Abstract

Introduction

   Multiple sclerosis (MS) affects the cerebral cortex, inducing cortical
   atrophy and neuronal and synaptic pathology. Despite the fact that
   women are more susceptible to getting MS, men with MS have worse
   disability progression. Here, sex differences in neurodegenerative
   mechanisms are determined in the cerebral cortex using the MS model,
   chronic experimental autoimmune encephalomyelitis (EAE).

Methods

   Neurons from cerebral cortex tissues of chronic EAE, as well as
   age-matched healthy control, male and female mice underwent RNA
   sequencing and gene expression analyses using RiboTag technology. The
   morphology of mitochondria in neurons of cerebral cortex was assessed
   using Thy1-CFP-MitoS mice. Oxygen consumption rates were determined
   using mitochondrial respirometry assays from intact as well as
   permeabilized synaptosomes.

Results

   RNA sequencing of neurons in cerebral cortex during chronic EAE in
   C57BL/6 mice showed robust differential gene expression in male EAE
   compared to male healthy controls. In contrast, there were few
   differences in female EAE compared to female healthy controls. The most
   enriched differential gene expression pathways in male mice during EAE
   were mitochondrial dysfunction and oxidative phosphorylation.
   Mitochondrial morphology in neurons showed significant abnormalities in
   the cerebral cortex of EAE males, but not EAE females. Regarding
   function, synaptosomes isolated from cerebral cortex of male, but not
   female, EAE mice demonstrated significantly decreased oxygen
   consumption rates during respirometry assays.

Discussion

   Cortical neuronal transcriptomics, mitochondrial morphology, and
   functional respirometry assays in synaptosomes revealed worse
   neurodegeneration in male EAE mice. This is consistent with worse
   neurodegeneration in MS men and reveals a model and a target to develop
   treatments to prevent cortical neurodegeneration and mitigate
   disability progression in MS men.

   Keywords: multiple sclerosis, experimental autoimmune
   encephalomyelitis, sex differences, mitochondria, cerebral cortex, RNA
   sequencing, estrogen, sex chromosomes

1. Introduction

   Multiple sclerosis (MS) is an autoimmune and neurodegenerative disease
   for which many disease-modifying treatments (DMTs) that target immune
   mechanisms exist. The discovery of neuroprotective treatments that
   target neurodegenerative mechanisms within the brain remains a goal.
   Genome-wide transcriptomics have shown differences in gene expression
   when comparing one brain region to another both when healthy and when
   affected by a disease, even within the same cell type, namely, neurons
   ([29]1), microglia ([30]2), astrocytes ([31]3–5), and oligodendrocytes
   ([32]6–8). Optimal therapeutic strategies aiming to reverse deleterious
   gene expression changes during disease may therefore need to be
   identified for distinct brain regions. One brain region of major
   interest in MS is the cerebral cortex. Cerebral cortex atrophy in MS
   correlates with worsening disability, as measured by the Expanded
   Disability Status Scale (EDSS) ([33]9–11). Cortical pathology in MS
   includes neuronal damage and synaptic loss ([34]12–17). A better
   understanding of neurodegeneration in the cerebral cortex at the
   molecular level is needed.

   In addition to brain region-specific gene expression differences, when
   healthy and when affected by disease, there are also sex-specific
   differences. Sex differences can be mediated by biologic effects,
   environmental effects, or both. Observations that sex differences
   extend beyond humans, for example, to female and male mice in the same
   environment of a vivarium, reveal the importance of biologic effects.
   Biologic sex differences can be due to differences in sex hormones
   (estrogen vs. testosterone), sex chromosomes (XX vs. XY), or both
   ([35]18, [36]19). Findings from the genotype-tissue expression (GTEx)
   project revealed the pervasiveness of sex differences in gene
   expression in humans. When 44 tissues were examined, 37% of all genes
   exhibited a sex bias in expression in at least one tissue ([37]20).
   Whole-genome expression profiles showed distinct sex-biased regulatory
   networks that were tissue-specific ([38]21), and sex differences in
   gene expression in the brain differed by the subregion ([39]20,
   [40]22). These sex differences in gene expression in the human brain
   when healthy carry implications for sex differences in the brain when
   affected by a disease.

   There are sex differences in MS ([41]23). MS affects women more
   frequently than men ([42]19, [43]24, [44]25). Despite the increased
   susceptibility of women, men with MS often demonstrate worse disability
   progression ([45]26–29) and brain subregion atrophy ([46]30–33). Thus,
   the effect of sex on inflammation vs. neurodegeneration appears to
   diverge. The female sex is more predisposed to develop autoimmune
   diseases, while the male sex appears to have a more neurodegenerative
   response to injury, reviewed in Voskuhl ([47]34). Is there a model to
   disentangle these sex differences? Experimental autoimmune
   encephalomyelitis (EAE) is the most widely used MS model, and it was
   used to develop many of the current treatments for MS. Similar to MS,
   EAE is characterized by both inflammation and neurodegeneration. There
   are several EAE models, and the EAE model used depends on the MS
   question asked, as reviewed in Voskuhl ([48]35). Relapsing remitting
   EAE in SJL mice shows an increase in susceptibility in female mice and
   is appropriate for the study of sex differences in immune responses in
   MS. EAE induced in SJL mice results in worse walking scores, spinal
   cord inflammation, and demyelination in female mice. In contrast,
   chronic EAE in C57BL/6 mice shows no sex difference in EAE walking
   scores or spinal cord pathology. Interestingly, the chronic EAE model
   exhibits brain atrophy as measured by in vivo magnetic resonance
   imaging (MRI) during later stages ([49]35). This atrophy occurs in the
   cerebral cortex and cerebellum, and it is correlated with neuronal cell
   loss and a decrease in synapses ([50]36–41). Cerebral cortex atrophy
   continues over time even after disability, as measured by standard EAE
   walking scores, has reached a plateau of severity. Such is also the
   case in MS patients whereby cerebral cortex atrophy by MRI continues
   even after disability, as measured by the EDSS, has reached a plateau
   of severity.

   Here, we address whether there are sex differences in molecular
   mechanisms in the cerebral cortex during chronic EAE, focusing on
   transcriptomics of cortical neurons and mitochondrial respiration in
   synaptosomes.

2. Materials and methods

2.1. Ethics statement

   All animal procedures were done in accordance with the guidelines of
   the National Institutes of Health and the Chancellor’s Animal Research
   Committee of the University of California, Los Angeles Office for the
   Protection of Research Subjects.

