Abstract

   Sexual dimorphism is a fundamental characteristic of various
   physiological and pathological processes in humans, including immune
   responses, senescence, and metabolism. Most studies on the sex bias
   have focused on sex hormones or female‐biased genes, whereas
   male‐biased genetic factors remain understudied. Here, we show that the
   Y‐linked noncoding RNA, RP11‐424G14.1, is expressed in human male
   keratinocytes. Microarray study suggests the NF‐κB pathway as the top
   biological pathway affected by RP11‐424G14.1 knockdown, consistent with
   known sex differences in inflammation. Additionally, IGFBP3 is
   identified as the top gene supported by RP11‐424G14.1 in male
   keratinocytes. Conversely, in female keratinocytes, IGFBP3 is the top
   gene repressed by the X‐linked long noncoding RNA XIST, suggesting a
   central role of IGFBP3 in mediating sexual dimorphism. Knockdown of
   RP11‐424G14.1 or IGFBP3 in male keratinocytes inhibits cellular
   senescence, consistent with increased longevity in females. IGFBP3
   expression is dependent on insulin, and metabolomics analysis suggests
   that RP11‐424G14.1 and IGFBP3 regulate acrylcarnitine metabolism. Our
   study identifies the role of the RP11‐424G14.1‐IGFBP3 pathway in
   coordinating sex differences in immunity, senescence, and metabolism.
   With RP11‐424G14.1 being a human‐specific genetic element, our study
   suggests the evolving feature of sexual dimorphisms in biological
   processes.

   Keywords: inflammation, metabolism, noncoding RNA, senescence, sexual
   dimorphism
     __________________________________________________________________

   Sexual dimorphism is a fundamental characteristic of various
   physiological and pathological processes in humans. Here we identify a
   human‐specific, Y‐linked noncoding RNA, RP11‐424G14.1, in male
   keratinocytes. Transcriptomic analysis shows that RP11‐424G14.1
   regulates IGFBP3 and the NF‐κB pathway, affecting male‐biased cellular
   senescence and immunosuppression. Metabolomic study identifies
   acrylcarnitine metabolism as a specific pathway downstream of
   RP11‐424G14.1‐IGFBP3, suggesting coordinated regulation of cellular
   aging, immune activity, and lipid metabolism.

   graphic file with name FBA2-7-e1479-g007.jpg

1. INTRODUCTION

   Sexual dimorphism is a fundamental characteristic of various
   physiological and pathological processes in humans. Generally, it is
   accepted that females mount stronger innate and adaptive immune
   responses than males, which is supported by the clinical observation of
   strikingly increased prevalence of autoimmune diseases in females.[26]
   ^1 , [27]^2 , [28]^3 In human populations, women are known to live
   longer than men, consistent with lower biological ages as assessed by
   molecular markers.[29] ^4 , [30]^5 Furthermore, sex differences have
   been identified in many aspects of energy balance regulation including
   glucose metabolism and insulin sensitivity as well as lipid storage and
   utilization patterns, which may influence the pathogenesis of diseases
   including obesity and diabetes.[31] ^6 , [32]^7

   Studies on the molecular mechanisms underlying sex differences in human
   physiology have been focused on sex hormones.[33] ^8 , [34]^9 , [35]^10
   Additionally, gene‐dosage effects of the X‐ and Y‐chromosomes have been
   described and on the single‐gene level, sex‐biased, protein‐coding
   genes have been identified with roles in regulating sexually dimorphic
   processes.[36] ^11 , [37]^12 , [38]^13 , [39]^14 The map of sexually
   dimorphic genetic elements is just starting to unfold, as
   protein‐coding regions only make up approximately 1 percent of the
   human genome and the rest 99 percent is noncoding.[40] ^15 The active
   transcription of non‐protein coding genes gives rise mainly to long
   non‐coding RNA (lncRNAs, 80%–90%).[41] ^16

   Long non‐coding RNA (lncRNA) are typically transcripts longer than
   200 nt that do not code for proteins but may still serve a diverse
   array of functions.[42] ^17 They may regulate expression epigenetically
   by coordinating with DNA or histone‐modifying enzymes or their products
   and can also directly affect transcription by interacting with or
   acting as transcription factors. Post‐transcriptionally, lncRNA can
   inhibit or coordinate with small interfering (siRNA) or microRNA
   (miRNA).[43] ^18 Though not as common, they are also capable of
   post‐translational modification.[44] ^19

