Early-life stress elicits peripheral and brain immune activation differently in wild type and 5xFAD mice in a sex-specific manner

All experiments were performed in accordance with the guidelines on experimental animal research approved by the Malmö-Lund Ethical Committee for Animal Research in Sweden (Dnr. 5.8. 18-01107/2018). The experiments were conducted on age-matched B6SJL 5xFAD Tg6799 transgenic mice and their wild-type (WT) littermates using the following breeding strategy: two WT female mice were paired with one heterozygous 5xFAD male mouse (9–12 weeks old) per cage. Thus, from the same breeding cage, we obtained both WT and 5xFAD heterozygous male and female littermates. Pups were weaned at postnatal day 30 (P30). Age- and sex-matched littermates were group-housed (4–5 animals/cage) in standard cages on a 12 h light/dark cycle, and water, food and nesting material were provided ad libitum.

Mice were genotyped for WT or 5xFAD status by PCR as previously described [13]. Briefly, DNA from ear punches collected at P30 was extracted using a kit (Extract-N-Amp™, Sigma-Aldrich) and amplified for PCR using the 2× PCR Bio HS Taq Mix Red enzyme (PCR Biosystems) with the following primers (5′–3′): APP forward: AGGACTGACCACTCGACCAG; APP reverse: CGGGGGTCTAGTTCTGCAT; PSN1 forward: AATAGAGAACGGCAGGAGCA; PSN1 reverse: GCCATGAGGGCACTAATCAT; WT APP forward: CTAGGCCACAGAATTGAAAGATCT; WT APP reverse: GTAGGTGGAAATTCTAGCATCATCC.

Maternal separation (MS) was performed as recently reported [14]. Briefly, pups were separated from their dams every day from P2 to P14, 3 h per day (09.00 AM–12.00 PM), and placed in a different room to avoid vocalized communication with their dams. Extra nesting material (cotton pieces) was added to keep them warm. After 180 min, pups were returned to their home cage and left undisturbed until the following MS session. Control litters (non-MS) were handled similarly to the MS pups from P2 to P14 without separation from their dams. No differences in body weight were found at the end of the MS manipulation.

5xFAD and WT littermate mice were randomly assigned to the Non-MS and MS conditions between P2 and P14. At P15, two cohorts of animals (n = 5 animals/group) were sacrificed to analyze microglia and cytokine levels. Behavioral characterization started at 4 months old and was performed in five different cohorts (5–13 animals/group). Following three consecutive days of handling, mice were tested for anxiety-like behavior in the open field test (n = 6–13 animals/group), learning and memory using the novel object recognition memory test (5–7 animals/group) and depressive-like behavior in the forced swim test (6–12 animals/group) (see Fig. 2A). Three days after the last behavioral session, a random subset of animals from each group was perfused, and their brains were collected for subsequent molecular analysis (≥ 4 animals/group). Flow cytometry analyses were performed on the spleens from 25 animals included in the behavioral experiments and 42 animals that did not undergo behavioral testing for a minimum of 5 animals/group.

Behavioral tests were performed in specific behavioral rooms during the light phase (09.00 AM–3.00 PM) by the same researcher. First, mice were habituated to the behavioral room for at least 30 min before starting the test. All equipment was thoroughly cleaned with 70% ethanol at the beginning of the session and between trials to remove olfactory cues. The testing and analyses of videotaped sessions were performed by a blinded, experienced researcher.

At P15 and 4 months of age, mice were anesthetized using isoflurane (5%) in oxygen (Virbac) and transcardially perfused with 0.9% saline solution.