2.2. Animals and breeding strategy

   All mice used in this study were from the C57BL/6 J background. Neuron
   Specific Enolase (NSE) Cre:RiboTag mice were generated by crossing the
   rNSEII-Cre line ([51]42) with the RiboTag mice ([52]43)
   (B6J.129(Cg)-Rpl22^tm1.1Psam/SjJ, Jackson Lab, Bar Harbor, ME). In
   breeding of our mouse lines, Cre recombinase alleles were always
   inherited from female mice (mother) and not from male mice (father).
   Thy1-CFP-MitoS ([53]44) mice generated by fusing the cyan fluorescent
   protein (CFP) construct with the mitochondria targeting sequence of
   human cytochrome C oxidase subunit 8 A (COX8A) under the regulatory
   elements derived from mouse Thymus cell antigen 1, theta (Thy1), were
   obtained from Jackson Lab (B6.Cg-Tg(Thy1-CFP/COX8A)S2Lich/J) for the
   visualization of neuron-specific mitochondria in C57BL/6 mice. The
   regulatory construct of Thy1 gene defects in the sequence region was
   required for the expression in non-neuronal cells, making it specific
   to the neurons.

   Genotypes were assessed by PCR using the following primer sequences
   (5′-3′): Cre, GCACTGATTTCGACCAGGTT and GCTAACCAGCGTTTTCGTTC;
   LoxP-STOP-loxP-Rpl22-HA (RiboTag), GGGAGGCTTGCTGGATATG and
   TTTCCAGACACAGGCTAAGTACAC; and Thy1-CFP-MitoS, AAG TTC ATC TGC ACC G and
   TCC TTG AAG AAG ATG GTG CG.

   Animals were maintained under standard conditions in a 12-h dark/light
   cycle with free access to food and water ad libitum.

2.3. Active experimental autoimmune encephalomyelitis induction and clinical
scoring

   Mice were induced with active EAE as described previously for C57BL/6
   mice ([54]45). Briefly, mice aged 12–20 week were injected
   subcutaneously with myelin oligodendrocyte glycoprotein (MOG) amino
   acids 35–55 (200 μg/animal, Mimotopes, Richmond, VA) emulsified in
   Complete Freund’s Adjuvant, supplemented with Mycobacterium
   tuberculosis H37Ra (200 μg/animal, Becton, Dickinson and Company,
   Sparks, MD), over two sites drained by left inguinal and auxiliary
   lymph nodes in a volume of 0.1 mL/mouse. After 1 week, a booster
   immunization was applied subcutaneously over contralateral lymph nodes.
   Pertussis toxin (500 ng/mouse; List Biological Laboratories, Inc.,
   Campbell, CA) was injected intraperitoneally on days 0 and 2 ([55]46).
   The animals were monitored daily for EAE signs based on a standard EAE
   0–5 scale scoring system: 0, healthy; 1, complete loss of tail
   tonicity; 2, loss of righting reflex; 3, partial paralysis; 4, complete
   paralysis of one or both hind limbs; and 5, premoribund state, as
   described ([56]47).

2.4. Histological analysis

   Standard histological analysis methods were applied as we described
   previously ([57]42) with some modifications.

2.4.1. Tissue preparation

   Mice were deeply anesthetized in isoflurane and perfused transcardially
   with ice-cold 1x phosphate-buffered saline (PBS) for 5–7 min, followed
   by 4% paraformaldehyde for 5–7 min. Central nervous system (CNS)
   tissues were dissected and submerged in 4% paraformaldehyde overnight
   at 4°C, followed by 30% sucrose for >24 h. Brains were embedded in 7.5%
   gelatin/15% sucrose solution. The gelatin-embedded tissues were
   submerged in 4% paraformaldehyde overnight at 4°C, followed by 30%
   sucrose for >24 h, and stored at −80°C after flash frozen by dry ice.
   The 40-μm thick free-floating brain sagittal sections were prepared
   using a cryostat (Leica Biosystems, Nussloch, Germany) at −20°C.
   Tissues were collected serially and stored in 1xPBS with 0.1% sodium
   azide in 4°C.

2.4.2. Immunohistochemistry

   Prior to histological staining, 40-μm thick brain free-floating
   sections were thoroughly washed with 1xPBS to remove residual sodium
   azide. Tissues were incubated with 50% methanol/PBS (1:1) at RT,
   followed by 1xPBS washing. After heat antigen retrieval with heated
   10 mM citric acid, 0.05% Tween 20, pH6.0, all tissue sections were
   permeabilized with 0.3% Triton-X in 1xPBS for 10 min on ice and blocked
   with 10% normal serum in 1x phosphate-buffered saline with Tween-20
   (PBST) for 1 h. Tissues were then incubated with primary antibodies in
   1xPBST-2% normal goat serum overnight at 4°C. The next day, tissues
   were washed and incubated with secondary antibodies in 1xPBST-2% normal
   serum for 2 h at RT. Sections were mounted on slides, allowed to dry,
   and cover slipped in fluoromount G (Southern Biotech, Birmingham, AL)
   for confocal microscopy. The following primary antibodies were used:
   rabbit anti-neuronal nuclear protein (NeuN; at 1:500 dilution, Cell
   Signaling Technology, Danvers, MA, cat#12943), rat anti-glial
   fibrillary acidic protein (GFAP; at 1:500, Thermo Fisher, Rockford, IL,
   cat# 13–0300), rabbit anti-ionized calcium-binding adapter molecule 1
   (IBA1; at 1:1000, Wako Chemicals United States Inc., Richmond, VA, cat#
   019–19,741), rabbit anti-glutathione S-transferase pi (GST-pi; at
   1:1000, Enzo Life Sciences, Farmingdale, NY, cat#ADI-MSA-102,), mouse
   anti-parvalbumin (PV; at 1:1000, Swant, Burgdorf, Switzerland, cat#
   PV235), mouse anti-hemagglutinin (HA; at 1:500, BioLegend, San Diego,
   CA, cat#901522), and rabbit anti-HA (at 1:1000, Thermo Fisher
   Scientific, Rockford, IL, cat#MA5-27915). The following secondary
   antibodies were used at 1:500 dilution for staining the tissues: goat
   anti-rabbit Alexa Fluor Plus 647 (Thermo Fisher Scientific, Rockford,
   IL, cat#A32733), goat anti-rat Cy5 (Jackson Immunoresearch, West Grove,
   PA, cat# 112–175-167), goat anti-mouse- Cy5 (Jackson Immunoresearch,
   West Grove, PA, cat# 115–175-166), goat anti-rabbit TRITC (Jackson
   Immunoresearch, West Grove, PA, cat# 111–025-144), and goat anti-mouse-
   Cy3 (Jackson Immunoresearch, West Grove, PA, cat#115–545-166).