   The importance of lncRNA in sex‐biased expression has only started to
   gain traction in the last decade. One of the most studied in this
   regard is X‐inactive specific transcript, XIST, known for its role in
   female development through X‐chromosome inactivation (XCI).[45] ^20
   During XCI, XIST RNA triggers gene silencing and drives a major
   structural reorganization to shut down one copy of the X‐chromosome,
   which encodes for various immune‐associated genes, in females. XIST
   overexpression has been associated with systemic autoimmune disease
   including systemic lupus erythematosus (SLE) and rheumatoid arthritis,
   both of which are female‐biased.[46] ^21 Beyond immune regulation,
   lncRNA is implicated in both glucose and lipid metabolism, with the
   focus being XIST as well.[47] ^22 To the best of our knowledge, little
   is known about the function of other lncRNAs in mediating sexual
   dimorphism.

   To this end, in this study we have address the expression and function
   of RP11‐424G14.1, a lncRNA of 2666 bp length encoded by the Y
   chromosome. Intriguingly, it is also a human‐specific genetic elements
   and does not show homology in alignment with the rhesus monkey, mouse,
   or other sequenced vertebrates. We confirm the male‐specific expression
   of RP11‐424G14.1 in human keratinocytes and identify its multifaceted
   function in mediating sex differences in inflammation, senescence, and
   metabolism.

2. MATERIALS AND METHODS

2.1. Keratinocyte culture and transfection

   ATCC primary epidermal keratinocyte cultures were grown in Keratinocyte
   Growth Medium, or KGM, (Lifeline Cell Technology) on labeled 10 cm cell
   culture‐treated dishes. KGM was refreshed after seeding and every other
   day after during culture. Upon sufficient cell growth and 70%–90% cell
   confluency, keratinocytes were collected via trypsinization, counted by
   hemocytometer, and transfected on the Lonza 4‐D X unit nucleofector
   according to P3 Lonza keratinocyte nucleofection protocols (V4XP‐3024).
   If the cells were treated with immunostimulants, they (and their
   controls) were given 24 h post‐nucleofection to recover. 24 h
   post‐treatment, cells were either harvested for RNA or protein, or were
   collected and used for experiments. siRNA used was purchased through
   Thermofisher: Silencer Select negative control siRNA (referred to as
   scrambled siRNA), pre‐designed siRNAs for XIST (assay ID: n272634) or
   IGFBP3 (assay ID: s7228), or custom designed for RP11‐424G14.1
   (GGGAAGUGUUUGUACAGUATT and UACUGUACAAACACUUCCCGA).

2.2. qPCR

   RNA isolated from keratinocyte culture either by Trizol extraction or
   Miniprep was quantified and used to make cDNA with reverse
   transcriptase. Standardized amounts of cDNA template were pipetted into
   96‐well plates with respective primers, Power SYBR Green master mix,
   and nuclease‐free water. Optical adhesive film was used to cover the
   plates, then centrifuged and run on a QuantStudio 6 at standard speed
   for 40 cycles. Differential gene expression was calculated using the
   ΔΔct method.

2.3. Microarray

   mRNA was isolated from cultured male and female keratinocytes using
   Trizol following manufacturer instructions. mRNA was hybridized to the
   Thermofisher human Clariom S array. Raw signals were obtained by the
   GeneChip Scanner and analyzed by the Transcriptome Analysis Console
   Software according to standard procedures. Differentially expressed
   genes were identified using a false discovery rate of 0.05 and a fold
   change of 2.

2.4. Metabolomics

   Keratinocyte lysates were profiled using Biocrates MxP Quant500 plates
   with LC‐ and FIA‐MS following manufacturer instructures. Metabolites
   were identified, quantified, and normalized by the MetIDQ workflow
   following standard analysis workflows. Differentially expressed
   metabolites were identified using an adjusted p‐value of 0.05.