Brain sections from P15 and 4-month-old mice were labeled as previously described [14]. Briefly, free-floating coronal sections (40 µm) were permeabilized with 1% (v/v) Triton X-100 in PBS (PBS-T 1%, Sigma-Aldrich) for 1 h and then incubated in blocking solution with 5% normal donkey serum (NDS) in PBS-T 1% for Gal-3, 6E10 and Iba1 co-labeling. For Iba1 and CD68 co-labeling, sections were blocked in 3% NDS in PBS-T 0.3% for 1 h. Sections were then incubated overnight at 4ºC with primary antibodies for microglia (Iba1, Wako, 1:500), Gal-3 (R&D, 1:300), CD68 (BioRad, 1:1000) and APP/amyloid-beta (Aβ) (6E10, Covance, 1:500). The following day, the tissue was rinsed for 1 h in PBS-T 0.1%, incubated with the corresponding secondary antibody (1:500, donkey-anti-rabbit 647 for Iba1, donkey-anti-goat 488 for Gal3, donkey-anti-rat conj. Cy2 for CD68 and donkey-anti-mouse 555 for 6E10, Invitrogen) for 1 h and then mounted with Diamond Antifade Mountant (ThermoScientific, Sweden) for visualization. Images of Iba1 labeling of P15 mice and Iba1, Gal3, and 6E10 co-labeling of 4-month-old mice were taken using a Nikon A1RHD laser-scanning confocal microscope using a 20 × air objective (numerical aperture 0.5). CD68-Iba1 co-labeled sections were visualized with a Nikon Eclipse 80i upright microscope using a 20 × objective (numerical aperture 0.75). All acquisition parameters were kept constant for a given experiment and were taken by the same researcher, blinded to the genotype, sex, and ELS exposure. The fluorescently labeled structures were analyzed using Fiji Image J software (W. Rasband, National Institutes of Health). For analysis, image background was first subtracted before a brightness threshold was set for the measurement of each marker. Confocal images were analyzed from a maximum intensity projection. At P15, hippocampal microglia were morphologically classified into three morphological phenotypes: with ramified processes, with stout processes, and round/ameboid like previously reported [18]. At least 2–3 brain sections per animal and brain region were analyzed at the following Bregma coordinates: hippocampus (− 2.0, − 2.5 mm), prefrontal cortex (+ 2.22, + 1.98 mm), and amygdala (− 1.7, − 2.0 mm).

To measure plaque burden, coronal brain sections (40 µm) from 4-month-old 5xFAD mice were mounted onto SuperFrost Plus slides (Thermo Scientific) 24 h before staining with Congo red. Sections were then incubated for 3 min in an alcoholic sodium chloride solution (0.03% (w/v) NaCl in 100% ethanol) with 20% Congo red (Sigma Aldrich) before being dehydrated for visualization. Slides were scanned using a 40 × objective in brightfield mode on a digital scanner (Nanozoomer 2H). Densitometric analysis of the labeled plaques was performed using NDP2-viewer software (Hammamatsu, JP) by a researcher blinded to the sex and ELS exposure.

Spleens were collected and homogenized as described in Sect. "Spleen". Subsequently, a fraction of each sample was divided into two different tubes for labeling with specific antibody cocktails: one for identifying mainly lymphoid cells and the other for identifying mostly myeloid cells. Cells were labeled in FACS buffer for 30 min at 4 °C with the corresponding master mix containing the following antibodies: lymphoid master mix containing CD4-APC Cy7 (GK1.5), NK1.1- Biotin (PK136), CD45.2-FITC (104), CD45.1-FITC (A20), Streptavidin-BV605 (Biolegend), CD8a- PeCy5 (53-6.7), CD11b-APC (M1/70), and CD62L-BV421; and Myeloid master mix containing B220- Biotin (RA3-6B2), NK1.1- Biotin (PK136), CD4-Biotin (GK1.5), CD8a-Biotin (53-6.7), CD45.2-FITC (104), CD45.1-FITC (A20), CD11b-APC (M1/70), CD115-BV605 (AF598), Ly6G-APC/Fire750 (1A8), CD11c-BV570 (N418) (Biolegend), Ly6c-PE (AL-21) and streptavidin-PerCP-Cy5.5 (BD bioscience). After labeling and prior to FACS processing, cells were stained using propidium iodide (Life Technologies) diluted 1:1000 in FACS buffer to determine viability. All samples were processed in an LSR Fortessa analyzer (Becton Dickinson), wherein compensation controls with singly labeled anti-rat/anti-hamster or anti-mouse compensation beads (BD Bioscience) were set up prior to the acquisition. Fluorescence minus one (FMO) controls with unfractionated splenocytes were run for proper gating definition. Acquired data were analyzed using FlowJo software (TreeStar). Immunophenotypic descriptions of analyzed populations include the following: inflammatory monocytes CD45 + Lin-(CD4; CD8; Nk1.1; B220) CD11bhiLy6ChiLy6G-/loCD115 + ; T cytotoxic cells CD45 + CD8 + CD4-Nk1.1-CD11b-; dendritic cells CD45 + CD11c + ; eosinophils CD45 + Lin-(CD4; CD8; Nk1.1; B220) CD115-CD11bhiSSChi; neutrophils CD45 + Lin-(CD4; CD8; Nk1.1; B220) CD115- CD11bhiSSCloLy6G + , activated T helper cells CD45 + CD4 + CD8-Nk1.1-CD11b-CD62L- and Activated T cytotoxic cells CD45 + CD8 + CD4-Nk1.1-CD11b-CD62L-.