2.4.3. Confocal microscopy and image analysis

   Stained sections were examined and imaged using Olympus BX61 DSU
   fluorescence microscope with a Hamamatsu ORCA-FLASH4.0LT+ SCMOS CAMERA.
   NeuN and Thy1-MitoCFP images were captured at 243.75 × 243.75 um^2 with
   0.2 μm thickness of 20 stacks. Three images at the cortical region
   (bregma −3 to 0 mm/ lateral 0 to 1.2 mm/ Layer II-V) have been taken
   from one animal. All images were taken and processed using the
   integrated software program Olympus cellSens software version 4. ImageJ
   (NIH) was used to perform integration and analysis of images.

2.5. RNA co-immunoprecipitation, sequencing, and analysis

   Mice were exposed to a lethal dose of isoflurane and transcardially
   perfused with ice-cold 1x PBS for 3–5 min, followed by ice-cold 1%
   paraformaldehyde-1x PBS for 4–5 min. Cortex were collected and
   snap-frozen in liquid nitrogen. Tissues were stored at −80°C. Frozen
   tissues were homogenized on ice using a dounce homogenizer with freshly
   made homogenization buffer containing: 50 mM Tris–HCl (Invitrogen,
   Carlsbad, CA) pH 7.5, 100 mM KCl (Thermo Fisher Scientific, Rockford,
   IL), 12 mM MgCl[2] (Thermo Fisher Scientific, Rockford, IL), 1% Nonidet
   P-40 (Roche, Mannheim, Germany), 1 mM Dithiothreitol (DTT;
   Sigma-Aldrich, St. Louis, MO), 1× Proteinase Inhibitors Cocktail
   (Sigma-Aldrich, St. Louis, MO), 200 units/mL RNAsin (Promega, Madison,
   WI), 100ug/ml cycloheximide (Sigma-Aldrich, St. Louis, MO), and 1 mg/mL
   heparin (Sigma-Aldrich, St. Louis, MO). The homogenates were then
   centrifuged at 20,000× g, 4°C for 15 min. The supernatant was collected
   and incubated with pre-washed anti-HA conjugated magnetic beads
   (Pierce, Rockford, IL) overnight on a rotating wheel at 4°C. After
   removal of the supernatant, the magnetic beads were washed three times
   with high salt buffer containing 50 mM Tris (pH 7.5), 300 mM KCl, 12 mM
   MgCl[2], 1% Nonidet P-40, 1 mM DTT, and 100 μg/mL cycloheximide. Then,
   25 μL of proteinase K (4 mg/mL; Zymo Research, Irvine, CA) was added to
   the samples and incubated in a 55°C water bath for 30 min. After
   incubation, 300 μL tri-reagent was added, and the Direct-zol^™ RNA
   MiniPrep Plus kit (Zymo Research, Irvine, CA) was used for RNA
   isolation according to manufacturer’s instructions. RNA quantity and
   quality were measured using the NanoDrop 2000 spectrophotometer (Thermo
   Fisher Scientific, Rockford, IL) and Agilent High Sensitivity RNA
   ScreenTape System (Agilent, Santa Clara, CA).

   The RNA sequencing library was made using the KAPA Stranded RNA-Seq Kit
   (Kapa Biosystems, Wilmington, MA) which consists of mRNA enrichment,
   cDNA generation, end repair, A-tailing, adaptor ligation, strand
   selection, and PCR amplification. Barcoded adaptors were used for
   multiplexing samples in one lane. Sequencing was performed on Illumina
   HiSeq3000 for a paired-end 2x50 run. Data quality check was done on
   Illumina SAV. De-multiplexing was performed with the Illumina
   Bcl2fastq2 v 2.17 program. These procedures were performed at the UCLA
   Technology Center for Genomics and Bioinformatics core facility.

   Sequencing analyses and production of figures were performed in R (R
   Core team, 2023).[58]^1 Qualities of raw sequence data were examined
   using FastQC,[59]^2 and Trimmomatic ([60]48) was used for cleaning. The
   R package “QuasR” ([61]49) was used for the read alignment to the mouse
   genome (mm10) followed by counting at the gene level. The genes with
   count numbers less than 1 in more than half of the samples were
   filtered out. Differentially expressed genes between male control and
   male EAE, as well as between female control and female EAE mice, were
   identified with R package “edgeR” ([62]50). A false discovery rate of
   0.1 was used as the threshold of differentially expressed genes.
   Canonical pathway enrichment analysis was performed for differentially
   expressed genes using ingenuity pathway analysis (QIAGEN, Redwood
   City).[63]^3 GO (gene ontology) enrichment analysis was performed using
   the package “clusterProfiler” in R ([64]51).

2.6. Mitochondrial morphology analysis

   Quantitative analysis of mitochondrial morphology was conducted using
   the Mitochondria analyzer plugin for ImageJ ([65]52, [66]53).[67]^4 The
   plugin’s thresholding settings were adjusted as follows: subtract
   background at 1.25, sigma filter plus at 2, enhance local contrast at
   1.4, and block size at 1.05 with a C-value of 11. Subsequently, a
   comparison was made between the number of mitochondria per volume and
   the average volume and diameter of mitochondria in both the control and
   EAE groups.

2.7. Isolation of cortical synaptosomes

   Mice were deeply anesthetized with isoflurane and perfused
   transcardially with ice-cold 1× PBS for 1–2 min. Cerebral cortices were
   dissected out and subjected to a synaptosome isolation method as
   described by Dunkley et al. ([68]54) and Flynn et al. ([69]55) with
   slight modifications. Briefly, the cortex was rapidly isolated and
   rinsed with ice-cold sucrose/EDTA buffer (320 mM sucrose, 1 mM EDTA,
   and 5 mM HEPES, pH 7.4). It was then transferred to a pre-chilled
   Dounce glass homogenizer containing 1 mL of homogenizing buffer (320 mM
   sucrose, 1 mM EDTA, 5 mM HEPES, and 0.25 mM DTT, pH 7.4) and gently
   homogenized by 10 strokes with pestle A first, followed by 10 strokes
   with pestle B after the addition of an additional 1 mL of homogenizing
   buffer. The homogenate was then centrifuged at 1,200 g for 10 min at
   4°C. The supernatant was carefully layered on top of a discontinuous
   Percoll gradient (3, 10, 15, and 23% Percoll in homogenizing buffer) in
   a 10 mL centrifuge tube and centrifuged at 31,000 g for 5 min at 4°C
   using a JA-25.50 fixed-angle rotor in a Beckman Avanti J-25 high-speed
   centrifuge. Synaptosomes were isolated and recovered from the band
   between the 10 and 15% and the 15 and 23% Percoll layers
   ([70]Supplementary Figure 1[71]A). The isolated synaptosomes were
   washed with ice-cold sucrose/EDTA buffer and then diluted into an ionic
   medium (20 mM HEPES, 10 mM D-Glucose, 1.2 mM Na[2]HPO[4], 1 mM MgCl[2],
   5 mM NaHCO[3], 5 mM KCl, and 140 mM NaCl, pH 7.4) for the measurement
   of respiration. The final synaptosome pellet was then resuspended in an
   ionic medium, and the protein concentration was measured by the BCA
   Protein assay (Cat# [72]A53226, Thermo Fisher Scientific, Rockford,
   IL).

