSF, granulocyte-macrophage colony-stimulating factor; IL, interleukin; LPS, lipopolysaccharide; PPI, prepulse inhibition of the startle reflex; S, stress; iMGs, microglia-like cells; PBS, phosphate buffered saline; GPER, G protein-coupled estrogen receptor pellet was resuspended in fetal bovine serum (FBS), diluted in trypan blue solution, and 10 µL was used to count viable cells in a hemocytometric Neubauer chamber. Finally, cells were frozen in FBS with 10% DMSO in liquid nitrogen until further use.

Differentiation of Induced Microglia-Like Cells (iMGs) and Lipopolysaccharide (LPS) Stimulation
The induction of the iMGs was adapted from a previously described model (Ohgidani et al., 2015; doi:10.1038/ mp.2016.220). After thawing, PBMCs were washed with PBS, centrifuged at 400 g for 8 min, and added at a density of 1.5 × 10 6 cells/cm 2 to Geltrex-coated 12-well culture plates for 1 day in RPMI 1640 medium, supplemented with 10% FBS and 1% penicillin/streptomycin. The next day, unbound cells were removed. Then, the cell medium changed to differentiation media (RPMI 1640 media, 100 ng/ ml IL-34, 10 ng/ml GM-CSF, and 1% penicillin/streptomycin) to induce differentiation towards a microglial-like fate. After 9 days of differentiation, media was removed, cells were washed with PBS twice, and the fresh differentiation medium was replenished. After 12-14 days of differentiation, the iMGs obtained from individuals in the SCZ (n = 6 by sex) and control (n = 6 by sex) groups were stimulated, in duplicates, with LPS (100 ng/mL) or vehicle (1 × PBS) during 24 h for mRNA expression analysis by RT-qPCR.

mRNA Expression
The total RNA of the cells was isolated with the TRIzol reagent (Thermo Fisher Scientific, São Paulo, Brazil). The RNA was resuspended in a final volume of 12 μl of nuclease-free water and quantified by absorbance at 260/280 nm in a spectrophotometer (NanoDrop 2000; Thermo Fisher Scientific). Then, 500 ng of total mRNA was reverse transcribed in a reaction volume of 20 μl ( [36] using, in the case of human samples, the geometric mean of the housekeeping genes RPLP0 and IPO8, and for animal samples, the gene GAPDH. Animal data are normalized relative to the control condition, "group saline-S," except for the comparisons between animals exposed to the two-hit condition, which are normalized relative to the "male poly(I:C) + S + group." Data obtained from human cells are normalized relative to the control PBS condition.

Statistical analysis
Data analysis and graphs were performed using GraphPad Prism software, version 9.1.1 for Mac (San Diego, CA). The Kolmogorov-Smirnov test was used to evaluate the data's normal distribution. Two-way ANOVA followed by Tukey's post hoc test was used for the analysis of the mean % PPI considering "sex" (male, female-in-diestrus, and femalein-proestrus) and "hit exposure" (Saline + S, Saline + S + , poly(I:C) + S − and poly(I:C) + S + ) as factors. Gene expression was analyzed by one-way ANOVA followed by the Bonferroni post hoc test or Student T-test. The results displayed in the graphics are expressed as the mean ± standard error of the mean (SEM). In all analyses, significance was set at 95%, with alpha error at 5% (P ≤ 0.05).