Total hippocampal RNA from 4-month-old mice was extracted using TRI-reagent (Sigma-Aldrich) following manufacturer’s instructions. RNA concentrations were measured using a NanoDrop (2000C, Thermo), and 1 μg of total RNA was converted to cDNA using an iScript™ cDNA synthesis kit (BioRad). Real-time RT-qPCR was performed using the following primer sequences (5′–3′): Bdnf [19] (forward: GCGGACCCATGGGACTCT; reverse: CTGCTGCTGTAGTGACCGA) and Arc [19] (forward: GCTGAGCTCTGCTCTTCTTCA; reverse GGTGAGCTGAAGCCACAAAT). Amplification was done using a CFX96™ Real-Time System-C1000™ Thermal Cycler (BioRad). Relative gene expression was normalized to mRNA levels of the housekeeping gene Gapdh (forward: ACCCAGAAGACTGTGGATGG; reverse: ACACATTGGGGGTAGGAACA).

Sequential protein extractions of prefrontal cortex (PFC) samples from 4-month-old mice were performed as previously described [20]. Briefly, to obtain the S1 fraction, tissue was homogenized with a dounce homogenizer in PBS (1 ml/ 100 µg of tissue) and ultracentrifuged for 1 h at 40,000 rpm. The supernatant (S1 fraction) was aliquoted and stored at – 80 ºC. The pellet was dissolved in RIPA buffer (Sigma-Aldrich) and centrifuged at 30,000 rpm for 1 h. The resulting supernatant (S2 fraction) was aliquoted and stored at – 80 ºC. S1 and S2 fractions contained soluble amyloid. RIPA and PBS solutions were prepared with a protein inhibitor (Protein Inhibitor Cocktail, ThermoScientific; PhosphoStop, Roche). To obtain the S3 fraction, which contained protofibrils, the pellet obtained in the S2 fraction was re-dissolved in buffered-SDS (2% SDS, 20 mM Tris–HCl, pH 7.4, 140 mM NaCl) and centrifuged at 30,000 rpm for 1 h, and the supernatant (S3 fraction) was stored at – 80 ºC. To obtain the P3 fraction, which included dense plaques, the pellet obtained in the S3 fraction was dissolved in SDS-urea (20 mM Tris–HCl, pH 7.4, 4% SDS and 8 M urea) and stored at – 80 ºC.

Total protein concentrations were estimated using the BCA assay following the manufacturer’s guidelines (BCA Protein Assay-Kit, ThermoScientific, Sweden). Proteins were separated out using SDS-PAGE with pre-cast gels (4–20%, Bio-Rad) and transferred to nitrocellulose membranes using the TransBlot Turbo system (Bio-Rad). The membranes were blocked in a blocking solution (3% skim milk in PBS) for 1 h R/T and incubated with anti-human amyloid precursor protein (APP)/amyloid-β (6E10, 1:3000, Covance) overnight at 4 ºC. The following day, the membranes were washed three times in PBS-T20 0.25% followed by incubation with the secondary antibody (peroxidase-conjugated anti-immunoglobulin, 1:10,000, DAKO) and a conjugated actin antibody for 2 h R/T. Blots were developed using the ECL Clarity kit (Bio-Rad) and visualized using a ChemiBlot XRS + (Bio-Rad). Bands were analyzed by densitometry using Fiji ImageJ software (W. Rasband, National Institutes of Health) and normalized to actin signal (arbitrary units).

Cytokine levels in hippocampal extracts of P15 mice were measured using a Meso Scale Discovery V-Plex Plus Kit (MSD Mesoscale Discovery, USA) with proinflammatory mouse panels (IFNγ, IL-1β, IL-2, IL-4, IL-5, IL-6, IL-10, IL-12p70, TNF-α and CXCL1) following manufacturer’s protocol as done previously [21]. Cytokines below the detection limit were removed from the analysis.

All statistical analyses were performed using GraphPad Prism 8.0 Software for Macintosh (GraphPad Software, San Diego, CA, USA). In graphs, individual dots represent single mice. Two-way ANOVA followed by Tukey’s multiple comparisons test was performed and the main effects due to sex, MS and the interaction effect between those factors were reported. Statistical 2-way ANOVA differences between groups are stated in Additional file 3: Tables S1–S5. Additionally, three-way ANOVA followed by Tukey’s multiple comparisons test was performed and the main effects due to sex, MS, genotype and the interaction effect between those factors are reported in Additional file 4: Tables S1.1–S5.1. Data are presented as mean ± SD. P values ≤ 0.05 were considered statistically significant and are indicated in the figure legends.