2.8. Seahorse respirometry assay

   The respirometry assay was performed in the Mitochondrial and
   Metabolism Core at UCLA. For monitoring respiration, freshly isolated
   synaptosomal protein in the ionic medium was aliquoted into a 96-well
   microplate (Seahorse Bioscience, North Billerica, MA) coated with a
   1:15,000 diluted polyethyleneimine (PEI). For intact respirometry, 8 μg
   per well of synaptosomal protein was loaded; for permeabilized
   synaptosomes, 6 μg per well was loaded for complex I substrates
   (pyruvate and malate) and 4 μg per well for complex II substrate
   (succinate with rotenone). The plate was centrifuged at 3,400 g for
   20 min at 4°C. Then, the medium was replaced with 175 μL Seahorse
   incubation Medium [3.5 mM KCl, 120 mM NaCl, 1.3 mM CaCl[2], 0.4 mM
   KH[2]PO[4], 1.2 mM Na[2]SO[4], 10 mM Glucose, 2 mM Pyruvate, 2 mM
   MgSO[4], 5 mM HEPES, and 0.05% (w/v) bovine serum albumin (BSA), pH
   7.4] for intact respirometry. For permeabilized synaptosomes, the
   samples were washed once with MAS (70 mM sucrose, 220 mM mannitol, 5 mM
   KH[2]PO[4], 5 mM MgCl[2], 2 mM HEPES, 1 mM EGTA, and 0.05% fatty
   acid-free BSA, pH 7.2) and brought to a final volume of 150 μL per well
   containing 4 mM ADP, 3 nM recombinant perfringolysin O (rPFO), and
   either 5 mM pyruvate and 5 mM malate for complex I or 5 mM succinate
   and 2 μM rotenone for complex II. The microplate was incubated and
   loaded into the Seahorse XF96 extracellular flux analyzer (Agilent,
   Santa Clara, CA). During measurement in intact synaptosomes, four
   compounds were injected in the following order: oligomycin A (3 μM),
   FCCP (2 μM), and antimycin A in combination with rotenone (2 + 2 μM),
   all at final concentrations.

   All experiments were performed at 37°C. The protocol for the
   measurement of oxygen consumption rate (OCR) is as previously described
   ([73]56). A minimum of three technical replicates were performed in
   each measurement.

   Synaptosome isolation and the respirometry assay were performed for
   3 days from EAE day 42 to 44 in the first experiment and for 4 days
   from EAE day 42 to 45 in the second experiment. Each procedure day had
   n = 2 per group from each group. The absolute OCR was normalized by the
   controls for each day to account for day-to-day variability in OCR. A
   minimum of 6 mice per group per experiment were used as biological
   replicates for each measurement to determine statistical differences
   per disease.

2.9. Electron microscopy

   In the study, 50 μg synaptosomes were fixed in 2.5% glutaraldehyde and
   2% formaldehyde in 0.1 M sodium cacodylate buffer for 1 h. After wash,
   samples were embedded in 4% agarose gel and post-fixed in 1% osmium
   tetroxide. After wash, samples were dehydrated through a graded series
   of ethanol concentrations and propylene oxide. After infiltration with
   Eponate 12 resin, the samples were embedded in fresh Eponate 12 resin
   and polymerized at 60°C for 48 h. Ultrathin sections of 70 nm thickness
   were prepared and placed on former carbon-coated copper grids and
   stained with uranyl acetate and lead citrate. The grids were examined
   using a JEOL 100CX transmission electron microscope at 60 kV, and
   images were captured by an AMT digital camera (Advanced Microscopy
   Techniques Corporation, model XR611; Electron Microscopy Core Facility,
   UCLA Brain Research Institute).

2.10. SDS–PAGE and western blotting

   Whole cortical lysate or synaptosome fraction were diluted in 4x
   Laemmli Sample Buffer (Bio-Rad, Hercules, CA, #1610747) with a
   supplement of DTT to a final concentration of 50 mM and were denatured
   by boiling for 5 min. For the detection of mitochondrial oxidative
   phosphorylation system (OXPHOS) proteins, instead of boiling, the
   samples were incubated at 37°C for 5 min. In the study, 5 μg of total
   protein were loaded into 4–15% Mini PROTEAN TGX precast gels (Bio-Rad,
   Hercules, CA, #4561084), and gel electrophoresis was performed in
   Mini-PROTEAN Tetra Vertical Electrophoresis Cell (Bio-Rad, Hercules,
   CA, #165–8004) under constant voltage at 200 V for 30 min.

