The Role of Interleukin-10 in Mediating the Effect of Immune Challenge on Mouse Gonadotropin-Releasing Hormone Neurons In Vivo

Abstract Immune challenge alters neural functioning via cytokine production. Inflammation has profound impact on the central regulation of fertility, but the mechanisms involved are not clearly defined. The anti-inflammatory cytokine interleukin (IL)-10 is responsible for balancing the immune response in the brain. To examine whether IL-10 has an effect on the function of the gonadotropin-releasing hormone (GnRH) neurons, we first examined the effect of immune responses with distinct cytokine profiles, such as the T cell-dependent (TD) and T cell-independent (TI) B-cell response. We investigated the effect of the TD and TI immune responses on ERK1/2 phosphorylation in GnRH neurons by administering fluorescein isothiocyanate/keyhole limpet hemocyanin (KLH-FITC) or dextran-FITC to female mice. Although dextran-FITC had no effect, KLH-FITC induced ERK1/2 phosphorylation in GnRH neurons after 6 d. KLH-FITC treatment increased the levels of IL-10 in the hypothalamus (HYP), but this treatment did not cause lymphocyte infiltration or an increase in the levels of proinflammatory cytokines. In IL-10 knock-out (KO) mice, KLH-FITC-induced ERK1/2 phosphorylation in the GnRH neurons was absent. We also showed that in IL-10 KO mice, the estrous cycle was disrupted. Perforated patch-clamp recordings from GnRH-GFP neurons, IL-10 immunohistochemistry, and in vitro experiments on acute brain slices revealed that IL-10 can directly alter GnRH neuron firing and induce ERK1/2 phosphorylation. These observations demonstrate that IL-10 plays a role in influencing signaling of GnRH neurons in the TD immune response. These results also provide the first evidence that IL-10 can directly alter the function of GnRH neurons and may help the maintenance of the integrity of the estrous cycle.


Introduction
Immune challenge is often associated with perturbances in brain function that disrupt homeostatic control mechanisms such as the neuroendocrine balance by modulating neuronal activity in the brain (ThyagaRajan and Priyanka, 2012). Interleukin (IL)-10 is an essential anti-inflammatory cytokine that regulates peripheral immune cells during excessive inflammation and it also maintains the balance of the immune response in the brain (Lobo-Silva et al., 2016;Norden et al., 2016). IL-10 is released from several cell types, including astrocytes and microglia, after immune challenge and brain injury (Schroeter and Jander, 2005;Rodney et al., 2018). Although there is growing evidence that the role of IL-10 in neurons is to induce anti-apoptotic effects during neurodegenerative diseases and brain injuries (Xin et al., 2011;Lobo-Silva et al., 2016;Lan et al., 2017), little if any attention has been given to the potential regulatory role of IL-10 in neuronal functions that regulate fertility.
Basal forebrain neurons that synthesize and secrete gonadotropin-releasing hormone (GnRH) are the final output cells of a complex neuronal network that regulates fertility in all mammalian species (Herbison, 2016). In the hypothalamic-pituitary-gonadal (HPG) axis, periodic release of GnRH into the hypophyseal portal system via the median eminence induces a pulsatile pattern of pituitary gonadotrophs regulating the gonadal function during the estrous cycle. Although clinical data show that IL-10 plays a key role in regulating fertility (Wegmann et al., 1993;van Dunné et al., 2006;Cheng and Sharma, 2015), and it has been demonstrated that various cytokines have the capacity to act on GnRH neurons (Bouret et al., 2004;Igaz et al., 2006;Kuwahara-Otani et al., 2017;Sarchielli et al., 2017), the effect of IL-10 on GnRH neurons remains unknown.
As secreted cytokines act via different intracellular signaling pathways, such as ERK1/2, to fine-tune neuronal functions (Kim et al., 2013;Sticozzi et al., 2013), we have used ERK1/2 phosphorylation as an index of changes in intracellular alterations in GnRH neurons. We applied multiple in vivo and in vitro techniques to explore the effect of IL-10 on GnRH neurons. Our findings demonstrate that IL-10 alters the intracellular signaling of GnRH in the immune response and maintains the integrity of the estrous cycle. These studies provide the first evidence for IL-10 as a regulator of the functions of GnRH neurons.