We first wondered whether maternal separation (MS) could induce any immediate neuroinflammatory alterations that we could detect. To study this, we labeled brain tissue from WT and 5xFAD mice 24 h following the last separation session at P15 (Fig. 1A) for the microglial marker Iba1, focusing on the hippocampus (Fig. 1B top and Fig. 1C) and prefrontal cortex (Fig. 1B bottom).

Our results showed significant differences in Iba1 + area between WT groups due to sex (F(1,13) = 16.4; P = 0.0014) and MS (F(1,13) = 13; P = 0.0032) in the hippocampus (Fig. 1B top) but not prefrontal cortex (Fig. 1B bottom) (see Additional file 3: Table S1). Similar to a previous study [22], post hoc analysis revealed significantly increased hippocampal Iba1 + area in MS WT male mice compared to their Non-MS counterparts (Fig. 1B) (P = 0.0113). Moreover, this increase in microglial area was significantly higher in MS WT males compared to MS WT females (P = 0.0093), suggesting a sex-specific microglial response immediately after exposure to MS in WT mice.

In P15 5xFAD mice, we did not find any significant effects on Iba1 + area in the hippocampus nor prefrontal cortex when accounting for sex and MS (see Additional file 3: Table S1). To further characterize microglia at this age, we performed a morphological analysis in both hippocampus and prefrontal cortex and classified microglia as into one of three categories: being ramified, having stout processes or being round/ameboid (Fig. 1D). Despite the differences that we found in Iba1 + area in WT hippocampus, we did not find any differences in the distribution of microglial morphologies between groups in WT and 5xFAD mice (Fig. 1D left, Additional file 3: Table S1). Interestingly, in prefrontal cortex, WT mice showed increased percentages of activated microglia in MS females compared to Non-MS counterparts (with stout processes microglia: P = 0.0055; round/ameboid microglia: P = 0.0021) and MS males (round/ameboid microglia: P = 0.0173) though no differences were found between 5xFAD mouse groups (see Additional file 3: Table S1).

Additionally, we also measured cytokine levels from hippocampal extracts of P15 WT and 5xFAD mice (Fig. 1E–H). In WT mice, a significant sex-dependent difference was observed in CXCL1 (Fig. 1F) and TNF-α (Fig. 1H) levels (CXCL1: F(1,16) = 8.93, P = 0.0087; TNF-α: F(1,16) = 7.36, P = 0.0154). In 5xFAD mice, we could only detect a significant interaction effect on TNF-α levels (F(1,16) = 4.54, P = 0.0489), meaning that the effect of MS differed depending on sex (Fig. 1H; Additional file 3: Table S1). Still, these findings indicate that exposure to MS may alter microglia and cytokine release in a sex-, brain region- and genotype-dependent manner.

We and others have reported sex-dependent behavioral alterations due to MS in adolescent [14] and adult [23] WT mice. Extending this, we wondered whether mice with an AD-like phenotype would demonstrate behavioral alterations due to MS and considered sex in our analyses. Therefore, to evaluate the effects of MS and sex on the behavioral outcomes of adult WT and 5xFAD mice, we performed the open field test (OFT), novel object recognition memory test (NOR) and forced swim test (FST) (Fig. 2A) on 4-month-old mice.

In WT mice, MS significantly affected the behavioral outcomes measured in NOR (Fig. 2D; F(1,19) = 17.0; P = 0.0006) and FST (Fig. 2C; F(1,24) = 20.8, P = 0.0001) but not OFT (Fig. 2B). Post hoc analysis revealed that both MS WT females and males presented significantly lower discrimination indices (DI) in NOR compared to their Non-MS counterparts (Fig. 2D). Additionally, MS WT male mice spent more time immobile in FST compared to Non-MS male WT mice (P = 0.003), and a trend towards increased immobility time was also observed in MS WT female mice compared to its Non-MS counterpart (Fig. 2C). In WT mice, sex did not have a statistically significant effect on the behavioral outcomes measured (see Additional file 3: Table S2).