   Proteins were transferred to nitrocellulose membrane (pore size: 0.2um,
   GE Healthcare Life science, Little Chalfont, United Kingdom, #10600004)
   in Mini-PROTEAN^® Mini Trans-Blot^® Module (Bio-Rad, Hercules, CA,
   #170–3935) under 400 mA constant current for 45 min at 4°C. Blots were
   blocked with 1% nonfat dry milk in 1xPBST and incubated with primary
   antibodies diluted in 1% nonfat dry milk /PBST overnight at 4°C. The
   next day, blots were washed 3 × 5 min in 1xPBST, probed containing
   HRP-conjugated secondary antibodies (1:10,000, Thermo Fisher
   Scientific, Rockford, IL, #G-21234 and #31430), diluted in 1xPBST for
   2 h at room temperature, and rinsed again for 3 × 5 min in 1xPBST.
   Detection was achieved using Clarity™ Western ECL Substrate (BioRad,
   Hercules, CA, #1705060) and imaged by Azure c300 (Azure Biosystems,
   Dublin, CA). The following primary antibodies were used: mtOXPHOS
   cocktail (at 1:250, for combined ATP synthase F1 subunit alpha (ATP5B),
   ubiquinol cytochrome c reductase core protein 2 (UQCRC2), cytochrome c
   oxidase I, mitochondrial (MTCO1), succinate dehydrogenase complex,
   subunit B (SDHB), and NADH:ubiquinone oxidoreductase subunit B8
   (NDUFB8) detections, Abcam, Waltham, MA, cat#ab110413), mouse
   anti-β-actin (at 1:3000, Cell Signaling Technology, Danvers, MA,
   cat#3700), mouse anti-postsynaptic density protein 95 (PSD95; at1:1000,
   BioLegend, San Diego, CA, cat#810301), mouse anti-glutamate receptor 2
   (GLUR2; at 1:100, Biolegend, cat#810501), rabbit anti-GST-pi (at 1:500,
   Enzo Life Sciences, Farmingdale, NY, cat#ADI-MSA-102), and rabbit
   anti-voltage-dependent anion channel ½ (VDAC1/2; at 1:500, Proteintech,
   Rosemont, IL, cat# 10866-1-AP).

2.11. Statistical analysis

   Mice were randomly assigned to experimental groups. Investigators were
   blinded to group allocation during data collection and analysis. Points
   represent individual animals. For all box-and-whisker plots, the center
   represents the median, the boxes represent the interquartile range, and
   the whiskers represent the highest and lowest values excluding
   outliers. Prism 9 (GraphPad) was used for all histological and
   respirometry data analysis. The statistical difference between groups
   was determined using the Mann–Whitney U-test. Statistical significance
   was reported as a *p < 0.05; **p < 0.01; *** p < 0.001; and
   ****p < 0.0001.

2.12. Data availability

   Datasets generated in this study are available in the GEO (Gene
   Expression Omnibus)[74]^5 under the accession number [75]GSE239455.

3. Results

3.1. Decreased neurons in the cerebral cortex in male mice with chronic
experimental autoimmune encephalomyelitis and sex differences in the
transcriptome of cortical neurons

   A decrease in neurons in cortical layer V of the cerebral cortex during
   chronic EAE has been described in a previous study ([76]36). Whether
   there is a sex difference in this loss of cortical neurons during
   chronic EAE remains unknown. Here, immunostaining of neurons (NeuN^+)
   was done on cerebral cortex tissues from male and female mice with
   chronic EAE (1st experiment EAE day 47; 2nd experiment EAE day 44) and
   from healthy age- and sex-matched controls. As expected, there was no
   sex difference in EAE walking scores in the C57BL/6 model
   ([77]Supplementary Figure 1). The average day of onset was a mean of
   15.25 in the 1st experiment (15.63 in female mice; 14.83 in male mice)
   and 14.47 in the 2nd experiment (14.90 in female mice; 13.86 in male
   mice), maximum scores (scale of 0 to 5) were 4.0 in the 1st experiment
   (4.0 in female mice; 4.0 in male mice) and 4.5 in the 2nd experiment
   (4.5 in female mice; 4.5 in male mice), and average scores over all
   days were 2.09 in the 1st experiment (1.98 in female mice; 2.20 in male
   mice) and 2.18 in the second experiment (2.04 in female mice; 2.39 in
   male mice). Despite no sex difference in standard EAE walking scores,
   the degree of cortical neuron loss in male mice with EAE compared to
   healthy male mice was more pronounced than in female mice with EAE
   compared to healthy female mice ([78]Figure 1A; [79]Supplementary
   Figure 2).

Figure 1.

   [80]Figure 1
   [81]Open in a new tab

   Worse neuronal loss in the cerebral cortex in male mice with chronic
   EAE and sex differences in the neuron-specific transcriptome. (A)
   Quantification of NeuN^+ cortical neurons from healthy age- and
   sex-matched controls (NL, white) and EAE (EAE, gray), with male mice
   represented in blue and female mice in red. Data from two separate
   experiments (circle: 1st experiment, triangle: 2nd experiment). EAE
   showed a significant decrease of NeuN^+ number as compared to control
   in male (p < 0.0001) and female mice (p = 0.0014). Additionally, male
   mice showed a significant decrease in NeuN^+ number compared to female
   mice in EAE (p = 0.0028) but not in controls (p = 0.5919; n = 11 mice
   in EAE, n = 9 mice in control; **** p < 0.0001, **p < 0.01). p-values
   were calculated by the Mann–Whitney U-test. (B) RNA enrichment
   analysis. In immunoprecipitated RNA samples from FIGURE 1
   (Continued)NSE-Cre:RiboTag mice, neuronal markers (Nestin, NeuroD1, and
   Tbr1) were enriched, while astrocytes (Aldhl1l and Gfap),
   oligodendrocytes (Mbp and Mog), and microglia genes (Cd11b and Tmem119)
   showed no enrichment. Asterisks: FDR < 0.1 (edgeR). (C) Quantification
   of HA co-localization with NeuN^+ (neurons), PV^+ (interneurons),
   GFAP^+ (astrocytes), GST-pi^+ (oligodendrocytes), and IBA1^+ cells
   (microglia/macrophages). n = 3, error bars represent SEM. ****
   p < 0.0001. p-values were calculated by the one-sample t-test. (D)
   Representative 40× images of NeuN (red), HA (green), and DAPI (blue) in
   a merged form showing co-localization (yellow), all from the cerebral
   cortex of NSE-Cre:RiboTag mice. Bar = 40um. (E) Differentially
   expressed gene (DE gene) numbers comparing EAE male mice with control
   male mice and EAE female mice with control female mice. FDR < 0.1,
   logFC<−0.25 or 0.25 < logFC. EAE n = 5 (male) or n = 4 (female) and
   control n = 4 (male) or n = 4 (female). (F) Top 10 canonical pathways
   for upregulated and downregulated genes in male EAE mice compared to
   male control mice (left) and female EAE mice compared to female control
   mice (right; FDR < 0.1, logFC<−0.25 or 0.25 < logFC). Downregulated
   genes in EAE male mice were enriched in mitochondrial dysfunction
   (black), oxidative phosphorylation (red), and sirtuin signaling (green)
   pathways. Pathways are shown in a descending order of significance.
   (G–I) Volcano plots to visualize the genes in mitochondrial dysfunction
   (G), oxidative phosphorylation (H), and sirtuin signaling (I) pathways,
   within each volcano panel male EAE mice compared to male control mice
   (left) and female EAE mice compared to female control mice (right).
   Green line: FDR = 0.1.

