Abstract

   Periodontitis in menopausal women tends to exacerbate, potentially
   resulting in tooth loss and increased risk of systemic diseases. The
   prior research demonstrates that pro‐inflammatory changes in
   macrophages under estrogen deficiency exacerbate periodontitis.
   However, the primary environmental factors contributing to the
   alteration of monocyte‐macrophages remain unknown. Recent studies,
   including animal models and clinical trials, have found a correlation
   between accumulation of visceral adipose tissue (VAT) and periodontitis
   progression in postmenopausal women. In estrogen‐deficient mice,
   macrophages in VAT show a pro‐inflammatory state. Removing VAT
   alleviates periodontitis in OVX mice. DNA methylation sequencing shows
   increased methylation in macrophages, especially Jazf1
   hypermethylation, inhibiting its expression and promoting inflammation.
   Subsequently, small extracellular vesicles (sEVs) derived from
   pro‐inflammatory macrophages further intensify M1‐like polarization in
   resting macrophages, carrying inflammatory microRNAs like miR‐30e‐5p.
   Overall, this study proposes a novel perspective: periodontal pathogens
   act as initial triggers for inflammation, while chronic systemic
   inflammation, worsened by estrogen deficiency, is the main factor that
   exacerbates periodontitis. Pro‐inflammatory macrophages in VAT release
   sEVs, which activate resting macrophages into a pro‐inflammatory state
   upon encountering periodontal pathogens, resulting in persistent
   inflammation. Addressing the root causes of this dysregulation can lead
   to new therapeutic strategies for periodontitis and systemic
   inflammatory conditions, particularly in postmenopausal women.

   Keywords: estrogen deficiency; extracellular vesicles; macrophages,
   periodontitis, therapy
     __________________________________________________________________

   Chronic systemic inflammation, worsened by estrogen deficiency,
   underlies and exacerbates periodontitis. Proinflammatory macrophages
   within visceral adipose tissue are hyperactivated due to hyper‐DNA
   methylation, releasing small extracellular vesicles (sEVs) similar to
   those of their parent cells. When periodontal pathogens invade,
   quiescent macrophages at the site are activated into a proinflammatory
   state by endocytosing circulating sEVs, causing persistent
   inflammation.

   graphic file with name ADVS-12-e06121-g009.jpg

1. Introduction

   Periodontitis constitutes a significant public health concern with
   profound implications for both the economy and mortality rates. This
   condition not only leads to tooth loss and compromised oral function
   but also contributes to the development of chronic ailments such as
   cardiovascular diseases, diabetes, hypertension, and cancer. According
   to the latest study published in the Journal of Clinical Periodontology
   in 2023, there is a notable prevalence of periodontitis. The study
   estimated that ≈62% of adults were affected by periodontitis between
   the years 2011 and 2020, with a significant ratio of 23.6% experiencing
   severe forms of the condition.^[ [36]^1 ^]

   Although bacterial plaque is acknowledged as the primary etiological
   factor in periodontitis, the host response to pathogenic stimulation is
   crucial in determining the disease's outcome. Consequently, systemic
   factors can also influence the progression of periodontitis. Previous
   research indicates a correlation between estrogen deficiency and the
   aggravated periodontitis, as it has been observed to contribute to an
   inflammatory status.^[ [37]^2 ^] The aged tendency of population has
   led to the emergence of menopause as an inevitable issue, prompting the
   use of complementary hormone therapy. However, this therapy is
   associated with side effects such as tumorigenesis. Accordingly,
   preventing periodontitis in postmenopausal women becomes a burning
   problem. Addressing this issue, our prior research identified that
   macrophages within periodontal tissues exhibit pro‐inflammatory
   alterations and impede the osteogenic differentiation of mesenchymal
   stem cells in the context of estrogen deficiency, thereby exacerbating
   periodontitis.^[ [38]^3 ^] Nonetheless, we consider that periodontal
   macrophages predominantly originate from monocytes in the bloodstream
   following activation by local pathogenic bacteria. This raises
   questions regarding the primary environmental triggers responsible for
   their pro‐inflammatory transformation.