Animal treatments
Immunization of animals was performed as described previously (Adori et al., 2010). Briefly, for immunization, intact WT (C57BL/6) or IL-10 KO (129(B6)-Il10 tm1Cgn /J) female mice and their WT littermates (129(B6)/J) were treated with 200 g of T cell-dependent (TD) antigen [keyhole limpet hemocyanin/fluorescein isothiocyanate (KLH-FITC); KLH: Sigma, FITC: Invitrogen] or 200 g of T cell-independent (TI) antigen (dextran-FITC, Fluka Bio-Chemika). The TD antigen, KLH-FITC (in 50 l of PBS) mixed with 250 l of complete Freund's adjuvant (CFA) was administered subcutaneously into the tail base as the KLH-FITC injection is more effective when introduced subcutaneously since in the TD immune response the lymph nodes and the dendritic cells play the critical role (Murphy and Weaver, 2017). In case of the TI response, the peritoneal B cells give a robust response (Murphy and Weaver, 2017); therefore, the dextran-FITC (in PBS) was administered intraperitoneally. The control mice for the KLH-FITC and dextran-FITC treatments received CFA and PBS, respectively. Vaginal smears were taken before KLH-FITC, dextran-FITC, CFA, or PBS injection, and animals in the diestrus stage were selected for each immunization.
Immunized and control animals were deeply anesthetized by an overdose of Avertin at different time points (6 h and 3, 6, and 12 d) after treatment. Blood was taken for ELISA, and the mice were transcardially perfused with ice-cold physiologic saline for flow cytometry or 4% paraformaldehyde (PFA) for immunohistochemistry. C57BL6/J mice, IL-10 KO mice and their WT littermates were perfused with 4% PFA on day 6 after immunization. For immunohistochemistry, brains were removed from the skulls, perfused with 4% PFA, postfixed for 2 h at 4°C, and placed in 0.1 M Tris-buffered saline (TBS) solution containing 30% sucrose overnight at 4°C. Coronal sections (30-m thick) were cut on a freezing microtome, and four sets of sections were collected in TBS.
In separate experiments, 5-mg/kg indomethacin (Fluka, Sigma; diluted in 2.5% ethanol with 0.9% NaCl) was administered intraperitoneally to mice (n ϭ 5) 30 min (Elmorsi et al., 2013) before KLH-FITC immunization and then daily until the sixth day, when the mice were deeply anesthetized by an overdose of Avertin and perfused with 4% PFA. Control animals received the vehicle of indomethacin.

Evaluation of the estrous cycle
To evaluate the effect of IL-10 on the estrous cycle, vaginal smears were taken daily (10 A.M.) from IL-10 KO mice and WT littermates for a period of five weeks. Vaginal material was stained by a 0.1% aqueous solution of methylene blue and examined under a microscope to evaluate the stage of the estrous cycle.
The specificities of the primary antibodies of ERK1/2, pERK1/2, and GnRH have been reported previously in rodent species Barabás et al., 2006). The specificity of IL-10RA antibody have been validated by preadsorption of the primary antibody with its corresponding fusion protein. The omission of primary antibodies resulted in a complete absence of the corresponding immunoreactivity in these studies.

Image analysis
A Zeiss (Zeiss LSM 510, Carl Zeiss Jena GmbH) confocal laser scanning system equipped with a multi-argon laser, and helium-neon lasers was used to detect ERK1/2, pERK1/2, and IL-10RA immunoreactivity within GnRH neurons. An argon laser was used to excite Alexa Fluor 488-labeled GnRH neurons at a wavelength of 488 nm, and a helium-neon laser was used to excite Cy5-labeled ERK1/2-or pERK1/2-containing neurons at a wavelength of 649 nm and Alexa Fluor 594-labeled IL-10RA at a wavelength of 594 nm. The diameter of the pinhole aperture was 80 m, resulting in a 1.5-m optical thickness. Quantitative evaluation of the double-labeled tissue was performed by an investigator blinded to the experimental groupings. Analysis of the double-labeled tissue was conducted by manually counting the number of single-and dual-labeled GnRH neurons. GnRH neurons located in the MS and preoptic area (POA) and at the level of the anterior hypothalamic area (AH) were examined. Using the Paxinos and Franklin (2001) mouse brain atlas, three sections from each region were selected for analysis: MS (plates 23-26), POA (plates 25-28) and AH (plates 34 -36). In case of IL-10RA immunohistochemistry, confocal stacks were prepared from GnRH neurons (0.5-m steps, resolution: 1544 ϫ 1544 pixels) and analyzed using Imaris 8.1.2 image analysis software (Bitplane AG). Convoluted surfaces were created and used to identify the location and number of IL-10RA on GnRH neurons in the three-dimensional space. IL-10RA and GnRH neuron coexpressions were analyzed using threedimensional reconstruction of the confocal images.