2.5. Immunofluorescence staining and imaging

   Keratinocytes were transferred to 8‐well chamber slides after
   transfection at a seeding density of 70,000–100,000 per well. After
   control or target gene knockdown such as XIST, RP11‐424G14.1, or IGFBP3
   knockdown (oligo info specified in the ‘keratinocyte culture and
   transfection’ section), cells were washed with PBS, fixed with 4%
   paraformaldehyde (PFA) for 15 min, and permeabilized with 0.1% Triton
   X‐100 for 5 min. Cells were then washed in wash buffer (PBS with 0.1%
   Triton X‐100) for 5 times and stained with the respective primary
   antibodies for 1 h at room temperature at manufacturer‐recommended
   dilutions (IGFBP3, Thermofisher, 10,189‐2‐AP; NFkB, Cell Signaling
   Technologies, 3033; pIkBa, Cell Signaling Technologies, 2859; p21, Cell
   Signaling Technologies, 2947; THBS2, Thermofisher, MA5‐46953; TSPAN1,
   Thermofisher, PA5‐119458; WNK1, Abcam, ab174854), followed by 5 washes.
   Cells were subsequently stained the respective secondary antibodies at
   manufacturer‐recommended dilutions (goat anti‐rabbit IgG (H + L) Alexa
   Fluor 488, Thermofisher, A11008; donkey anti‐mouse IgG (H + L) Alexa
   Fluor 488, Thermofisher, A21202) for 1 h. Cells were then washed with
   milliQ water, stained with 1 μg/mL DAPI for 5 min, and washed again.
   Upon completion of staining, the cells were then covered with
   Vectashield antifade mounting medium and imaged on a Nikon TI‐2 E
   inverted fluorescent microscope. The same acquisition parameters,
   including laser intensity and exposure settings, were used for all
   conditions in comparison.

2.6. Western blot

   Keratinocytes were transferred to 6‐well tissue culture‐treated plates
   after transfection. After treatments, cells were washed with ice‐cold
   PBS and placed on ice. Cells were then treated with RIPA buffer and
   scraped to isolate protein. Cell‐lysate protein concentration was
   measured for each sample using a bicinchoninic acid assay with bovine
   serum albumin for a standard on an Agilent BioTek Epoch plate reader
   and Take3 plate. For each sample, 35 μL were loaded into wells on a
   precast mini‐PROTEAN gel, containing 20 μg protein, 8.75 μL 4× Laemmli
   buffer, and milliQ water. 4 μL of BioRad Protein Precision Plus
   standard ladder were loaded into surrounding wells. Gels were run in a
   BioRad mini‐PROTEAN tetra cell with Tris‐glycine‐SDS running buffer at
   100 V for 2 h. Upon separation, gels were then equilibrated and
   transferred to an activated PVDF membrane for blotting, according to
   the BioRad transfer manual. Blots were then cut based on
   ladder‐indicated size ranges to capture protein targets for efficient
   antibody probing. Blots were rinsed in Tris buffered saline with 0.1%
   Tween‐20 (TBS‐T) and then blocked in 5% non‐fat dry milk in TBS‐T for
   1 hour at RT. Primary and HRP‐linked secondary antibodies were diluted
   to indicated concentrations. Blots were incubated with indicated
   primary antibodies for 1 h at RT and then washed 3x in TBS‐T. Blots
   were then stained with respective secondary antibodies for 1 h at RT
   and then washed 5x in TBS‐T and 1x in TBS. Stained blots were marked
   with WesternSure pen to identify marker lanes upon imaging. Stained
   blots were incubated with LI‐COR WesternSure chemiluminescent substrate
   for 5 min and placed in a laminated pouch, aspirating any excess
   solution. Blots were imaged on a C‐Digit, after 12‐min exposure.

3. RESULTS

3.1. Sex‐biased expression of lncRNAs in human keratinocytes

   To understand the potential roles of lncRNAs in regulating sexual
   dimorphism in human male keratinocytes, we used qPCR to examine the
   expression levels of top male‐biased lncRNAs suggested by previous
   RNA‐sequencing studies using whole skin biopsies.[48] ^14 Of the three
   candidates tested, including RP11‐372M18.2, RP11‐424G14.1, and
   RP11‐331F4.1, we were able to detect male‐specific expression of
   RP11‐424G14.1 in keratinocytes (Figure [49]1A). It is possible that the
   other two candidates, RP11‐372M18.2 and RP11‐331F4.1, are expressed in
   other types of skin cells such as fibroblasts instead. Consistent with
   its male‐specific expression, RP11‐424G14.1 maps to chromosome Y:
   21,853,827–21, 856, 492 with no overlap to known protein‐coding genes.
   The expression level of RP11‐424G14.1, but not the control gene GAPDH,
   was significantly reduced by knockdown using its sequence‐specific
   siRNA, suggesting specificity of expression detection (Figure [50]1B,
   Figure [51]S1). We did not detect a significant association between
   RP11‐424G14.1 and the age of donors (Figure [52]1A).