   In prior research, it was observed that the majority of ovariectomized
   (OVX) mice exhibited a pronounced phenotype characterized by the
   accumulation of visceral adipose tissue. Subsequent investigations
   indicated a correlation, to a certain degree, between the accumulation
   of abdominal fat tissue in postmenopausal women and the progression of
   periodontitis. Additionally, it was discovered that macrophages within
   the adipose tissue of OVX mice exhibited a pro‐inflammatory phenotype,
   and their periodontitis symptoms were mitigated following adipotomy.
   Consequently, we are interested in exploring the potential, yet
   unexplored, connection between adipose tissue and periodontal
   macrophages. Recent studies have further demonstrated an increase in
   visceral fat accumulation in mice subsequent to ovariectomy (OVX).^[
   [39]^4 ^] It also suggests that the accumulation of abdominal fat is a
   significant characteristic of postmenopausal women.^[ [40]^5 ^]
   Additionally, it has been reported that postmenopausal women with
   obesity suffer from exacerbated periodontitis,^[ [41]^6 ^] but the
   underlying mechanisms remain unknown.

   Adipose tissue is an active endocrine organ that secretes a variety of
   cytokines and molecules involved in immune regulation.^[ [42]^7 ,
   [43]^8 ^] Depending on the distribution, adipose tissue can be
   categorized into subcutaneous adipose tissue (SAT) and visceral adipose
   tissue (VAT).^[ [44]^9 ^] The immune cells in visceral adipose tissue
   play a crucial role in regulating the homeostasis and inflammation
   levels of adipose tissue during pathological conditions, as well as in
   modulating their quantity and functions.^[ [45]^10 ^]

   Among the various immune cells, macrophages are particularly abundant
   in adipose tissue. Adipose tissue macrophages play a crucial role in
   modulating systemic cellular functions.^[ [46]^11 ^] On one hand, it is
   believed that adipose tissue serves as an additional reservoir for
   macrophages originating from the bone marrow, where undifferentiated
   monocytes infiltrate and differentiate into mature macrophages,
   potentially contributing to an increased macrophage population that
   might play greater roles in systemic inflammation.^[ [47]^12 ^] In
   healthy adipose tissue, the proportion of macrophages is typically less
   than 10% of the total cell population. On the contrary, in pathological
   circumstances, this ratio can increase significantly, reaching up to
   50% due to the accumulation of circular mononuclear macrophages in
   adipose tissue.^[ [48]^13 , [49]^14 ^] Macrophages within visceral
   adipose tissue predominantly originate from the infiltration of
   monocyte‐macrophages derived from peripheral blood.^[ [50]^15 ^] Mature
   macrophages expressing F4/80 are more inclined to sustain an
   inflammatory phenotype. Researchers have demonstrated that visceral fat
   harbors a greater number of infiltrating macrophages and exhibits
   elevated levels of inflammatory cytokine expression relative to
   subcutaneous fat.^[ [51]^16 ^] The interaction between these
   macrophages and adipocytes exerts a prolonged influence on the body
   through metabolic processes and their dysfunction is closely linked to
   metabolic diseases.^[ [52]^17 , [53]^18 , [54]^19 ^] Notably, the
   polarization phenotypes of these macrophages play a crucial role in
   determining their diverse immune response functions. For instance, the
   M1 polarization of macrophages promotes inflammation, whereas the M2
   polarization exerts anti‐inflammatory effects.^[ [55]^20 , [56]^21 ^]

   On the other hand, the substantial blood flow and active secretory
   functions of macrophages in adipose tissue confer a more pronounced
   systemic impact through secretion compared to macrophages located in
   bone marrow.^[ [57]^22 ^] However, the mechanisms by which these
   macrophages influence distal immune responses remain inadequately
   understood.

