key: cord-0286616-5phhqeri
authors: Cheng, Mandy I.; Riggan, Luke; Li, Joey H.; Tafti, Rana Yakhshi; Chin, Scott; Ma, Feiyang; Pellegrini, Matteo; Hrncir, Haley; Arnold, Arthur P.; O’Sullivan, Timothy E.; Su, Maureen A.
title: Sex differences in NK cells mediated by the X-linked epigenetic regulator UTX
date: 2022-04-29
journal: bioRxiv
DOI: 10.1101/2022.04.21.489076
sha: c6ed4d93a99b98e8f23f4ae1bc99350613a6e84c
doc_id: 286616
cord_uid: 5phhqeri

Viral infection outcomes are sex-biased, with males generally more susceptible than females. Paradoxically, the numbers of anti-viral natural killer (NK) cells are increased in males compared to females. Using samples from mice and humans, we demonstrate that while numbers of male NK cells are increased compared to females, they display impaired production of the anti-viral cytokine IFN-γ. These sex differences were not due solely to divergent levels of gonadal hormones, since these differences persisted in gonadectomized mice. Instead, these differences can be attributed to lower male expression of X-linked Kdm6a (UTX), an epigenetic regulator which escapes X inactivation in female NK cells. NK cell-specific UTX deletion in females phenocopied multiple features of male NK cells, which include increased numbers and reduced IFN-γ production. Integrative ATAC-seq and RNA-seq analysis revealed a critical role for UTX in the regulation of chromatin accessibility and gene expression at loci important in NK cell homeostasis and effector function. Consequently, NK cell-intrinsic UTX levels are critical for optimal anti-viral immunity, since mice with NK cell-intrinsic UTX deficiency show increased lethality to mouse cytomegalovirus (MCMV) challenge. Taken together, these data implicate UTX as a critical molecular determinant of NK cell sex differences and suggest enhancing UTX function as a new strategy to boost endogenous NK cell anti-viral responses.

indicating that gonadal hormones are not solely responsible for sex differences in NK cells.

Thus, we hypothesized that chromosomal complement, in particular X chromosome dosage, 138 may also play an important role.

139 140 X-linked UTX escapes X-inactivation and has higher expression in female NK cells.

While 46XX females undergo X chromosome inactivation (XCI) to control dosages of X-linked 142 genes, a subset of genes escapes XCI (termed XCI escapees), often resulting in higher 143 expression in females compared to males. Thus, XCI escapees are prime candidates for 144 mediating phenotypic sex differences in NK cells. Five genes (XIST, DDX3X, KDM6A, EIF2S3, 145 KDM5C ) have previously been identified as XCI escapees in both humans and mice 30 . XIST 146 was excluded from further analysis because it is not expressed in male cells due to its known 147 role in X chromosome inactivation in female cells 1 . All 4 remaining genes were significantly 148 downregulated in male vs. female NK cells, from humans (Fig. 2a) and mice (Fig. 2b) . The 149 greatest differential expression in both human and mouse NK cells was seen with Kdm6a (also 150 known as UTX) (Fig. 2a,b) . Male NK cells also expressed lower UTX protein levels compared to 151 female NK cells in mice (Fig. 2c,d) . These data indicate that expression levels of Kdm6a (UTX) 152 is sex-biased in NK cells.

In NK cells derived from gonadectomized mice, differences persisted in Kdm6a 154 transcript levels (Fig. 2e) and UTX protein levels (Fig. 2f) . Additionally, using the four core 155 genotype (FCG) mouse model, which uncouples sex chromosome complement (XX or XY) and 156 gonadal sex organ (ovaries or testes) 31 , Kdm6a transcript levels were also lower in mice with 157 one X chromosome (XY) independent of gonadal composition (Extended Data Fig. 2a) .

Together these findings suggest that increased UTX expression in female mice is not due to 159 hormonal effects and instead point to a primary role for X chromosome dosage. 