Western blotting
Micro-dissected brain tissue was homogenized on ice in 0.1 M PBS-0.5% Tween buffer containing protease inhibitors [1 g/ml; leupeptin, pepstatin A, aprotinin, phenylmethanesulfonyl fluoride (PMSF) and vanadate (20 l of homogenization buffer/1 mg tissue)]. Homogenates were pelleted at high speed; supernatants were collected and stored at -80°C until further use. Proteins were separated by sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis and then electrically transferred to PVDF membranes. The membranes were incubated with rabbit anti-mouse pERK1/2 primary antibody (1:1000; Cell Signaling) or monoclonal rabbit ERK1/2-specific primary antibody (1:1000; Cell Signaling). The membranes were then incubated with HRP-conjugated polyclonal antirabbit antibody (1:20,000; Abcam). Bands were visualized on x-ray films using the Supersignal West Pico chemiluminescent substrate (Pierce). The membranes were washed extensively with TBS-0.5% Tween buffer between all the incubation steps. The ERK1/2 activity (relative to pERK1/2/ ERK1/2 levels) was calculated by normalizing the level of pERK1/2 to the total ERK1/2. pERK1/2/ERK1/2 data for each group were expressed as ratios of the mean values.

Isolation of lymphocytes and fluorescence-activated cell sorting (FACS)
KLH-FITC-treated and control mice were deeply anesthetized by an overdose of Avertin at 6 d after immunization, and the animals were perfused with physiologic saline. The brains of four KLH-FITC-injected and four control mice were homogenized separately in HBSS buffer (pH 7.4). Samples were then centrifuged at 1200 rpm for 10 min at 4°C. The pellets were resuspended in HBSS buffer containing 1 mg/ml collagenase (Sigma) and 0.1 mg/ml DNase I (Sigma), and the suspension was incubated for 1 h at 37°C. Then, EDTA was added to each sample to avoid cell aggregation at a final concentration of 20 mM for 5 min. Samples were filtered using a 70-m mesh and pelleted at 1200 rpm for 10 min at 4°C. To isolate lymphocytes, the Percoll separation method was used. Pellets were resuspended in 7 ml of 30% Percoll (Sigma) and it carefully superposed onto 4 ml of 70% Percoll. The samples were then centrifuged at 2000 rpm for 10 min at 4°C. The cells at the interface of the 30 and 70% Percoll solutions were collected and washed with HBSS buffer. At this stage, cells from the same experimental group were pooled. The live cells collected were counted by the standard Trypan blue method (Louis and Siegel, 2011).
To evaluate the lymphocyte infiltration following KLH-FITC immunization, isolated brain cells were washed with FACS buffer (PBS containing 0.5% FCS and 0.1% sodium azide). K9.361, a mouse anti-mouse FcgammaRII antibody, was used to block Fc receptors (5 min, 4°C). B cells were stained with FITC-conjugated anti-mouse B220 antibody. For T-cell detection, an Alexa Fluor 647-conjugated anti-mouse CD3-specific monoclonal antibody (Caltag) was used (30 min 4°C). Flow cytometric analysis was performed in FACSCalibur (Becton Dickinson). Regions for analysis were set according to the forward scatter (FSC, roughly proportional to the diameter of the cell) and side scatter (SSC, proportional to the granularity) characteristics of lymphocytes from the spleen. A total of 2 ϫ 10 5 fluorescence-labeled cells were analyzed with CellQuest Pro (Becton Dickinson).