   Among the secretions of VAT, a range of cytokines and molecules,
   including tumor necrosis factor (TNF), Interleukin‐6, and acute phase
   proteins, have the potential to induce a state of high oxidation.
   Although the underlying biological mechanism remains unclear, existing
   theories propose that alterations in the immune inflammatory profile of
   obese individuals might impact the immune response to periodontitis.^[
   [58]^8 , [59]^23 ^] For the past few years, extracellular vesicles have
   emerged as stable vehicles for long‐range communication, exerting
   notable impacts on immune regulation.^[ [60]^24 , [61]^25 , [62]^26 ,
   [63]^27 ^] Specifically, small extracellular vesicles (sEVs), measuring
   ≈30 – 200 nm in diameter, have been recognized as crucial intercellular
   regulators due to their ability to transfer bioactive molecules such as
   DNA, mRNA, non‐coding RNA, and proteins.^[ [64]^3 , [65]^27 , [66]^28
   ^]

   Consequently, this study seeks to elucidate the primary changes
   occurring in macrophages under conditions of estrogen deficiency and
   their impact on pro‐inflammatory regulation. Furthermore, we intend to
   investigate the influence of the small extracellular vesicle (sEV)
   pathway, derived from pro‐inflammatory macrophages, on inflammation and
   the exacerbation of periodontitis. This will involve identifying key
   inflammatory molecules present in sEVs and elucidating their underlying
   mechanisms. In general, our study is expected to provide new a strategy
   for the treatment of not only periodontitis, but also other systemic
   inflammatory diseases in postmenopausal women.

2. Results

2.1. Severity of Periodontitis and Abdominal Obesity was Aggravated in
Postmenopausal Women

   To explore the relationship between periodontitis and abdominal obesity
   in postmenopausal women. We investigated the periodontal probing depth
   (PD), attachment loss (AL), waist circumference (WC)
   and waist‐to‐height ratio (WHtR) of premenopausal and postmenopausal
   women (Figure [67] 1A). The results showed that the waist circumference
   (79.85 ± 3.35 cm) and waist‐to‐height ratio (0.49 ± 0.02) of
   postmenopausal women were significantly higher than those of
   pre‐menopausal women (74.65 ± 2.02 cm, P <0.001) and waist‐to‐height
   ratio (0.46 ± 0.01, P <0.001). In addition, probing depth (1.38 ±
   0.66 mm) and attachment loss (1.53 ± 0.62 mm) in postmenopausal women
   were also aggravated in premenopausal women (PD, 0.52 ± 0.24 mm, P
   <0.001; AL, 0.58 ± 0.31 mm, P <0.001). The correlation between
   menopause age and WC, WHtR, AL, PD was determined by linear regression
   correlation analysis. The results showed that with increasing age after
   menopause, there was a positive correlation with WC, WHtR, AL and PD
   (Figure [68]1B). Furthermore, the link between abdominal
   obesity and periodontitis severity was verified. Importantly, WHtR was
   strongly and significantly associated with attachment loss (R^2 = 0.54,
   P <0.001) and probing depth (R^2 = 0.45, P <0.0001).

Figure 1.

   Figure 1
   [69]Open in a new tab

   Periodontitis and abdominal obesity aggravated after estrogen
   deficiency. A) Comparison of anthropometric and biological parameters
   between premenopausal and postmenopausal women. B) Correlation between
   WC, WHtR, AL, PD and years after postmenopausal, respectively. And the
   correlation between AL, PD and WHtR. C) The body weight of C57BL/6 mice
   increased after OVX operation. Data represent mean ± SD. *P < 0.05 (OVX
   group compared with Sham group), #P < 0.05 (OVX+EP group compared with
   Sham+EP group). D,E) Visceral fat in the abdominal cavity accumulated
   after the OVX operation. Data are shown as mean ± SD, with n = 6 in
   each group, ns: not significant, **P < 0.01, ***P < 0.001 (One‐way
   ANOVA). F) Representative 3D reconstruction images of maxillae
   harvested 4 weeks after periodontitis induction from OVX+EP group (n =
   4) or OVX+EP with VATectomy group (n = 5). G) The average linear
   distance from CEJ‐ABC of maxillary molars on both the buccal and
   palatal sides for each group. ***P <0.001 (Student's t test).