   Changes in gene expression in cortical neurons during chronic EAE in
   each sex were then investigated using RiboTag technology ([82]8,
   [83]43). NSE-Cre mice were crossed with RiboTag mice (NSE-Cre:RiboTag)
   to generate mice expressing HA-tagged ribosomal protein specifically in
   neurons. Cerebral cortex tissues were acquired, and neuron-specific
   RNAs were isolated. The specificity of HA labeling in NSE-Cre:RiboTag
   mice was confirmed through RNA-seq analysis ([84]Figure 1B) and
   immunostaining ([85]Figures 1C,[86]D). At the RNA level, neuronal genes
   (Nestin, NeuroD1, and Tbr1) were enriched in the immunoprecipitated
   neuronal RNA, whereas astrocytes (Aldhl1l and Gfap), oligodendrocytes
   (Mbp and Mog), and microglia genes (Cd11b and Tmem119) were not
   enriched ([87]Figure 1B). At the protein level, HA staining
   co-localized with the neuron marker (NeuN), while there was no
   significant co-localization with other CNS cell markers:
   parvalbumin-expressing interneurons (PV), astrocytes (GFAP),
   oligodendrocytes (GST-pi), or microglia/macrophages (IBA1; [88]Figures
   1C,[89]D). These results validated the neuron-specificity of RNAs
   isolated from the cerebral cortex of NSE-Cre:RiboTag mice.

   Genome-wide transcriptome analyses were then carried out on
   neuron-specific RNAs from male and female NSE-Cre: RiboTag mice with
   chronic EAE, as well as from age- and sex-matched healthy controls. In
   male mice with EAE as compared to male controls, 7,487 genes were
   differentially expressed (upregulated 3,601, downregulated 3,886),
   while female mice with EAE compared to female controls showed only 495
   genes differentially expressed (upregulated 336, downregulated 159;
   FDR < 0.1, logFC<−0.25 or 0.25 < logFC; [90]Figure 1E). A canonical
   pathway analysis (Ingenuity Pathway Analysis)[91]^6 revealed that the
   top three pathways showing the most enrichment in male mice during EAE
   were the mitochondrial dysfunction, oxidative phosphorylation, and
   sirtuin signaling pathways, with each downregulated ([92]Figure 1F). In
   contrast, female mice showed less significant enrichment of pathways
   during EAE. Male-specific downregulation of mitochondria-related genes
   was further confirmed in the additional gene set enrichment analysis
   using GO (Gene Ontology) database, while female mice showed almost no
   significant enrichment ([93]Supplementary Figure 3). Volcano plots
   showing the expression of genes in each of the top three pathways most
   enriched in male mice are shown for both sexes (see [94]Figures
   1G–[95]I). In contrast to EAE male mice, the majority of pathway genes
   in EAE female mice were below the threshold (FDR = 0.1) and not
   significantly different when comparing EAE to the sex-matched healthy
   control.

   Together, transcriptome changes in cortical neurons during chronic EAE
   were robust in male mice and minimal in female mice, and differences in
   expression of mitochondria-related genes were primarily responsible for
   this sex difference.

3.2. Mitochondrial morphology in neurons of the cerebral cortex is abnormal
in male mice with experimental autoimmune encephalomyelitis

   Mitochondrial fusion and fission play a vital role in maintaining
   mitochondrial homeostasis. Our transcriptome analyses of neurons in the
   cerebral cortex during EAE revealed dysregulation in the expression of
   genes involved in fusion and fission (Fis1, Dnm1l, Mfn1, and Opa1) in
   male mice but not female mice ([96]Supplementary Table 1).

   Examination of the morphology of mitochondria in cortical neurons was
   then done using Thy1-CFP-MitoS mice ([97]Figure 2A). EAE male mice
   showed a significant reduction in the numbers of cortical MitoS^+
   mitochondria compared to normal control male mice ([98]Figure 2B).
   Moreover, MitoS^+ morphology in EAE male mice exhibited a significant
   decrease in volume ([99]Figure 2C) and diameter ([100]Figure 2D). In
   contrast, no significant differences in cortical MitoS^+ number,
   volume, or diameter were observed in EAE female mice compared to normal
   control female mice. Notably, a significant decrease in cortical
   MitoS^+ mitochondria volume and diameter was also observed when
   comparing EAE male to EAE female mice ([101]Figures 2C,[102]D). This
   demonstrated that neuron-specific mitochondrial morphology was affected
   in the cerebral cortex of EAE male mice more than EAE female mice.

Figure 2.

   [103]Figure 2
   [104]Open in a new tab

   Mitochondria morphology in the cerebral cortex. (A) Representative
   Thy1- CFP-MitoS^+ (MitoS^+) images in the cerebral cortex of male
   control (Male NL) and EAE (Male EAE) mice; female control (Female NL)
   and EAE (Female EAE) mice. Bar = 40 μm. Inset: Magnification at 100×.
   (B–D) Morphometric quantification of the MitoS^+ mitochondrial number
   (B), volume (C), and diameter (D) from control (NL, white) and EAE
   (EAE, gray); male mice (blue) and female mice (red). EAE male mice
   showed a significant decrease in MitoS^+ number (p = 0.0159, B), volume
   (p = 0.0159, C), and diameter (p = 0.0159; D) as compared to control
   male mice. n = 4–5 per group. MitoS^+ mitochondrial volume (p = 0.0079;
   C) and diameter (p = 0.0079; D) were significantly decreased in EAE
   male mice compared to EAE female mice (*p < 0.05; **p < 0.01). p values
   were calculated by the Mann–Whitney U-test.

3.3. Mitochondrial respiration in cortical synaptosomes is impaired in males
with experimental autoimmune encephalomyelitis

   Since cortical neurons showed changes in gene expression in the
   mitochondrial and oxidative pathways in EAE male mice, we next
   investigated mitochondrial function in EAE and healthy control male and
   female mice. To determine mitochondrial function in cortical neurons,
   we isolated synaptosomes since they are derived from the termini of
   neurons and contain mitochondria. The isolation of synaptosomes was
   verified using electron microscopy to confirm the presence of
   membrane-enclosed structures housing mitochondria ([105]Supplementary
   Figure 4[106]B) and by western blotting to confirm the enrichment of
   synaptic markers and the presence of mitochondrial proteins
   ([107]Supplementary Figure 4[108]C). Following this validation,
   isolated synaptosomes underwent functional respirometry assays.