FIGURE 1.

   FIGURE 1
   [53]Open in a new tab

   Sex‐biased expression of RP11‐424G14.1 and XIST as well as their
   downstream targets. (A) Expression levels RP11‐424G14.1 and XIST in
   female (F) and male (M) keratinocytes (left) by qPCR, showing lack of
   association with age (right). (B) Specificity of RP11‐424G14.1
   detection and knockdown, shown by expression levels of RP11‐424G14.1
   and GAPDH upon scrambled RNAi (Scr Ri, as control) or RP11‐424G14.1
   sequence‐specific RNAi by qPCR. (C) Venn diagram of genes supported or
   repressed by RP11‐424G14.1 knockdown in male keratinocytes and XIST
   knockdown in female keratinocytes compared to respective controls from
   microarray studies (FDR <0.05 for differentially expressed genes). (D)
   Expression levels of indicated genes upon scrambled (control) or
   RP11‐424G14.1 knockdown (KD) in male keratinocytes (top) and upon
   scrambled (control) or XIST knockdown (KD) in female keratinocytes
   (bottom) by qPCR. (E) Pathway enrichment analysis of microarray data
   showing the top 5 pathways impacted by the loss of RP11‐424G14.1 or
   XIST. FDR, false discovery rate. Mean ± stdev, *p < 0.05, Student's
   t‐test.

   As expected, as a lncRNA known to mediate X‐chromosome inactivation in
   females, XIST RNA is expressed in human keratinocytes in a female‐based
   manner (Figure [54]1A). The expression level of XIST, but not the
   control gene GAPDH, was reduced by its sequence‐specific knockdown,
   suggesting specificity of expression detection (Figure [55]S1). While
   XIST has been shown to decrease in expression in aged hematopoietic
   stem cells (HSCs),[56] ^23 we have found that in keratinocytes, the
   expression level of XIST did not associate with the age of donors
   (Figure [57]1A). Taken together, these data suggest tissue specificity
   in the expression change of XIST during aging. Based on these results,
   we focused our subsequent study on RP11‐424G14.1 in male keratinocytes
   and XIST in female keratinocytes.

3.2. Genes and pathways regulated by sex‐biased lncRNAs

   Because little is known about the roles of RP11‐424G14.1 and XIST on a
   genome‐wide level in human keratinocytes, we performed microarray to
   identify their downstream genes in an unbiased manner. We performed
   RP11‐424G14.1 and scrambled control knockdown in human male
   keratinocytes as well as XIST and scrambled control knockdown in human
   female keratinocytes (age‐ and race‐matched), followed by microarray
   analyses. We identified four groups of differentially expressed genes
   (FDR <0.05), namely genes that were downregulated by RP11‐424G14.1
   knockdown in male keratinocytes (i.e., RP11‐424G14.1‐supported), genes
   that were upregulated by XIST knockdown in female keratinocytes (i.e.,
   XIST‐repressed), genes that were upregulated by RP11‐424G14.1 knockdown
   in male keratinocytes (i.e., RP11‐424G14.1‐repressed), and genes that
   were downregulated by XIST knockdown in female keratinocytes (i.e.,
   XIST‐supported). Compared to XIST, RP11‐424G14.1 regulated a restricted
   set of genes, supporting the expression of 20 genes and repressing the
   expression of 4 genes (Figure [58]1C). Of the 20
   RP11‐424G14.1‐supported genes in male keratinocytes, 4 genes (20%) were
   repressed by XIST in female keratinocytes, suggesting that these four
   genes are male‐biased genes that are coregulated by the two lncRNAs
   (i.e. activated by RP11‐424G14.1 in males and repressed by XIST in
   females). These four genes are IGFBP3, CXCL14, SEMA5A, and WNK1. Of the
   4 RP11‐424G14.1‐repressed genes in male keratinocytes, 2 genes (50%)
   were supported by XIST in female keratinocytes, suggesting that these
   two genes are female‐biased genes that are coregulated by the two
   lncRNAs (i.e., repressed by RP11‐424G14.1 in males and activated by
   XIST in females). These two genes are TSPAN1 and THBS2. We further
   validated the co‐regulation of these six genes by RP11‐424G14.1 and
   XIST by independent qPCR experiments (Figure [59]1D) and summarized
   their known functions (Table [60]1).

TABLE 1.