PPI Levels and the Expression of TSPO, ERα, and ERβ Are Influenced by the Estrous Cycle Phase and Hit Exposure
PPI deficits are a stable vulnerability marker for schizophrenia over time whose magnitude appears related to the severity of clinical symptoms and is influenced by sex [37]. In this regard, Fig. 2 shows that male mice challenged with poly(I:C) or poly(I:C) + S + significantly decreased the mean %PPI compared with zero-hit animals (Saline + S − vs. poly(I:C) + S − mean diff. 33.75, P = 0.0056; Saline + S − vs. poly(I:C) + S + mean diff. 39.96, P = 0.0004). Regarding females-in-diestrus, two-hit mice presented significant PPI deficits compared with zero-hit and one-hit-stress groups (Saline + S − vs. Poly(I:C) + S + mean diff. 46.14, P < 0.0001; Saline + S + vs. Poly(I:C) + S + mean diff. 51.55, P < 0.0001). However, neither poly(I:C) nor stress altered PPI in female mice in the proestrus phase.
To investigate alterations in estrogen hippocampal synthesis, we searched for changes in genes that regulate the expression of the neurosteroidogenic proteins, i.e., TSPO and aromatase. In the analysis of TSPO (Fig. 3A), we observed that male mice exposed to one-hit-stress, one-hitpoly(I:C), and two-hit significantly decreased hippocampal mRNA expression of this factor compared with zero-hit mice (Saline + S − vs. Saline + S + mean diff. 0.8304, P < 0.0001; Saline + S − vs. Poly(I:C) + S − mean diff. 0.6449, P = 0.0002; Saline + S − vs. Poly(I:C) + S + mean diff. 0.3655, P = 0.0325). Females-in-diestrus (Fig. 3B) presented a marked increase in the mRNA expression of TSPO only when exposed to one-hit-stress (Saline + S − vs. Saline + S + mean diff.: -8.903, P < 0.0001). On the other hand, females-inproestrus (Fig. 3C) presented a marked increase in TSPO mRNA expression when exposed to two-hit compared with zero-hit (Saline + S − vs. Poly(I:C) + S + mean diff. -4,473, P = 0.0019). In the evaluation of aromatase mRNA expression, we observed decreased expression in male mice (Fig. 3D) exposed to one-hit-stress alone and two-hit compared with zero-hit mice (Saline + S − vs. Saline + S + mean diff. 0.8091, P = 0.0020; Saline + S − vs. Poly(I:C) + S + mean diff. 0.5467, P = 0.5467). Similarly to TSPO, aromatase mRNA expression in females-in-diestrus (Fig. 3E) presented a marked increase in the hippocampus of the one-hit-stress group (Saline + S − vs. Saline + S + mean diff. -12.15, P < 0.0001) and in females-in-proestrus (Fig. 3F) exposed to two-hit in comparison with zero-hit mice (Saline + S − vs. Poly(I:C) + S + mean diff. -6.069, P < 0.0001). By comparing the results obtained in the animals exposed to the two-hit, we observed that females-in-proestrus presented significantly higher expression of TSPO and aromatase (P < 0.0001) compared with males ( Fig. 3 G,H).

GPER1 Expression Is Higher in the Animals that Do Not Present PPI Deficits
We evaluated GPER1 mRNA expression in mice to further elucidate the hippocampal estrogen receptors' expression pattern along the estrous cycle in animals exposed to one-or two-hit. Increased GPER mRNA expression was observed in male mice (Fig. 5A) exposed to one-hit-stress, while twohit exposure caused a significant reduction when compared to zero-hit (Saline + S − vs. Saline + S + mean diff. -0.6631, P = 0.0130; Saline + S − vs. Poly(I:C) + S + mean diff. 0.7110, P = 0.0077). Females-in-diestrus (Fig. 5B) presented a marked increased expression of GPER1 when exposed to one-hitpoly(I:C), whereas the two-hit exposure significantly reduced its levels (Saline + S − vs. Poly(I:C) + S − mean diff. -0.8491, P = 0.0042; Saline + S − vs. Poly(I:C) + S + mean diff. 0.4014, P < 0.05). Females-in-proestrus (Fig. 5C) showed increased mRNA expression of GPER1 when exposed to one-hitpoly(I:C) and to two-hit in comparison with zero-hit mice (Saline + S − vs. Poly(I:C) + S − mean diff. -1.194, P = 0.0299; Saline + S − vs. Poly(I:C) + S + mean diff. -1.662, P = 0.0023). By comparing the results obtained in the animals exposed to the two-hit, we observed higher expression of GPER in females-in-proestrus (P < 0.0001) and lower expression in females-in-diestrus (P < 0.01) compared with males (Fig. D).

iMGs from Patients with Schizophrenia Present Lower GPER Expression
In a pilot study with iMGs obtained from male subjects (Fig. 6), we observed that cells from patients with schizophrenia under control conditions (exposed to PBS) had lower expression of GPER compared to control subjects (P < 0.01). Moreover, the exposure of these cells to the endotoxin LPS maintained the decrease in GPER mRNA expression observed with PBS exposure (P < 0.01). The same pattern of alterations was observed in iMGs obtained from female patients with schizophrenia, which showed decreased expression of GPER, as compared to control patients (P < 0.01) (Fig. 5E). In addition, as observed in the case of males, LPS exposure maintained the decrease in the expression of GPER observed with PBS exposure (P < 0.01).
We observed no difference in GPER mRNA expression between control male and female samples (Fig. 6C).