To determine if UTX mediates sex differences in NK cells, we generated mice with a conditional 164 deletion of UTX in NK cells (Kdm6a fl/fl x Ncr1 Cre+ , referred to as UTX NKD hereafter) with WT 165 (Kdm6a fl/fl x Ncr1 Cre-) littermates used as controls. To control for gonadal hormone and sex 166 chromosome effects, comparisons were made only in female mice (female WT vs. female 167 UTX NKD littermates). We confirmed decreased UTX protein expression in NK cells from UTX NKD 168 mice using flow cytometry (Extended Data Fig. 2b,c) . Similar to male mice, female UTX NKD 169 mice displayed increased splenic NK cell frequency (Fig. 3a,b) and absolute numbers (Fig. 3c) 170 in the spleen, blood, lungs, liver and bone marrow, demonstrating that this increase was not 171 tissue specific. Furthermore, IFN-γ protein production in response to IL-12 and IL-18 stimulation 172 was decreased in NK cells from UTX NKD vs. WT mice (Fig. 3d-f) . These results implicate UTX in 173 limiting NK cell numbers and promoting IFN-γ production, suggesting divergent UTX levels may 174 play a causal role in NK cell sex differences. (mBMCs) (Fig. 4a) . Using mBMCs allowed for an internally controlled experiment to minimize 185 environmental confounding factors. Principle Component Analysis (PCA) of both ATAC-seq and 186 RNA-seq data revealed that samples clustered together by genotype (Fig. 4b) . 4e) . Major pathways such as immune system process, cytokine production, IFN-γ 203 production, lymphocyte activation, and immune effector process were associated with 204 decreased expression in UTX NKD (Clusters 1, 2, 3, and 6) (Fig. 4e) . At the same time, pathways 205 such as developmental process, biosynthetic process, and metabolic process were significantly 206 associated with increased expression in UTX NKD (Clusters 4 and 5) (Fig. 4e) 

and Tbx6) 39 family TFs were more significant and had a higher percentage of target motifs 215 associated with clusters displaying decreased accessibility in UTX NKD (Clusters 1, 2, 3, and 6) 216 (Fig. 4f) . Conversely, TFs associated with proliferation, differentiation, and metabolism in the 217 zinc finger family TFs (KLF1, KLF5, KLF6, KLF14, Sp2 and Sp5) 40 were more significantly 218 associated with clusters displaying increased accessibility (Clusters 4 and 5) (Fig. 4f) . These 219 data suggest that UTX poises the chromatin accessibility of several genes at steady state 220 known to influence NK cell fitness and effector responses, while also controlling genome-wide 221 accessibility of transcription factor binding sites implicated in these processes. 

Previous studies in mice have demonstrated that Bcl-2 can inhibit the pro-apoptotic function of Bim to promote NK cell survival 41 , thus, we interrogated the expression of these proteins in 238 UTX NKD NK cells. While naïve UTX NKD NK cells showed increased intracellular protein 239 expression of Bcl-2 compared to WT NK cells (Fig. 5b,d) , UTX-deficient NK cells also displayed 240 a modest increase in intracellular Bim levels (Fig. 5c,e) . Importantly, the Bcl-2:Bim ratio was 241 significantly higher in UTX NKD NK cells (Fig. 5f) , suggesting UTX-deficiency likely results in a 242 higher proportion of pro-survival proteins present in NK cells. Notably, male NK cells also 243 displayed a significant increase in Bcl-2:Bim ratio, which may underlie the expanded NK cell 244 numbers observed in male mice (Fig. 5g) . These results implicate UTX in regulation of NK cell 245 fitness to restrict numbers at homeostasis. 

WT NK cells at rest (Fig. 6b) . Similarly, male NK cells, which express lower levels of UTX 255 (Fig.2c,d) , also had a similar decrease in Ifng transcript levels at rest (Fig. 6c) 

To determine whether UTX deficiency also led to decreased chromatin accessibility and Ifng and Csf2 transcripts in NK cells from UTX NKD mice at D1.5 PI (Fig. 6f) . Similarly, UTX NKD 264 NK cells showed decreased IFN-γ protein expression in UTX NKD NK cells on D1.5 PI indicating 265 that UTX expression in NK cells is required for optimal IFN-γ production following viral infection 266 (Fig. 6g,h) . To confirm whether dosage of UTX expression in mature NK cells associates with 267 NK cell production of IFN-γ during viral infection, we generated transgenic mice to achieve 

Together, these findings support a model in which divergent UTX expression contributes to sex 299 differences in NK cell numbers and effector function.