ELISA
To validate the immunizations, 0.2 ml of blood was collected from the right atrium of each mouse before perfusion with PFA to determine the immunoglobulin levels produced against FITC. Plates were coated with BSA-FITC (overnight, 4°C) and then blocked with 1% non-fat milk powder. Serial dilutions of serum samples were incubated on the plates for 2 h at RT, which was followed by incubation with HRP-conjugated polyclonal anti-rabbit IgM and IgG antibodies (1:2000, Dako). The reaction was visualized with H 2 O 2 and 3,30,5,50-tetramethyl benzidine and stopped with 2 N H 2 SO 4 . The end product was measured in an ELISA reader at a wavelength of 450 nm.
For determination of brain cytokine levels following immunization, the mice were perfused with ice-cold physiologic saline solution, and the dissected brain tissue, containing the HYP and MS regions, was homogenized in 0.1 M PBS-0.5% Tween buffer containing protease inhibitors. Samples were centrifuged, and IL-1␤, TNF-␣, and IL-10 concentrations in the supernatant were determined by using a quantitative sandwich enzyme immunoassay kit according to the manufacturer's protocol (R&D Systems). Briefly, the ELISA plates were coated overnight at 4°C with mouse IL-1␤-, TNF-␣-, or IL-10-specific antibodies. Plates were washed the following day, blocked, and incubated with the brain lysates at appropriate dilutions for 2 h at RT. After washing, the plates were incubated with cytokine-specific biotinylated detection antibodies for 1 h at RT, followed by incubation with avidin-HRP for an additional 30 min. The reaction was initiated with the kit substrate solution and stopped with 2 N HCl, and the plates were read at 450 nm.

Single-cell electrophysiology
Brain slices were acutely prepared as described (Bhattarai et al., 2014); 40-to 54-d-old female GnRH-GFP mice in the diestrus stage were decapitated, and the brains were rapidly removed and placed in ice-cold ACSF with the following composition: 126 mM NaCl, 2.5 mM KCl, 2.4 mM CaCl 2 , 1.2 mM MgCl 2 , 11 mM D-glucose, 1.4 mM NaH 2 PO 4 , and 25 mM NaHCO 3 (pH 7.4 when bubbled with 95% O 2 and 5% CO 2 ). Brain blocks were placed on the chilled stage of a vibratome (Microm), and 200m-thick coronal slices containing the POA were cut. The slices were placed in oxygenated ACSF for at least 1 h at RT. The slices were transferred to the recording chamber and superfused with ACSF. The fluorescent GnRH neurons were identified at 10ϫ and 40ϫ objective magnifications of an upright microscope (BX51WI; Olympus) by brief fluorescent illumination and then viewed and patched under Nomarski differential interference contrast optics. Patch pipettes were pulled from glass capillary tubing (PG52151-4; WPI) on a Flaming/Brown micropipette puller (P-97; Sutter Instruments Co.). The pipette solution was filtered and contained 130 mM KCl, 5 mM NaCl, 0.4 mM CaCl 2 , 1 mM MgCl 2 , 10 mM HEPES, and 1.1 mM EGTA (pH 7.3 with KOH). Gramicidin (Sigma) was first dissolved in dimethylsulfoxide (Sigma) to a concentration of 2.5-5 mg/ml and then diluted in the pipette solution just before use to a final concentration of 2.5-5 g/ml. The gramicidin-perforated patch recordings were performed using an Axopatch 200B amplifier (Molecular Devices). The tip resistance of the electrode was 4 -6 M⍀.
In initial experiments, the access resistance was monitored, and the experiments began when the resistance stabilized at 50 -90 M⍀. This stabilization typically took 15-20 min after gigaseal formation and always corresponded to the resting membrane potential (RMP) of the cell reaching a stable level below -45 mV. In all subsequent cells, experiments were initiated when the RMP reached a stable level below -45 mV. Spontaneous rupture of the seal was evident by a sudden overshooting of the action potentials to values Ͼ0 mV. Changes in membrane potential were sampled using a Digidata 1322A interface (Molecular Devices) connected to a PC. Any GnRH-GFP neuron that exhibited a shift in RMP of Ͼ2 mV was considered to have responded. Acquisition and subsequent analysis of the acquired data were performed using Clampex9 software (Molecular Devices). Traces were plotted using Origin7 software (MicroCal Software). All recordings were made at RT. Similar to the in vitro studies on acute brain slices, 2 ng/ml recombinant mouse IL-10 or TTX or BM containing TTX, CNQX, AP-5, and picrotoxin were used to investigate the effect of IL-10 on the activity of GnRH neurons.