   A schematic of a mitochondrial respirometry profile from intact
   synaptosomes is shown in [109]Figure 3A. Intact synaptosomes derived
   from EAE male mice compared to normal control male mice showed a
   significant reduction of oxygen consumption rate (OCR) in basal
   ([110]Figure 3B), ATP-linked ([111]Figure 3C), maximal ([112]Figure
   3D), and reserve ([113]Figure 3E) parameters. In contrast, EAE female
   mice compared to control female mice only showed a trend of decrease in
   two of the four parameters: basal OCR ([114]Figure 3B) and maximal OCR
   ([115]Figure 3D).

Figure 3.

   [116]Figure 3
   [117]Open in a new tab

   Intact synaptosome oxygen consumption. (A) Representative mitochondrial
   respiratory profile from intact synaptosomes. The figure shows across
   time (in min) basal OCR (blue); OCR after inhibition of ATP synthase
   through the addition of oligomycin (pink); maximal stimulatory capacity
   of synaptosomes in response to the uncoupling agent FCCP (orange);
   reserve (calculated from maximal OCR minus basal OCR, green); and
   non-mitochondrial respiration (measured after combined treatment with
   rotenone and antimycin A to completely inhibit mitochondrial
   respiration, yellow). (B–E) Quantitative analysis of mitochondrial OCR
   in EAE mice compared to normal healthy control mice for basal (B),
   ATP-linked (C), maximal (D), and reserve (E) parameters using intact
   cortical synaptosomes from control (NL, white) and EAE (EAE, gray);
   male mice (blue) and female mice (red). Data shown are from two
   separate experiments (circle symbol: 1st experiment, triangle symbol:
   2nd experiment). EAE male mice showed a significant decrease in basal
   OCR (p = 0.0061; B), ATP-linked (p = 0.014; C), maximal (p = 0.0012;
   D), and reserve (p = 0.0191; E) parameters, each as compared to NL
   control male mice. In contrast, EAE female mice compared to NL female
   mice showed a trend of decrease in two of four parameters: basal OCR
   (p = 0.0501; B) and maximal OCR (p = 0.0501; D) with no significant
   differences in any parameters. OCR in EAE mice normalized by that NL
   control. n = 14 per group. * p < 0.05; ** p < 0.01. p values were
   calculated by the Mann–Whitney U-test.

   Next, we conducted complex-specific respiration assays using
   permeabilized synaptosomes, using a combination of either pyruvate and
   malate (substrates for complex I) or succinate (substrate for complex
   II) and rotenone (complex I inhibitor; see a diagram in [118]Figure
   4A). There were no differences in any complex I-mediated respiration
   parameters in EAE compared to control in either sex ([119]Figures
   4E–[120]G). However, there were effects of EAE on complex II-mediated
   respiration parameters, and this revealed a sex difference. EAE male
   mice showed a significant decrease in OCR complex II-mediated
   respiration for state 3 ([121]Figure 4B), state 4o ([122]Figure 4C),
   and uncoupled ([123]Figure 4D) parameters each as compared to control
   male mice. In contrast, OCR in EAE female mice was not decreased for
   complex II-mediated respiration.

Figure 4.

   Figure 4
   [124]Open in a new tab

   Permeabilized synaptosome oxygen consumption. (A) Oxidation
   phosphorylation diagram for complex I- (orange) and complex II- (green)
   mediated respirometry. Permeabilized synaptosomes were treated with two
   different substrate and/or inhibitor combinations: pyruvate in
   combination with malate for complex I-mediated respirometry or with
   succinate with rotenone for complex II-mediated respiration.
   Permeabilized synaptosomes were stimulated with ADP to induce coupled
   respiration (State 3), followed by oligomycin to induce proton-leak
   respiration (State 4o), as well as with the mitochondrial uncoupler
   FCCP to reveal maximal uncoupled respiration (uncoupled). (B–G)
   Quantitative analysis of mitochondrial OCR in EAE mice compared to NL
   control for state 3 (B,E), state 4o (C,F), and uncoupled (D,G)
   parameters in complexFIGURE 4 (Continued)II-mediated (B–D) and complex
   I-mediated (E–G) respirometry using permeabilized cortical synaptosomes
   from control (NL, white) and EAE (EAE, gray); male mice (blue) and
   female mice (red). Data shown are from two separate experiments (circle
   symbol: 1st experiment, triangle symbol: 2nd experiment). EAE male mice
   showed a significant decrease in OCR complex II-mediated respiration
   for state 3 (p = 0.0016, K), state 4o (p = 0.0141, L), and uncoupled
   (p = 0.00102, M) parameters, each as compared to NL control male mice.
   In contrast, OCR in EAE female mice was intact in complex II-mediated
   respiration. There were no differences in any complex I-mediated
   respiration parameters in EAE compared to NL in either sex. n = 14 per
   group. * p < 0.05; ** p < 0.01. p values were calculated by the
   Mann–Whitney U-test.

   Finally, we compared functional respiration parameters in healthy
   control male and female mice to ascertain if absolute OCR differences
   might exist independent of disease. This was not the case, since no
   significant OCR differences were observed between male and female mice
   in the control groups across basal, ATP-linked, maximal, and reserve
   parameters in intact synaptosomes ([125]Supplementary Figure 5).
   Additionally, there were no differences in complex II- or I-mediated
   respiration assays for state 3, state 4o, and uncoupled parameters in
   permeabilized synaptosomes ([126]Supplementary Figure 6).

   Mitochondria respirometry assays demonstrated that EAE induces
   impairment of mitochondrial respiration in cortical synaptosomes, which
   was worse in male mice than in female mice, particularly in complex
   II-mediated respiration.

4. Discussion

   Neuronal transcriptome, mitochondrial morphology, and functional
   respirometry assays in synaptosomes revealed worse outcomes in the
   cerebral cortex of male EAE mice than in female EAE mice of the C57BL/6
   strain. This is consistent with worse neurodegeneration in MS men and
   reveals a model and a target to develop treatments with the potential
   to prevent cortical pathology and mitigate disability progression in MS
   men.

   In addition to considering translation of our preclinical findings to
   MS, one can also discuss possible mechanisms leading to our preclinical
   observations. Findings that male mice have more significant
   mitochondrial and oxidative gene expression changes in cortical
   neurons, abnormal mitochondrial morphology, and decreased oxygen
   consumption rates in synaptosomes could be driven by an effect of sex
   chromosomes, sex hormones, or both. Sex chromosomes and sex hormones
   can act in a synergistic or antagonistic manner in a given process
   ([127]57). Compensatory mechanisms may have arisen during evolution to
   promote survival of each sex, reaching an optimal balance between sex
   chromosome and sex hormone influences. How each sex reaches this
   balance is likely distinct given the differences in sex hormones
   (estrogen vs. testosterone) and sex chromosomes (XX vs. XY) ([128]58).