   To further investigate the role of macrophages in exacerbated
   periodontitis under conditions of estrogen deficiency, we developed an
   experimental periodontitis (EP) model and simulated estrogen deficiency
   through ovariectomy (OVX). Subsequently, we randomly allocated the mice
   into four groups: Sham, OVX, Sham + EP, and OVX + EP. In the OVX mice
   with periodontitis, we observed an increase in body weight
   (Figure [70]1C) and visceral fat accumulation (Figure [71]1D,E). Our
   previous studies have demonstrated that periodontitis is exacerbated in
   OVX mice following periodontitis induction,^[ [72]^3 ^] and we recently
   observed significant accumulation of abdominal adipose tissue. Notably,
   periodontitis‐related bone loss was reduced in OVX mice following the
   removal of abdominal fat (Figure [73]1F,G), suggesting that estrogen
   deficiency facilitates abdominal fat accumulation, which is potentially
   associated with the progression of periodontitis.

2.2. M1‐Polarization of Macrophages was Enhanced in Periodontal Tissue and
Visceral Adipose Tissue of OVX Mice with Periodontitis

   As macrophages have been reported as an important source of
   inflammation in adipose tissues, we collected the periodontal tissue of
   mice from Sham, OVX, Sham + EP, and OVX + EP groups (Figure [74] 2A),
   observing a significant increase of F4/80+ CD11b+ CD86+ macrophages in
   periodontal tissue, indicating enhanced M1‐polarized macrophages
   (Figure [75]2B,C). Given that the accumulation of abdominal adipose
   tissue is also a key feature of postmenopausal women, we obtained
   visceral adipose tissue of the above groups and conducted
   immunofluorescence staining and flow cytometry to detect the state of
   macrophages under different conditions. As data shown, there were
   numerous macrophages infiltrated in the visceral adipose tissue
   (Figure [76]2D). Consistent with the above data, the proportion of M1
   macrophages in the adipose tissue increased in OVX mice, especially
   after periodontitis modeling (Figure [77]2E,F). Overall, the findings
   point to a possible connection between dysfunctional macrophages in the
   periodontium and those in visceral adipose tissue.

Figure 2.

   Figure 2
   [78]Open in a new tab

   Enhanced M1 polarization in periodontal tissue and adipose tissue in
   OVX mice with periodontitis. A) A schematic overview of establishing a
   combined OVX mouse model for periodontitis. B,C) Representative live
   cell gates and scatter plots from Sham, Sham+EP, OVX and OVX+EP
   periodontal tissues identifying CD11b+ F4/80+ cells, further subdivided
   by low or high CD86 expression (percentage of total live cells), and
   corresponding group data. D) Representative confocal images of
   F4/80 immunofluorescence in abdominal adipose tissues from Sham,
   Sham+EP, OVX and OVX+EP mice. E F) Representative live cell gates and
   scatter plots from Sham, Sham+EP, OVX and OVX+EP adipose tissues
   identifying CD11b+F4/80+ cells, further subdivided by low or high CD86
   expression (percentage of total live cells), and corresponding group
   data. Data represent mean ± SEM. *P <0.05, **P <0.01, ***P <0.001, ns:
   not significant. (One‐way ANOVA).

2.3. Macrophages were Critical for the Aggravated Periodontitis in OVX Mice

   To explore the role of macrophages in periodontitis, macrophages were
   cleared from mice through clodronate liposomes. According to F4/80
   staining by fluorescence activated cell sorting (FACS) and
   immunofluorescence, more than 80% of macrophages in the spleen were
   cleared after 24 h of clodronate injection, demonstrating a successful
   depletion of macrophages (Figure [79] 3A–C). Furthermore,
   immunofluoresence of F4/80 in visceral adipose tissue also verified the
   removal of macrophages (Figure [80]3D). As data shown, the elimination
   of macrophages reduced alveolar bone loss (Figure [81]3E,F) and
   inflammatory‐cell infiltration in periodontal tissue in OVX mice with
   periodontitis (Figure [82]3G). Furthermore, qPCR analysis of the mRNA
   expression level of inflammatory genes Il1b and Tnfα also revealed
   alleviated inflammation in periodontal tissue after macrophage
   depletion (Figure [83]3H,I). Collectively, the data demonstrates the
   critical role of macrophages in aggravated periodontitis under estrogen
   deficiency.