Statistical analysis
Data are expressed as the mean Ϯ SEM. Generally, we applied ANOVA with Tukey post hoc to test the significance. Arcsine transformation was performed on data collected as percentages before using ANOVA ( Fig. 3E;  Fig. 5D1,D2). In case the assumption of the homogeneity of variance and normality were violated for ANOVA, nonparametric Mann-Whitney U test was used to test the difference between independent groups (Figs. 2, 3C,D, 4) (STATISTICA 13.3 for Windows, TIBCO).

KLH-FITC but not dextran-FITC administration induces delayed ERK1/2 phosphorylation in GnRH neurons
Based on antigens and related cytokine production, humoral immune responses can be classified as TD B-cell activation and TI B-cell activation (Defrance et al., 2011;Zubler et al., 2001). In rodent experimental models, the TD and TI immune responses can be effectively dissected using selective immunization protocols such as KLH-FITC and dextran-FITC antigen treatment, respectively (Moreno et al., 1981;Adori et al., 2010). Accordingly, in the first set of experiments, we examined the effects of TD and TI antigendriven immune challenge on ERK1/2 phosphorylation in GnRH neurons using KLH-FITC and dextran-FITC immunization, respectively.
In contrast, immunization with a TI antigen, namely, dextran-FITC, failed to have any effect on ERK1/2 phosphorylation in GnRH neurons ( Fig. 1C1; n ϭ 6/group), and the number of GnRH neurons expressing ERK1/2 was not affected by the treatment (Fig. 1C2 ; n ϭ 6/group). The number of GnRH neurons did not change after dextran-FITC treatment (data not shown).
The efficiency of the immunization protocols was tested by measuring whether KLH-FITC or dextran-FITC administration effectively increased the FITC-specific antibody levels in the blood. Six days after KLH-FITC or dextran-FITC injection (for the TD and TI antigens, respectively), the levels of the FITC-specific antibodies of the IgM and IgG isotypes were elevated in the sera of immunized animals compared to the levels in the sera of control animals that received vehicles only (Fig. 1D1,D2,E1,E2 ).

IL-10 mediates KLH-FITC-induced ERK1/2 phosphorylation in GnRH neurons
Next, to understand how ERK1/2 phosphorylation is mediated in GnRH neurons on peripheral immune challenge, potential contributing factors were examined.
Our whole-brain flow cytometry measurements revealed that there was no lymphocyte infiltration into the brain following immunization with the KLH-FITC antigen (n ϭ 4 pooled samples/group, two separate experiments; Fig. 3A). Administration of indomethacin did not alter the number of pERK1/2-positive GnRH neurons following KLH-FITC treatment on day 6 (n ϭ 5/group; Fig. 3B), indicating that prostaglandins do not play a role in mediating ERK1/2 phosphorylation in GnRH neurons. The levels of the proinflammatory cytokines IL-1␤ and TNF-␣ did not change in the brain following KLH-FITC injection (IL-1␤: control Յ 3 pg/100 mg of tissue, KLH-FITC Յ 3 pg/100 mg of tissue; TNF-␣: control ϭ 13.88 Ϯ 4.94 pg/100 mg of tissue, KLH-FITC ϭ 16.14 Ϯ 2.27 pg/100 mg of tissue; n ϭ 5/group).
IL-10 levels were also quantified in the HYPs and spleens of mice following KLH-FITC injection. Interestingly, IL-10 levels were elevated in the HYPs of KLH-FITCimmunized mice compared to the levels in the controls 6 d after treatment (n ϭ 4/group, p ϭ 0.03, Mann-Whitney; Fig. 3C). In contrast, the spleens from the same immunechallenged mice did not show differences in the concentrations of IL-10 (n ϭ 4/group; Fig. 3D).
To examine whether IL-10 mediates the effect of KLH-FITC on ERK1/2 phosphorylation in GnRH neurons, we examined the expression of pERK1/2 in GnRH neurons 6 d after KLH-FITC immunization in IL-10 KO animals and their WT littermates. Similar to our findings above, the percentage of pERK1/2-positive GnRH neurons increased in WT animals following immunization (n ϭ 5/group, WT: control versus KLH-FITC: p ϭ 0.0018, F ϭ 11.9, df ϭ 1, two-way ANOVA; Fig. 3E). However, the KLH-FITCinduced ERK1/2 phosphorylation was blocked in IL-10 KO mice (n ϭ 5/group, two-way ANOVA; Fig. 3E). Expression levels of non-phosphorylated ERK1/2 were not altered by KLH-FITC injection in WT or IL-10 KO mice (Fig.  3F). The number of GnRH neurons was not affected in IL-10 KO animals (data not shown).