   Known functions for co‐targets of RP11‐424G14.1 and XIST as identified
   by microarray.
   Gene         Known roles Expression
   IGFBP3
     * Negative regulator of IGF signaling.[61] ^39
     * Antagonist of TGF‐βII receptor, arresting growth[62] ^34 , [63]^38
       , [64]^40
     * Facilitate NHEJ.[65] ^41
     * Inhibit keratinocyte proliferation.[66] ^42
     * Modulate carbohydrate and lipid metabolism.[67] ^43

   Male‐biased
   CXCL14
     * Important for differentiation of dendritic cells into Langerhans
       cells.[68] ^44
     * Immune surveillance through chemotaxis.[69] ^45
     * Neutrophil, DC, and uterine NK cell chemoattractant.[70] ^46

   Male‐biased
   SEMA5A
     * Inhibits synaptogenesis during neurogenesis.[71] ^47
     * Promotes angiogenesis and inhibits apoptosis.[72] ^48

   Male‐biased
   WNK1
     * Ion homeostasis in kidney and nervous system.[73] ^49
     * Autophagy inhibition.[74] ^50
     * Suppresses NLRP3 inflammasome activation, and therefore
       inflammation and pyroptosis.[75] ^51

   Male‐biased
   TSPAN1
     * Promotes cell migration.[76] ^52
     * Suppresses NFκB signaling.[77] ^53

   Female‐biased
   THBS2
     * Inhibits angiogenesis.[78] ^54
     * Promotes cell motility and migration.[79] ^55

   Female‐biased
   [80]Open in a new tab

   Furthermore, gene ontology analyses suggested that the top pathways
   affected by RP11‐424G14.1 knockdown included the NF‐κB pathway,
   consistent with known sex differences in inflammation. Intriguingly,
   both RP11‐424G14.1 and XIST knockdown significantly affected
   senescence‐related pathways, in line with sexual dimorphism in aging
   (Figure [81]1E).

   With available antibodies, we were able to confirm the regulation of
   IGFBP3 (Insulin‐like growth factor binding protein 3), among other
   potential leads, by RP11‐424G14.1 (Figure [82]2A,B, Figure [83]S2).
   Because IGFBP3 was the top candidate from the microarray experiment, we
   focused subsequent study on IGFBP3 (insulin‐like growth factor binding
   protein) on its potential role in cellular senescence, the co‐regulated
   biological pathway.

FIGURE 2.

   FIGURE 2
   [84]Open in a new tab

   Regulation of senescence by male‐biased molecules RP11‐424G14.1 and
   IGFBP3. (A) Expression level of IGFBP3 in female or male keratinocytes
   by qPCR, with RNAi‐mediated knockdown (Ri) of indicated genes. Scr,
   scrambled control. (B, C) Immunofluorescence staining (B) and
   quantification (C) of IGFBP3 with DAPI counterstain in male (M) or
   female (F) keratinocytes, with indicated knockdown or scrambled
   control. (D, E) Immunofluorescence staining (D) and quantification (E)
   of p21 with DAPI counterstain in male (M) or female (F) keratinocytes,
   with indicated knockdown or scrambled control. (F) Quantification of
   BrdU assay in male or female keratinocytes, with indicated knockdown or
   scrambled control. Average fluorescence intensity per cell is used in
   quantification. Scale bar, 10 μm. Mean ± stdev, *p < 0.05, Student's
   t‐test.

3.3. IGFBP3 as a lncRNA‐regulated male‐biased gene as well as their
modulation of senescence

   As an RP11‐424G14.1‐supported and XIST‐repressed gene from microarray
   results, we predicted that IGFBP3 was a male‐biased gene. Indeed, in
   independent experiments we confirmed the male‐biased expression of
   IGFBP3 in human keratinocytes on the mRNA level and in subcellular
   compartments (Figure [85]2A,B). Additionally, we confirmed that in male
   keratinocytes, RP11‐424G14.1 knockdown reduced IGFBP3 observed within
   the cell (Figure [86]2A–C). On the contrary, in female keratinocytes,
   XIST knockdown increased IGFBP 3 observed within the cell
   (Figure [87]2A–C).