In contrast, we observed a significant interaction effect on OFT results in 5xFAD mice (F(1,37) = 4.89, P = 0.0332; Fig. 2B) with MS having different effects between sexes. We also noted that MS had a significant impact on immobility time in FST (Fig. 2C) and that sex and MS have significant but separate effect on NOR results in 5xFAD mice (Fig. 2D). Specifically, MS 5xFAD females spent more time immobile in FST (P = 0.0269) and had a lower DI in NOR (P = 0.04) compared to Non-MS 5xFAD females.

We next analyzed hippocampal gene expression levels of brain-derived neurotrophic factor (Bdnf) and activity-regulated cytoskeleton-associated protein (Arc), which are involved in cognition and AD (reviewed in [24, 25]). MS significantly affected the expression of Bdnf (Fig. 2F) in both WT (F(1,18) = 12.1, P = 0.0027) and 5xFAD mice (F(1,19) = 6.72, P = 0.0179). Post hoc analysis showed significantly lower ΔCt values in both MS WT (P = 0.0160) and MS 5xFAD (P = 0.034) male mice compared to their Non-MS counterparts though sex did not have an overall statistically significant effect (see Additional file 3: Table S2). Moreover, no differences in the ΔCt values of Arc (Fig. 2E) were observed between WT groups (see Additional file 3: Table S2). However, we did find a significant interaction effect in 5xFAD mice on ΔCt values of Arc (Interaction: F(1,22) = 16.6, P = 0.0005; Fig. 2E), indicating a sex-dependent MS effect. Here, post hoc analysis revealed significantly lower ΔCt values in Non-MS females compared to Non-MS males (P = 0.0253) and in MS males compared to Non-MS males (P = 0.0185). Altogether, these findings suggest that sex may dictate the expression changes of certain hippocampal genes in response to MS and that males may be more affected compared to females.

Because ELS and AD occur at opposite life stages, we wondered whether microglial alterations would persist in adult WT and 5xFAD mice. We observed significant MS-associated microglial alterations in the prefrontal cortex of adult WT mice (F(1,21) = 9.61, P = 0.0054; Fig. 3A, B). Post hoc analysis revealed higher Iba1 + area in MS male mice compared to Non-MS counterparts (P = 0.0365) though the results are not significantly sex-dependent (see Additional file 3: Table S3). In addition, no significant differences were detected between WT groups in the percentage of CD68 + area within Iba1 + area (Fig. 3A, bottom). Moreover, no differences were found in the Iba1 + area in the other regions that we studied, including the hippocampus (dentate gyrus, CA1 and CA3) and amygdala (Additional file 1: Fig. S1A–E, see Additional file 3: Table S5).

Looking at the effect of MS in AD mouse models, it was recently reported that ms can induce the activation of microglia in PFC of App knock-in mice [26]. Here, MS alone did not have a statistically significant effect on microglial activation measures in PFC of 5xFAD mice (F(1,22) = 1.87, P = 0.1850) (Fig. 3A, down and B). However, there was an interaction effect due to MS dependent on sex effect (F(1,22) = 5.49, P = 0.0286) and a significant difference in general between sexes (F(1,22) = 4.77, P = 0.04). These effects were similar when comparing the percentages of CD68 + area within Iba1 + area between 5xFAD mouse groups (Sex: (F(1,16) = 4.56; P = 0.0485). Specifically, MS females had a higher percentage of Iba1 + area and CD68 + /Iba1 + area compared to MS males (P = 0.020; P = 0.0431, respectively). In addition, a MS-related trend towards increased Gal-3 immunoreactive area (a marker of microglial activation) was observed (Fig. 3C, E; Additional file 3: Table S3).

In AD, amyloid plaques and microglia are often associated. It was previously reported that MS enhanced AD pathology in 9-month-old male APPswe/Ps1dE9 mice [27]. In 5xFAD mice, we only noted a MS-related trend when looking at APP/Aβ immunoreactivity via the antibody 6E10 (Fig. 3D). However, with Congo red staining, we found that plaque burden was significantly different due to sex (F(1,23) = 5.11, P = 0.0336) and MS (F(1,23) = 10.9, P = 0.0031) as separate factors in PFC of 4-month-old 5xFAD mice (Fig. 3F). Furthermore, post hoc analysis did not show any MS-dependent differences in males (P = 0.4032) but did in females (P = 0.0276) (Fig. 3F). In comparison, we detected sex-dependent differences on plaque density in the hippocampus of 5xFAD mice (Additional file 1: Fig. S1F, G) with overall higher plaque density in MS females compared to MS males (P = 0.01). No differences were observed in the plaque load in the amygdala (Additional file 1: Fig. S1F–H, see Additional file 3: Table S5).