Discussion
Accumulated evidence points to the male sex being a risk factor for schizophrenia, with women seemingly protected-later onset and less severe symptoms [38]. On the other hand, novel evidence points to the relative vulnerability to symptom severity of the perimenstrual phase of the menstrual cycle. Altered estrogen signaling in the perimenstrual phase seems to be a putative mechanism for increased symptom intensity. Providing initial preclinical evidence on this subject, our results revealed, in a validated two-hit animal model of schizophrenia, that the expression of the non-genomic estrogen receptor, GPER [39,40], is influenced by hit exposure, sex, and the estrous cycle phase. In other words, we found that adult female mice exposed to two-hit had no PPI deficits in the proestrus phase of the cycle, overlapping with a higher expression of GPER and the proteins related to estrogen synthesis, TSPO, and aromatase. In contrast, lower or no alteration in GPER expression was observed in two-hit adult males and diestrus females, showing PPI deficits. Therefore, our results provide evidence for GPER overexpression as a protective mechanism against developing sensorimotor deficits in animals exposed to risk factors related to schizophrenia based on neonatal immune activation combined with peripubertal stress.
Sensorimotor gating deficits are evaluated in humans and animals by reduced PPI and interpreted as impaired information processing [41]. Accumulated evidence points to sex influence on PPI, with healthy men exhibiting higher PPI than women [42]. Additionally, 17betaestradiol increases PPI in rats 30 min after injection. This increase was observed in animals treated with estrogen under control conditions and exposed to drugs that disrupt PPI response, such as apomorphine and MK-801 [11].
To our knowledge, this is the first study describing sex and the estrous cycle influence on PPI in animals exposed to the two-hit model of schizophrenia, based on viral mimetic exposure in neonatal life and combined with peripubertal unpredictable stress. It deserves to be mentioned that the two-hit animal model of schizophrenia, with the first hit being a neonatal (not a prenatal) exposure to poly I:C, has previously been implemented in our laboratory, resulting in behavioral and neurobiological alterations like those observed in schizophrenia [43,44] This model resembles various aspects of schizophrenia, such as epidemiology, pathophysiology, symptomatology, treatment [45], as well as gut microbiota alterations in the prodrome phase of the disorder [46].
Indeed, we observed that adult male mice presented PPI deficit depending on the nature of the first hit, i.e., if stress or poly I:C exposure, and also after two-hit exposure. Indeed, after one-hit stress, we did not observe a PPI deficit in male mice, which was accompanied by hippocampal reduced levels of TSPO, aromatase, ERα, ERβ, and increased GPER expression. On the other hand, we observed PPI deficits in male mice neonatally exposed to poly I:C, which was accompanied by hippocampal reduced levels of TSPO, ERα, and ERβ, with no alterations in GPER and aromatase expression. In females, we observed that PPI deficits occurred only in female-in-diestrus animals exposed to the two-hit, which was accompanied by no alterations in TSPO, aromatase, ERα, ERβ, and GPER expression, while one-hit exposure to poly I:C to females-in-diestrus resulted in increased GPER without PPI deficit. In other words, by analyzing the results obtained in the present study, we can infer that hippocampal GPER overexpression may be a protective mechanism against PPI deficits in animals exposed to the two-hit model. Our findings bring mechanistic evidence for the Fig. 5 Relative mRNA expression of the G protein-coupled estrogen receptor 1 (GPER) in adult male (A), female-indiestrus (B), and female-inproestrus (C) mice exposed to neonatal poly I:C or saline as a first hit and peripubertal unpredictable stress (S +) or left undisturbed (S-) as a second hit. GPER expression in male, female-in-diestrus, and femalein-proestrus animals exposed to the two-hit (D). Columns represent means ± s.e.m. increased vulnerability to schizophrenia due to exposure to prenatal infection and psychological trauma in peripubertal life, as previously reported [47], as well as reveal that this mechanism may be related to estrogen hippocampal signaling imbalance.
A previous study of our research group with an animal model of schizophrenia based on neonatal N-methyl-D-aspartate (NMDA) receptor blockade by ketamine revealed similar sex and estrous cycle influence in PPI measures [32]. Ketamine neonatal administration caused PPI deficits and a hippocampal oxidative imbalance (decreased GSH levels, increased lipid peroxidation, and nitrite levels) in both male and female-in-diestrus rats. Although our aim with this previous study was not to search for the influence of estrogen, we can speculate that the oxidant alterations observed may also be related to alterations in estrogen signaling since females-in-proestrus were protected from these alterations. Indeed, estrogen has antioxidant properties related to the upregulation of mitochondrial antioxidant enzymatic expression via intracellular signaling pathways [48]. In line with this, GPER is present in mitochondria protecting them from oxidative imbalance [49] In another study of our research group, we observed that male and female rats exposed to the two-hit model of schizophrenia presented oxidative imbalance. Nonetheless, only females had a significant decrease in GPER hippocampal protein expression. At the same time, males showed a tendency toward decreased expression. This study was not controlled by the estrous cycle phase. The peripubertal administration of the antioxidant N-acetylcysteine effectively prevented GPER alterations [44], providing the first evidence that correcting an oxidative imbalance may normalize GPER levels. Further studies are needed to confirm this hypothesis.