   Using p21 as a marker for cell senescence,[88] ^24 we showed that age‐
   and race‐matched male keratinocytes expressed increased levels of p21
   compared to female (Figure [89]2D,E), consistent with previous reports
   that females exhibit lower biological ages as assessed by molecular
   markers.[90] ^4 , [91]^5 Knockdown of RP11‐424G14.1 or IGFBP3 reduced
   p21 levels in male keratinocytes, while knockdown of XIST increased p21
   levels in female keratinocytes (Figure [92]2D,E). Similar results were
   obtained using the bromodeoxyuridine (BrdU) assay,[93] ^25 which showed
   limited proliferative capability of male keratinocytes compared to
   females (Figure [94]2F, Figure [95]S3). The knockdown of RP11‐424G14.1
   or IGFBP3 restored proliferation of male keratinocytes (Figure [96]2F).
   Our results suggest that the RP11‐424G14.1 – IGFBP3 axis promotes
   senescence in male keratinocytes.

3.4. Regulation of IGFBP3 expression by insulin and nutrient levels

   IGFBP3 has been historically used as a marker for growth hormone levels
   and is known to associate with high body mass index (BMI).[97] ^26 ,
   [98]^27 In addition to its role in binding IGF (insulin‐like growth
   factor), IGFBP3 can signal through the IGF receptor, which can also
   relay signal from insulin, to regulate cell growth and survival.[99]
   ^27 Based on the known link of IGFBP3 to insulin signaling, we tested
   the hypothesis that IGFBP3 itself is regulated by insulin, a component
   of the keratinocyte growth media. Indeed, removal of insulin from the
   keratinocyte growth media reduced IGFBP3 levels, and similar effects
   were achieved by removing all growth supplements from the media
   (Figure [100]S4). Removal of insulin or nutrients and the reduction in
   IGFBP3 were concurrent with the decrease in p21 (Figure [101]S4). The
   decrease in p21 and cell number with insulin treatment is consistent
   with previous reports that increased insulin levels promote senescence
   of various human cell types and that nutrient restriction reduces
   senescence.[102] ^28 , [103]^29 Insulin is known to signal through AKT
   signaling,[104] ^30 and in male keratinocytes we have found that AKT
   phosphorylation was dependent on RP11‐424G14.1 and IGFBP3, which
   suggests that IGFBP3 can both respond to and regulate insulin signaling
   (Figure [105]S4).

   Our ancestors have been dubbed “fat hunters”, and to accommodate the
   large brains, the need for fat is thought to propel the evolution of H.
   erectus.[106] ^31 Given the upregulation of IGFBP3 by insulin and
   nutrient levels, we were curious to examine its expression regulation
   in other organisms. Intriguingly, we found that RP11‐424G14.1, the
   promoter of IGFBP3 expression in human keratinocytes, was
   human‐specific and not found in the rhesus monkey, mouse, or other
   lower organisms (Figure [107]S5A). In the mouse skin, without
   RP11‐424G14.1, there was no longer a significant male bias in Igfbp3
   expression (Figure [108]S5B). Therefore, the male‐biased RP11‐424G14.1
   – IGFBP3 axis is human‐specific.

3.5. Metabolism regulation by RP11‐424G14.1 and IGFBP3

   Given the observed link between IGFBP3 and nutrition, we directly
   tested whether RP11‐424G14.1 and IGFBP3 regulated cell metabolism by
   metabolomic studies. Over 600 metabolites from 26 biochemical classes
   were profiled and quantified in keratinocytes with control or knockdown
   of RP11‐424G14.1 or IGFBP3. Consistent with a putative role of
   RP11‐424G14.1 – IGFBP3 in lipid metabolism, the most significant effect
   upon knockdown of RP11‐424G14.1 or IGFBP3 was observed in the
   acylcarnitine group. RP11‐424G14.1 or IGFBP3 knockdown resulted in
   increases in various long‐chain acylcarnitines such as
   octadecenoylcarnitine, hydroxyhexadecenoylcarnitine,
   hydroxyhexadecanoylcarnitine, hydroxytetradecadienoylcarnitine, and
   tetradecanoylcarnitine, suggesting impairment of beta‐oxidation of
   fatty acids (Figure [109]3A). In addition, RP11‐424G14.1 or IGFBP3
   knockdown led to decreases in medium‐chain acycarnitines including
   dodecenoylcarnitine, dodecanedioylcarnitine, decenoylcarnitine, and
   nonaylcarnitine (Figure [110]3A). Meanwhile, limited changes were
   detected for other lipid groups including diglycerides, triglycerides,
   phosphatidylcholines, lysophosphatidylcholines, or free or
   non‐covalently bound fatty acids (Figures [111]4 and [112]5). Small
   molecules including nucleobases, amino acid, vitamins, and other
   metabolites remain unchanged upon the knockdowns (Figure [113]3C,
   Figure [114]5). The pattern of metabolite changes with knockdown in
   RP11‐424G14.1 and IGFBP3 is consistent with them working in the same
   pathway. Therefore, the male‐biased factors RP11‐424G14.1 and IGFBP3
   regulate lipid metabolism in keratinocytes.