   With regard to autoimmunity and neuroprotection, we hypothesize that XX
   is proinflammatory and estrogen is neuroprotective. This is consistent
   with female mice developing more robust immune responses to infections
   and being predisposed to autoimmunity but having more resilience to
   brain injury. It follows that male mice would have less robust immune
   responses to infections, be less susceptible to autoimmune diseases,
   and exhibit more neurodegeneration during brain injury. Evidence
   supporting this hypothesis is supported by the use of the four core
   genotype model ([129]59). The XX genotype was shown to be more
   proinflammatory in the relapsing remitting EAE model in the SJL genetic
   background, in the chemically induced lupus model in the SJL, and
   spontaneous lupus in the NZM.2328 genetic background ([130]60,
   [131]61). Conversely, the XY genotype was shown to confer worse
   neurodegeneration in the chronic EAE model on the C57BL/6 background
   ([132]62) and in the human amyloid precursor protein model of
   Alzheimer’s disease ([133]63). Notably, the proinflammatory effects of
   XX in active EAE in SJL mice were verified by using adoptive transfers
   of immune cells that were XX or XY into a common recipient ([134]61).
   More neurodegenerative effects of XY compared to XX in active EAE in
   C57BL/6 mice were shown using bone marrow chimeras to ensure the same
   immune system with the central nervous system that differed in sex
   chromosomes ([135]62). Findings here where the neuronal transcriptome,
   mitochondrial morphology, and functional respirometry assays in
   synaptosomes demonstrated more abnormalities in the cerebral cortex of
   male C57BL/6 mice with chronic EAE are consistent with a more
   neurodegenerative effect of XY using this same chronic EAE model
   ([136]62). Limitations to this interpretation include the following:
   Bone marrow chimeras were not used here, so male immune cells may have
   contributed to worse neuronal outcomes in the cerebral cortex of males
   ([137]64). Findings here could theoretically be due to testosterone in
   male mice causing worse neurodegeneration or estrogen in female mice
   being neuroprotective. These are not mutually exclusive. Repeating
   experiments in gonadectomized mice to remove activational effects of
   sex hormones would be informative. There could also be organizational
   effects of sex hormones, e.g., an effect of differences in testosterone
   vs. estrogen during development, which cannot be addressed merely by
   gonadectomy at adult ages. To investigate organizational effects of sex
   hormones, experiments using the four core genotype mice ([138]59) are
   needed. Comparison would be made between XY mice with ovaries vs.
   XY-Sry mice with testes, where developmental hormones differ, and sex
   chromosome complement is the same. These are interesting future
   directions to discern mechanisms underlying sex differences observed
   here in cortical neurons during chronic EAE.

   Further discussion is warranted regarding the role of sex hormones on
   molecular outcomes observed here. An effect of sex chromosome
   complement is not mutually exclusive of a contribution of circulating
   testosterone or estradiol in gonadally intact male and female adult
   mice. Indeed, both testosterone and estradiol have neuroprotective
   properties in MS models. Specifically, testosterone has been shown to
   bind androgen receptors to induce remyelination and can also be
   protective via its conversion to estradiol by aromatase in the brain
   ([139]65–67). Estradiol mediates neuroprotection by activating estrogen
   receptors alpha (ERα) and/or beta (ERβ) in the central nervous system
   of EAE mice. ERα has been shown to act on astrocytes to confer
   protection by modifying immune cell recruitment ([140]45, [141]68,
   [142]69), while ERβ acts on oligodendrocytes and microglia to induce
   remyelination ([143]70–72). In conditions other than MS or EAE, ERα and
   ERβ expression was demonstrated in mitochondria, and ligation of
   estrogen receptors conferred protection in mitochondrial function
   within neurons as well as other cells and tissues ([144]73–76).

   In humans, the loss of either testosterone with andropause or estradiol
   with menopause is believed to be deleterious for neurogenerative
   changes in MS, brain aging, and Alzheimer’s disease, as reviewed in
   Voskuhl and Itoh ([145]18). The difference in timing of andropause and
   menopause may play a role in men having worse disability progression
   and more gray matter substructure atrophy than women. Data for these
   findings were generally derived from MS patients aged 18–60 years, with
   the mean age of early 40s. Testosterone declines in men by 1–2% per
   year starting at the age of 30 years. Meanwhile, women in their early
   40s maintain the levels of estradiol since menopause occurs at a mean
   age of 52 years. Indeed, when women undergo menopause and have loss of
   estradiol, it is associated with disability worsening ([146]77–83).
   Thus, worse disability progression in men compared to women with MS may
   be age-dependent, e.g., an age when men exhibit effects of andropause
   but before women undergo menopause ([147]18).

   In summary, the ultimate effect of biologic sex on adults with MS is
   due to changes in sex hormones during aging in the setting of
   inheritance of a given sex chromosome complement. Sex hormones and sex
   chromosomes can have opposing effects on disease, in part through
   differential effects on the immune system vs. the CNS. Even when there
   is no overall sex difference in a clinical outcome, there could be very
   important disease-modifying effects of sex hormones or sex chromosomes
   that reach equivalence regarding the net impact on disease. This is the
   rationale for studying the effects of sex hormones and sex chromosomes
   not only in diseases where there is a sex difference but also in
   diseases not characterized by an overall sex difference. It can lead to
   the discovery of naturally existing disease modifiers in each sex.

Data availability statement

   Datasets generated in this study are available in the GEO (Gene
   Expression Omnibus; www.ncbi.nlm.nih.gov/geo/) under the accession
   number [148]GSE239455.

Ethics statement

   All animal procedures were done in accordance to the guidelines of the
   National Institutes of Health and the Chancellor’s Animal Research
   Committee of the University of California, Los Angeles Office for the
   Protection of Research Subjects. The study was conducted in accordance
   with the local legislation and institutional requirements.

Author contributions

   NI: Writing – review & editing, Formal analysis, Investigation,
   Methodology. YI: Formal analysis, Investigation, Methodology, Writing –
   review & editing. LS: Formal analysis, Investigation, Methodology,
   Writing – review & editing. RV: Writing – review & editing,
   Conceptualization, Funding acquisition, Writing – original draft.

Acknowledgments