It deserves to be mentioned that oxidative mechanisms may be involved in PPI deficits and pro-inflammatory ones induced by a systemic maternal immune response influencing neurodevelopment [50]. In other words, prenatal/neonatal infection can act as a "neurodevelopmental disease primer" relevant for several chronic mental illnesses by mechanisms related to increased interleukin (IL)-6 plasma . Bonferroni's multiple comparisons test for post hoc analysis: *P < 0.05 and **P < 0.01 for the comparisons between the groups indicated by the connectors levels during gestation/early life and abnormal, long-lasting increase of excitatory synapses, and brain connectivity in the offspring [51]. Importantly, GPER suppresses IL-6 production [52] The hippocampus is one of the main target regions for brain estrogen action. Fluctuations in plasma estrogen levels during the estrous cycle modify hippocampal spine density, enhancing synaptic NMDA receptor current and the magnitude of longterm potentiation, cellular correlates of learning, and memory [53], which are compromised in schizophrenia [54]. Estrogen hippocampal effects are modulated by different molecular mechanisms and signaling pathways [55]. Indeed, genomic and non-genomic receptors contribute to estrogen-mediated regulation of hippocampal synaptic plasticity [55]. In line with this, GPER is expressed in hippocampal neurons of both mice and humans by genes identified as putative homologues ("Gper1-G protein coupled estrogen receptor 1-Mus musculus (Mouse) | UniProtKB | UniProt," n.d.; "HomoloGene-NCBI," n.d.).
In the present study, we observed that two-hit male and female animals with PPI deficits showed decreased mRNA expression of proteins related to estrogen synthesis and altered expression of estrogen receptors. The hormone 17β-estradiol is effective as adjunctive treatment in schizophrenia patients, both females and males [56,57]. Hence, knowing that higher expression of GPER may protect against schizophrenia symptoms and that this receptor is not linked to sexual characteristics like ERα and ERβ, GPER agonists appear as potentially interesting alternatives for treating schizophrenia [26].
Therefore, since GPER is highly expressed in microglia [58] but not studied in cells from schizophrenia patients, we conducted a pilot protocol to evaluate GPER expression in iMGs from male and female patients with schizophrenia. We observed that GPER expression was significantly lower than controls in iMGs derived from patients of both sexes and was not influenced by the exposure of these cells to an immune challenge with the endotoxin LPS. These results are the first evidence for GPER as a possible novel target for treating schizophrenia [26].
Genetic variations in estrogen receptors are related to schizophrenia risk. In this regard, ERα seems to be involved in neurobiological reproductive systems [59]. At the same time, ERβ is important in modulating nonreproductive neurobiological systems involved in anxiety, locomotion, fear, memory, and learning and is expressed in the cortex, hippocampus, and cerebellum [60]. In addition, GPER seems responsible for the anti-inflammatory activities of estrogen and is also present in brain areas related to cognition and behavior regulation [18].
The present study has some limitations since we did not evaluate the circulating levels of sex hormones or the estrogen-response elements. Nevertheless, the circulating estrogen levels in mice may be inferred by the alterations in vaginal cells used to confirm the estrous cycle phase. Another limitation of the human study was not asking female participants about the menstrual cycle phase nor evaluating the expression of ERα and ERβ in iMGs. Additionally, the sample size of human samples is small since our goal was to provide a shred of initial evidence for this target in patients with schizophrenia.

Conclusion
In conclusion, we observed that the animals protected from PPI deficits induced by the two-hit exposure, i.e., females-indiestrus, presented overexpression of GPER and also of proteins related to estrogen synthesis. On the other hand, male animals exposed to neonatal immune activation with PPI deficits presented decreased expression of proteins related to estrogen synthesis and of ERα and ERβ with no alteration in GPER and aromatase expression. Additionally, iMGs from male and female patients with schizophrenia showed reduced expression of GPER. Altogether, these results bring the first evidence for GPER alterations in schizophrenia, opening new avenues for developing agonists of this receptor for treating this mental disorder.