FIGURE 3.

   FIGURE 3
   [115]Open in a new tab

   Regulation of metabolism by male‐biased molecules RP11‐424G14.1 and
   IGFBP3. Metabolic profiling and quantification of acylcarnitine (A),
   amino acid (B) and nucleobases and vitamin and cofactors (C) in male
   keratinocytes with scrambled RNAi (scr Ri), RP11‐424G14.1‐specific RNAi
   (RP11‐424G14.1 Ri), and IGFBP3‐specific RNAi (IGFBP3 Ri). *, FDR <0.05.

FIGURE 4.

   FIGURE 4
   [116]Open in a new tab

   Regulation of diglycerides and triglycerides by male‐biased molecules
   RP11‐424G14.1 and IGFBP3. Metabolic profiling and quantification of
   diglycerides (A) and triglycerides (B) in male keratinocytes with
   scrambled RNAi (scr Ri), RP11‐424G14.1‐specific RNAi (RP11‐424G14.1
   Ri), and IGFBP3‐specific RNAi (IGFBP3 Ri). *, FDR <0.05.

FIGURE 5.

   FIGURE 5
   [117]Open in a new tab

   Regulation of phosphoglycerolipids and additional metabolites by
   male‐biased molecules RP11‐424G14.1 and IGFBP3. Metabolic profiling and
   quantification of phosphoglycerolipids (A) and free fatty acids (B) in
   male keratinocytes with scrambled RNAi (scr Ri), RP11‐424G14.1‐specific
   RNAi (RP11‐424G14.1 Ri), and IGFBP3‐specific RNAi (IGFBP3 Ri). *,
   FDR <0.05.

3.6. Regulation of NF‐κB pathway by RP11‐424G14.1

   Enrichment of the NFκB pathway upon RP11‐424G14.1 knockdown from the
   microarray study prompted us to investigate the role of RP11‐424G14.1
   in regulating inflammation, a female‐biased response. During
   inflammation, activated NFκB translocates from cytosol to nucleus to
   coordinate with cofactors and regulate expression of inflammatory
   genes.[118] ^32 We simulated inflammation using TNFα, LPS, and IL‐1β
   and found that the combination of RP11‐424G14.1 knockdown and TNFα
   stimulation promoted NFκB nuclear translocation and activation
   (Figure [119]6, Figure [120]S6). Mechanistically, RP11‐424G14.1
   inhibition of NFκB activation does not appear to function through
   canonical means, as TNFα induced IκBα degradation regardless of RP11
   knockdown (Figure [121]6, Figures [122]S6 and [123]S7). Therefore, the
   lack of RP11‐424G14.1 in female keratinocytes potentiates TNFα‐induced
   NFκB activation.

FIGURE 6.

   FIGURE 6
   [124]Open in a new tab

   Regulation of NFκB nuclear translocation by RP11‐424G14.1. (A)
   Immunofluorescence staining of NFκB and pIκBα in male keratinocytes
   with RP11‐424G14.1 knockdown (RP11‐424G14.1 Ri) or scrambled knockdown
   (Scr Ri), with or without stimulation with TNFa, LPS, or IL1B. Scale
   bar, 10 μm. (B) Keratinocytes stained with DAPI to delineate their
   nuclei were analyzed for nuclear NFkB localization. Regions of interest
   (ROI) were drawn around nuclei using DAPI counterstain, and then
   anti‐NFkB‐GFP probe mean fluorescence intensity (MFI) within the ROI
   was measured with standardized laser intensity and exposure settings
   between conditions. Whisker‐plot of nuclear NFκB signal calculated with
   mean fluorescence intensity (MFI) of n ≥ 199. Kruskal–Wallis test was
   performed due to half‐normal distribution and differences in group size
   (α = 0.05). **, P < 0.01. ****, P < 0.0001.