Estrous cyclicity is disrupted in IL-10 KO mice
GnRH neurons play a critical role in the regulation of the estrous cycle. To examine whether IL-10 has an effect on the estrous cycle under physiologic conditions, estrous cyclicity was evaluated using vaginal smears for a period of five weeks in IL-10 KO mice and WT littermates continued (‫ء‬p Ͻ 0.05, n ϭ 6/group). Antigen-specific serum IgM and IgG levels on day 6 of immunization with KLH-FITC (D1, D2) or dextran-FITC (E1, E2; ‫ء‬p Ͻ 0.05 n ϭ 6/group).

Figure 2.
Effect of KLH-FITC on ERK1/2 phosphorylation in different brain regions. Two representative immunoblots probed for pERK1/2 and ERK1/2 from the STR and the HYP of immunized and vehicle-treated mice (A). Effect of KLH-FITC on the levels of pERK1/2 in different brain areas shown by Western blot analysis (B): quantification of the proportions of pERK1/2 and ERK1/2 in the C, HC, S, T, MS, P, and HYP of control and KLH-FITC-treated animals (‫ء‬p Ͻ 0.05, n ϭ 4/group). (n ϭ 5/group; Fig. 4A). These data showed that the number of cycles decreased (p ϭ 0.016, Mann-Whitney) and the duration of each cycle increased (p ϭ 0.012, Mann-Whitney) compared to those in the WT littermates (Fig.  4B,C). The mutant mice spent significantly less time in the estrus stage (p ϭ 0.012, Mann-Whitney) and more time in the diestrus stage (p ϭ 0.012, Mann-Whitney) than the WT littermates (Fig. 4D,E). Representative estrous cycle profiles demonstrating these alterations are presented in Figure 4A.

IL-10 directly alters the electrical activity of GnRH neurons and increases their ERK1/2 phosphorylation
To evaluate the functional relevance of the effect of IL-10 on GnRH neurons, the effect of IL-10 on the electrical activity of GnRH neurons was investigated in GnRH-GFP mice, in which GnRH neurons can be readily identified.
To examine whether IL-10 directly acts on GnRH neurons, we investigated the effect of IL-10 on GnRH neurons in the presence of TTX, a Na ϩ -channel blocker. We found that 0.5 M TTX blocked the action potentials but did not inhibit the IL-10-induced membrane depolarization in 11 of the 28 (39%) GnRH neurons tested (RMP, -71.08 Ϯ 1.52 mV; Fig. 5B2 ), and membrane hyperpolarization was observed in nine of the 28 GnRH neurons tested from 18 animals (Fig. 5B1 ). Furthermore, the IL-10-induced de- In addition, to confirm that the IL-10-induced ERK1/2 phosphorylation is the result of direct action of IL-10 on GnRH neurons, we investigated ERK1/2 phosphorylation in GnRH neurons on coapplication of IL-10 and the previously used BM to acute brain slices. The number of GnRH neurons (MS: 6 Ϯ 1; POA: 20 Ϯ 1; AH: 4 Ϯ 1 GnRH neurons/section; data not shown) and the percentage of GnRH neurons containing ERK1/2 immunoreactivity in the slices were similar to those seen in the in vivo experiments under basal conditions. Treatment of brain slices with 2 ng/ml IL-10 significantly elevated pERK1/2 expression within GnRH neurons (n ϭ 6/group, ACSF: vehicle versus IL-10 treatment: p ϭ 0.00017, df ϭ 1, F ϭ 376; Fig. 5D1 ). IL-10-induced ERK1/2 phosphorylation in GnRH neurons was not affected by the blocking cocktail ( Fig. 5D2 ; blocking solution: vehicle versus IL-10 treatment: p ϭ 0.00017, df ϭ 1, F ϭ 376). The number of GnRH neurons expressing ERK1/2 was not changed by IL-10 treatment (data not shown).
To further evaluate whether GnRH neurons are direct targets of IL-10, we examined IL-10RA immunoreactivity in GnRH neurons (Fig. 5E). Confocal laser scanning microscopic examination revealed IL-10RA immunoreactiv- ity on the perikarya and dendrite of GnRH neurons. Using three-dimensional reconstruction our results showed that 26% of all GnRH neurons exhibited IL-10RA immunoreactivity (65 GnRH neurons were tested from two female mice).