We further measured the levels of APP in PFC by aggregation state, but found no differences in the levels in the soluble fraction nor dense plaque fraction (Fig. 3G, H; see Additional file 3: Table S3). However, in the fraction corresponding to protofibrils (S3), levels differed significantly due to sex (F(1,16) = 11.9, P = 0.0033) and MS (F(1,16) = 5.87, P = 0.0276) as separate factors with post hoc analysis revealing significantly higher levels of APP in MS females compared with MS males (P = 0.0256). Altogether, our data suggest that MS could be playing a sex-specific role in worsening Aβ-related pathology in 5xFAD mice with females more adversely affected than males.

Though microglia are the resident immune cells in the brain, increasing evidence is showing that microglial crosstalk with peripheral immune cells may be altered in AD and play a role in driving the pathology [10]. As we suspect a link between MS and AD via neuroinflammation, we wanted to assess whether MS could affect the peripheral immune system even into adulthood. To do this, we performed flow cytometry analyses of peripheral inflammatory cells in the spleen, including inflammatory monocytes (Fig. 4A, B), cytotoxic T cells (Fig. 4C), dendritic cells (Fig. 4D), eosinophils (Fig. 4E), neutrophils (Fig. 4F), activated T helper (Fig. 4G) and (Fig. 4H) activated T cytotoxic cells (for gating strategy, see Additional file 2: Fig. S2).

In WT mice, we first observed that the proportion of inflammatory monocytes differed significantly due to MS (F(1,32) = 9.61, P = 0.004, Fig. 4B) but not sex (see Additional file 3: Table S4). Cytotoxic T cell populations significantly differed based on MS and sex separately (F(1,32) = 5.70, P = 0.0231; F(1,32) = 4.82, P = 0.0356, respectively) (Fig. 4C). Non-MS WT females had a proportionally larger population of cytotoxic T cells compared to Non-MS WT males (P = 0.0478), and Non-MS WT females had a larger proportion of cytotoxic T cells compared to MS WT females (P = 0.0281). No differences in the proportion of dendritic cells (Fig. 4D) were observed between WT groups (see Additional file 3: Table S4). Like the cytotoxic T cell population, ANOVA analysis of eosinophils (Fig. 4E) revealed significant separate effects due to sex (F(1,32) = 5.04, P = 0.0317) and MS (F(1,32) = 5.17, P = 0.0298; see Additional file 3: Table S4). Analysis of neutrophils (Fig. 4F) showed a significant difference between sexes (F(1,32) = 7.66, P = 0.0093) but not MS groups (see Additional file 3: Table S4). Here, post hoc analysis showed that Non-MS males had a significantly larger proportion of neutrophils compared to Non-MS females (P = 0.0413). Lastly, the activated T helper (CD4) population (Fig. 4G) was affected by MS (F(1,32) = 7.80; P = 0.0087) but not sex (see Additional file 3: Table S4). Post hoc analysis revealed a significant increase in this population in MS males compared to Non-MS group (P = 0.0379).

In 5xFAD mice, sex (F(1,28) = 12.4, P = 0.0015) and MS (F(1,28) = 13.2, P = 0.0011) separately had significant effects on the proportion of inflammatory monocytes. Post hoc analysis revealed a significant increase in the proportion of inflammatory monocytes in MS males compared to Non-MS males (P = 0.0123) and MS females (P = 0.0085). Likewise, sex (F(1,28) = 4.90, P = 0.0353) and MS (F(1,28) = 14.4, P = 0.0007) had separate, significant effects on the proportion of cytotoxic T cells. In this case, Non-MS females had a larger proportion of cytotoxic T cells compared to their MS counterparts (P = 0.0227). However, only MS affected the dendritic cell population (F(1,28) = 4.74, P = 0.0381) in 5xFAD mice, and interestingly, no significant differences were observed in the proportion of eosinophils between groups (see Additional file 3: Table S4). Like inflammatory monocyte and cytotoxic T cell populations, neutrophil populations were also affected by sex (F(1,28) = 7.14, P = 0.0124) and MS (F(1,28) = 5.59, P = 0.0253). Lastly, no significant differences were found in the populations of activated T helper (CD4) nor activated T cytotoxic (CD8) cells (Fig. 4H) between the 5xFAD mouse groups (see Additional file 3: Table S4). No interaction effects were detected on any of the cell populations observed in WT and 5xFAD mice. Altogether, these findings suggest that MS can alter the peripheral immune cell profile in the adulthood differently in female and male mice, which may have a detrimental effect in the context of AD.