4. DISCUSSION

   With protein‐coding genes making up only approximately 1 percent of the
   human genome, it is important to understand how the non‐coding genetic
   elements may contribute to sexual dimorphism. Our study uncovers the
   multifaceted roles of a Y‐linked noncoding RNA, RP11‐424G14.1, in
   mediating sex differences in inflammation, senescence, and metabolism.
   We have found that in male keratinocytes, RP11‐424G14.1 signals through
   IGFBP3, which is regulated by XIST in an opposing manner in female
   cells. Through IGFBP3, we show that RP11‐424G14.1 and XIST modulate
   p21‐mediated senescence in a manner consistent with the increased
   longevity of females.

   The roles of the Y‐linked RP11‐424G14.1 and X‐linked XIST in regulation
   of inflammation, senescence, and metabolism suggest the contribution of
   sex chromosomes to these processes. While we have so far obtained no
   direct evidence for the regulation of RP11‐424G14.1 and XIST by sex
   hormones in cultured keratinocytes, future studies are warranted to
   test for a potential genetic and hormonal interaction. Indeed, studies
   show that XIST promotes insulin resistance in gestational
   diabetes,[125] ^33 suggesting a regulation of noncoding RNAs by dynamic
   changes in sex hormone levels.

   RP11‐424G14.1 is a human‐specific lncRNA. Consistently, we have found
   that RP11‐424G14.1‐supported male bias of IGFBP3 is not conserved in
   the mouse. Intriguingly, the evolution of humans is tightly linked to
   the need to feed on and metabolize fat.[126] ^31 Our metabolomic data
   suggests that RP11‐424G14.1 and IGFBP3 work in a coordinated manner to
   regulate fatty acid metabolism, with a decrease in mobilized
   medium‐chain acylcarnitines and increase in mobilized long‐chain
   acylcarnitines with their knockdown. The lack of change in diglycerides
   or triglycerides implies that the cells are not simply meeting the
   demands of increased proliferation on their membranes. Furthermore,
   RP11‐424G14.1 and IGFBP3 may regulate lipid metabolism through
   mitochondria homeostasis, as has been reported for IGFBP3.[127] ^34
   Further study on the mechanism by which RP11‐424G14.1 and IGFBP3
   regulate various classes of lipids is necessary to address these
   possibilities.

   RP11‐424G14.1 also has a role in protecting male keratinocytes from
   TNFα‐induced inflammation. Knockdown of RP11‐424G14.1, mimicking the
   female scenario, promotes the nuclear translocation of NFκB. Given the
   degradation of IκBα, we speculate that RP11‐424G14.1 functions on other
   regulatory components for NFκB subunits such as HDAC3 and IRFD
   themselves or the importins/exportins and modulates the nuclear
   translocation process.[128] ^35 , [129]^36 , [130]^37 lncRNAs are
   capable themselves of modifying proteins through ubiquitination or
   adenylation.[131] ^19 Otherwise, RP11‐424G14.1 is located on the Y
   chromosome upstream of KDM5D, a histone demethylase known for sexually
   dimorphic epigenetic regulation and interaction with HDACs.[132] ^38

   In conclusion, XIST and RP11‐424G14.1 are sex‐chromosome‐encoded
   lncRNAs that modulate sexual dimorphic pathways in females and males,
   respectively. While XIST has previously been described as an important
   factor in female bias towards autoimmunity, we have found that
   RP11‐424G14.1 plays an opposing role in males. In male keratinocytes,
   RP11‐424G14.1 promotes senescence, modulate fatty acid metabolism, and
   prevent TNFα‐induced NFκB activation. Our data establishes
   RP11‐424G14.1 as a Y‐linked, human‐specific lncRNA that plays a
   multifaceted role in regulating sexual dimorphism.

AUTHOR CONTRIBUTIONS

   Conceptualization, K.K. and Y.L.; Methodology, K.K. and Y.L.; Data
   collection and analysis, K.K. and Y.L.; Writing, reviewing and editing,
   K.K. and Y.L. All authors have read and agreed to the published version
   of the manuscript.

FUNDING INFORMATION

   Y.L. has received research support from the NIH National Institute of
   Arthritis and Musculoskeletal and Skin Diseases grant K01 AR073340, R01
   AR078781 and the National Institute of Dental and Craniofacial Research
   grant R21 DE031765.

CONFLICT OF INTEREST STATEMENT

   The authors have stated explicitly that there are no conflicts of
   interest in connection with this article.

Supporting information

   Figure S1.
   [133]FBA2-7-e1479-s001.zip^ (22.6MB, zip)

ACKNOWLEDGMENTS