Discussion
We report here that peripheral administration of the TD antigen KLH-FITC induces ERK1/2 phosphorylation in the HYP and in GnRH neurons. This effect showed region specificity as it was absent in C, HC, STR, T, MS, and P. Furthermore, the TI immune response induced by dextran-FITC did not lead to ERK1/2 phosphorylation in either the HYP or in GnRH neurons. While the levels of the proinflammatory cytokines IL-1␤ and TNF␣ did not change in the HYPs of immunized mice, the IL-10 levels in the HYPs increased 6 d after KLH-FITC treatment. The role of IL-10 was further confirmed by the fact that the KLH-FITC administration failed to induce ERK1/2 phosphorylation in IL-10 KO mice. Interestingly, the estrous cycle was perturbed in IL-10 KO mice, indicating that IL-10 deficiency could alter the estrous cycle even under pathogen-free, physiologic conditions. The functional relevance of IL-10 was demonstrated by single-cell electrophysiological experiments and in vitro experiments on acute brain slices, which showed that IL-10 had a direct effect on electrical activity and ERK1/2 phosphorylation in GnRH neurons. The direct effect of IL-10 on GnRH neurons is further supported by our finding that GnRH neurons express IL-10RA.
Peripheral infections activating innate soluble factors, cells and humoral or cell-mediated adaptive immunity result in reproductive system disease, which, in severe cases, might even lead to infertility (Sheldon et al., 2017;Liu et al., 2018). Several studies have investigated the effect of endotoxin-lipopolysaccharide (LPS) challenge on the reproductive system. Although immune challenge induced by LPS has various effects on the reproductive axis (He et al., 2003;Cardoso et al., 2010;Herman et al., 2013;Izvolskaia et al., 2016), this challenge is a model for a severe, complex, whole-body infection such as septic shock; this model does not mimic the fine-tuned, less severe immune challenges induced by bacteria and viruses. Antibody responses to bacterial or viral antigens require recognition by helper T cells and cooperation between the antigen-specific B cells and T lymphocytes. These antigens are called TD antigens. Many nonprotein antigens, such as polysaccharides and bacterial lipids, stimulate antibody production in the absence of helper T cells, the antigens of which are called TI antigens. Accordingly, the viral/bacterial antigen-induced responses are classified as TD or TI humoral immune responses. Although TD and TI responses are both mediated by cytokines, TI responses do not induce high levels of cytokine production. Based on our immunization protocol (Adori et al., 2010), we used FITC as an antigen conjugated to KLH or dextran to selectively elicit either a TD or TI immune response, respectively. This immunization protocol enabled us to selectively examine the effect of TD and TI antigens with distinct cytokine profiles on the HPG axis.
While the number of pERK1/2immunopositive GnRH neurons significantly increased following KLH-FITC injection, the number of ERK1/2-immunoreactive GnRH neurons was not altered, indicating that the increase is due to phosphorylation of the expressed ERK signal molecules and not due to the alteration of ERK1/2 expression. This increase was absent in dextran-FITC-injected animals, demonstrating that GnRH neurons respond differently to antigens activating distinct immune responses that require cytokine production. Indeed, the levels of IL-10 were significantly elevated in the HYP after KLH-FITC administration, and KLH-FITC-induced ERK1/2 phosphorylation in GnRH neurons was blocked in IL-10 KO mice, indicating the role of IL-10 in mediating GnRH neuron signaling after TD humoral immune responses.