Our findings indicate that UTX restricts NK cell numbers at steady state, since NK cells 301 are increased at baseline in UTX NKD mice. This is in contrast to UTX deficiency in other immune 302 cell types, which have been reported to result in moderate (CD8 + and CD4 + T cells) or severe 303 (iNKT) decreases in peripheral cell numbers 44-46 . Interestingly, T cell-specific UTX-deficiency is 304 associated with CD8 + T cell accumulation during viral infection. Thus, it is possible that UTX-305 mediated gene programs that inhibit CD8 + T cell numbers during inflammation are shared by NK 306 cells at rest 46 . Indeed, increased Bcl-2 levels were observed in both UTX-deficient NK cells and 307 UTX-deficient CD8 + T cells, suggesting that UTX down-regulates this anti-apoptotic factor in 308 both innate and adaptive cytotoxic lymphocytes. 

Weighing factors that define patient subsets with different immune responses will allow 354 us to move past a "one-size-fits-all" therapeutic approach to a precision medicine paradigm. 

ATAC-seq fastq files were trimmed to remove low-quality reads and adapters using 672 Cutadapt 61 (version 2.3). The reads were aligned to the reference mouse genome (mm10) with 673 bowtie2 62 (version 2.2.9). Peak calling was performed with MACS2 63 (version 2.1.1). The peaks 674 from all samples were merged into a single bed file, peaks from different samples that were 675 closer than 10bp were merged into a single peak. HTseq 59 (version 0.9.1) was used to count the 676 number of reads that overlap each peak per sample. The peak counts were analyzed with 677 DESeq2 60 (version 1.24.0) to identify differentially accessible genomic regions. Peaks with 678 adjusted p-value < 0.05 were considered significantly differentially accessible. The peak counts 679 were visualized with Integrated Genome Browser, (version 9.1.8).

Fuzzy c-means clustering was used for both ATAC-seq and RNA-seq using significant 681 (p-value and FDR <0.05, Log2FC +/-0.5) normalized counts generated from DESeq2. MFuzz 682 package (version 3.14) within R was used to perform this analysis into 6 clusters with a 683 membership score of >0.5. The differentially accessible ATAC peaks were analyzed using the 

Sex Differences in Immunity

Sex differences in immune responses

Exploring the biological contributions to human health: 708 does sex matter?

Sex Differences in CMV 711

Replication and HIV Persistence During Suppressive ART

Sex differences in immune responses

Gender effect on in vitro lymphocyte subset levels of healthy individuals

Age-and sex-related changes in lymphocyte subpopulations of 720 healthy Asian subjects: from birth to adulthood

Reference 723 intervals for leukocyte subsets in adults: Results from a population-based study using 724 10-color flow cytometry

Natural variation in the parameters of innate 731 immune cells is preferentially driven by genetic factors

Effects of sex and aging on the immune cell landscape as assessed by 735 single-cell transcriptomic analysis

X Inactivation and Escape: Epigenetic and 738

Sex difference in neural tube defects in p53-null mice is caused by differences in 741 the complement of X not Y genes

A role for sex chromosome 745 complement in the female bias in autoimmune disease

Natural killer cell specificity for viral 748 infections

Natural killer cell deficiency

Severe herpesvirus infections in an adolescent 752 without natural killer cells

Cmv-1, a 755 genetic locus that controls murine cytomegalovirus replication in the spleen

Adoptive transfer studies 758 demonstrating the antiviral effect of natural killer cells in vivo

Natural killer (NK) 761 cell response to virus infections in mice with severe combined immunodeficiency. The 762 stimulation of NK cells and the NK cell-dependent control of virus infections occur 763 independently of T and B cell function