The direct effect of IL-10 on GnRH neurons
Although it is not clear why the electrical response of GnRH neurons to IL-10 is so heterogeneous, the maintenance of IL-10-induced hyperpolarization/depolarization and pERK1/2 in the presence of BM and TTX clearly demonstrates a direct effect of IL-10 on the postsynaptic neuronal membranes of GnRH neurons. This result is supported by the fact that the GT1-7 mouse hypothalamic GnRH neuronal cell line is regulated by IL-10 (Tu et al., 2007). Previous single-cell microarray studies have suggested that mouse GnRH neurons express IL-10 ␣ and ␤ receptor mRNAs (Liposits, personal communication), providing an effective platform for a direct effect of IL-10 on GnRH neurons. Our IL-10RA-GnRH double immunostaining also confirm the possibility that IL-10 can act directly on GnRH neurons. However, the mechanism of the action of IL-10 on GnRH neurons remains unclear. In single-cell brain slice recordings of dentate gyrus neurons, IL-10 dose-dependently decreases the miniature inhibitory postsynaptic current and the picrotoxin-sensitive tonic current, indicating the existence of pre-and postsynaptic mechanisms (Suryanarayanan et al., 2016). Further studies are required to establish the mechanism of the effect on GnRH neurons. Since IL-10-induced electrical changes were maintained during the blockade of GABAergic and glutamatergic neurotransmission, it seems unlikely that IL-10 was acting via presynaptic modulation of amino acid inputs to GnRH neurons.
The source of the KLH-FITC-induced IL-10 elevation in the HYP is unclear. It is unlikely that IL-10 is derived from the periphery because our IL-10 measurements did not reveal IL-10 elevation in the periphery. It is possible that brain-infiltrating immune cells can produce cytokines such as IL-10 when they are in close proximity of the microglia (Chabot et al., 1999). Our results, however, did not show any difference in the number of brain-infiltrating leukocytes between the control and immunized mice 6 d after KLH-FITC immunization, when the ERK1/2 phosphorylation was most prominent. Changes in the ratio of T and B cell subpopulations, and a consequent functional change, however, cannot be excluded as a possible explanation. An increased proportion of regulatory B and/or T cells might enhance immunosuppression, since regulatory B cells themselves can produce IL-10 (Seifert et al., 2018), while regulatory T cells can shift the microglial polarization toward the M2 phenotype, which is characterized by the production of anti-inflammatory cytokines, including IL-10 production (Zhou et al., 2017). Indeed, activated microglia can secrete IL-10 (Park et al., 2007;Lim et al., 2013;Lobo-Silva et al., 2016). Furthermore, Fujioka et al. (2013) revealed evidence for a close relationship between microglia and GnRH neurons, providing a functional platform for microglia-derived IL-10 activity on GnRH neurons. Based on these findings, we have speculated that IL-10 produced by microglia might alter the functions of GnRH neurons. This theory warrants further investigation.

The possible functional consequences of IL-10mediated TD antigen-induced immune challenges in GnRH neurons
One of the most important questions associated with the effect of IL-10 on GnRH and IL-10-mediated TD antigen-induced ERK1/2 phosphorylation is that of the physiologic and clinical relevance.
Understanding the effect of the antibody-mediated immune response on the HPG axis has substantial clinical relevance. Several pathogens that primarily elicit humoral immune responses have an effect on the reproductive system. There is growing evidence to support the role of B lymphocyte-mediated immunity in several intracellular infections, including Chlamydia trachomatis (Li and Mc-Sorley, 2015). Chlamydia infections can cause severe fertility problems, such as pelvic inflammatory disease and involuntary infertility; therefore, Chlamydia infection is a serious threat to the health of women worldwide (Li and McSorley, 2015).
Clinical studies have demonstrated that as one of the major immunoregulatory cytokines, IL-10 is beneficial for a successful pregnancy and may increase fertility and fecundity (Wegmann et al., 1993;van Dunné et al., 2006;Cheng and Sharma, 2015). Decreased production of IL-10 is associated with pregnancy loss and increased preeclampsia (Hennessy et al., 1999). Although clinical studies have demonstrated the importance of IL-10 in the periphery, our results suggest for the first time that IL-10 directly alters the function of the central processing unit of fertility and fecundity, i.e., the GnRH neurons. Importantly, the estrous cycle was perturbed in IL-10 KO mice, suggesting an important role for IL-10 in the regulation of the HPG axis. Although the effect of IL-10 on the periphery cannot be ruled out completely, these data suggest that the action of IL-10 on GnRH can be important for regulation of the estrous cycle.
In summary, we report here that the TD humoral immune response results in delayed phosphorylation of ERK1/2 in GnRH neurons due to the effect of IL-10. These studies provide the first evidence of the direct action of IL-10 on GnRH neurons and its potential role in the maintenance of the integrity of the estrous cycle.