Vital involvement 766 of a natural killer cell activation receptor in resistance to viral infection

Murine cytomegalovirus is regulated by a discrete subset of natural killer cells reactive 770 with monoclonal antibody to Ly49H

Natural killer cell 772 depletion enhances virus synthesis and virus-induced hepatitis in vivo

Genetic influences on the augmentation 775 of natural killer (NK) cells during murine cytomegalovirus infection: correlation with 776 patterns of resistance

Demonstration of the antiviral role of natural killer cells in vivo with a natural killer cell-779 specific monoclonal antibody (NK 1.1)

Increased 781 susceptibility to cytomegalovirus infection in beige mutant mice

Sex-and age-dependent alterations of splenic immune cell profile and NK cell 785 phenotypes and function in C57BL/6J mice

Natural Killer Cells: From Innate to Adaptive 788

Synergistic proliferation and activation 791 of natural killer cells by interleukin 12 and interleukin 18

Effect of estrogens on the interferon-gamma 794 producing cell population of mouse splenocytes

Escape from X inactivation varies in mouse tissues

Four Core Genotypes and XY* mouse models: Update on impact on SABV 800 research

Enhancer Remodeling and Transcription Factor Repurposing Enable High Magnitude 805

Developmental Acquisition of Regulomes 809 Underlies Innate Lymphoid Cell Functionality

Jmjd3 and UTX play a demethylase-812 independent role in chromatin remodeling to regulate T-box family member-dependent 813 gene expression

Fuzzy C-means method for clustering microarray data

Profiler: a web server for functional enrichment analysis and conversions of gene lists 818 (2019 update)

Simple combinations of lineage-determining transcription 821 factors prime cis-regulatory elements required for macrophage and B cell identities

Core-binding factor beta and Runx transcription factors promote adaptive 825 natural killer cell responses

KLF/SP Transcription Factor Family 830 Evolution: Expansion, Diversification, and Innovation in Eukaryotes

Cell cycle progression dictates the requirement for BCL2 in natural 836 killer cell survival

NK cell-derived GM-CSF 840 potentiates inflammatory arthritis and is negatively regulated by CIS

Early IFN-alpha signatures and persistent dysfunction are distinguishing

Follicular Helper Cell-Dependent Clearance of a Persistent Virus Infection Requires T 856

The histone demethylase UTX regulates the 861 lineage-specific epigenetic program of invariant natural killer T cells

UTX promotes CD8(+) T cell-mediated antiviral defenses but reduces T 865 cell durability

Role of interferon-867 gamma in murine cytomegalovirus infection

Pleiotropic Functions of H3K27Me3 Demethylases in Immune

Transcriptional Regulatory Network Coordinately Shapes Active Enhancer Landscapes 874 for Eliciting Transcription

Oxidative stress controls 878 regulatory T cell apoptosis and suppressor activity and PD-L1-blockade resistance in 879 tumor

Inhibition of histone H3K27 demethylases selectively modulates 883 inflammatory phenotypes of natural killer cells

A selective jumonji 890 H3K27 demethylase inhibitor modulates the proinflammatory macrophage response

Differential effects of Cytomegalovirus carriage on 894 the immune phenotype of middle-aged males and females

Kabuki syndrome genes 898 KMT2D and KDM6A: functional analyses demonstrate critical roles in craniofacial, heart 899 and brain development

Adoptive cell therapy using engineered natural killer cells

Phenotypic and Functional Plasticity of CXCR6(+) Peripheral Blood NK Cells

ILC1 Confer Early Host 908 Protection at Initial Sites of Viral Infection

The transcription factor Fli1 restricts the formation of memory precursor NK cells 912 during viral infection

HTSeq--a Python framework to work with high-914 throughput sequencing data

Moderated estimation of fold change and dispersion 917 for RNA-seq data with DESeq2

A New Tool for Accurate 920 Cutting of Primers from Reads of Targeted Next Generation Sequencing

Fast gapped-read alignment with Bowtie 2