Figure 3.

   Figure 3
   [84]Open in a new tab

   Macrophage depletion alleviated periodontitis in OVX mice. A) Gating
   strategy and representative flow cytometric plots of F4/80 expressing
   macrophages from the spleens of control‐liposome‐injected and
   clodronate‐liposome‐injected mice. B) Percentage of F4/80+ macrophages
   from the spleen of the two groups above. C) Representative
   immunofluorescence of the spleen after 24 h of injection, stained for
   F4/80 (green) and Hoechst (blue). D) Representative immunofluorescence
   of the visceral adipose tissue after 24 h of injection, stained for
   F4/80 (green) and Hoechst (blue). E) Reconstructed 3D images of
   maxillae harvested 4 weeks after periodontitis induction. F) The
   average linear distance from CEJ‐ABC of maxillary molars on both the
   buccal and palatal sides for each group (n = 6). G) F4/80, CD3, MPO and
   MHC‐II were used to monitor macrophages, T cells, neutrophils and
   MHCII‐positive antigen‐presenting cells, respectively, in the
   periodontal tissues of each group. The arrows indicate infiltrating
   positive cells. H,I) The mRNA levels of Il1b and Tnfα in periodontal
   tissues were determined by qPCR analysis. Results are shown as a fold
   change in expression with respect to the baseline of the control group.
   Data represent mean ± SD. * P <0.05, **P <0.01, ***P <0.001 (Student's
   t test).

2.4. DNA Methylation in Macrophages Alters under Estrogen Deficiency

   Based on the above findings, we focused on the reasons for
   dysfunctional macrophages under estrogen deficiency. Considering
   estrogen deficiency as an environmental factor, it is plausible to
   hypothesize that DNA methylation serves as a primary epigenetic
   regulatory mechanism for environmental factors and chronic inflammatory
   regulation. Since hormonal alterations regulate epigenetics, we
   detected differential DNA methylation in macrophages from Sham and OVX
   mice (Figure [85] 4A). The proportion of methylated CpGs and the
   average methylation level were shown (Figure [86]4B,C). Comparisons of
   methylated C levels between different gene regions and functional
   elements revealed that both groups presented hypomethylation in the
   promoter region (Figure [87]4D). Interestingly, total methylated CG,
   CHG, and CHH frequencies in the promoter region were higher in
   OVX‐BMDMs than the control (Figure [88]4E). The methylation levels of
   every type C in the chromosome or in different functional elements in
   genome of Sham‐BMDMs and OVX‐BMDMs were shown in Tables [89]S1 and
   [90]S2 (Supporting Information), respectively. The results of Kyoto
   Encyclopedia of Genes and Genomes (KEGG) pathway analysis indicated
   that genes having hyper‐DMRs in their promoters are mainly involved in
   20 pathways, including Carbon metabolism, Pyrimidine metabolism,
   Protein digestion and absorption, GnRH signaling pathway, Breast
   cancer, Phospholipase D signaling pathway, Pathways in Cancer, Gastric
   cancer, Glucagon signaling pathway, Adipocytokine signaling pathway,
   Citrate cycle (TCA cycle), Pentose phosphate pathway, Hepatocellular
   carcinoma, Pancreatic cancer, AMPK signaling pathway, Ras signaling
   pathway, Biosynthesis of amino acids, Relaxin signaling pathway, ErbB
   signaling pathway and Insulin signaling pathway (Figure [91]4F). Then
   we further identified the top 30 genes exhibiting the most significant
   differences in DNA methylation (Table [92]S3, Supporting Information).
   Notably, the Jazf1 gene emerged as closely associated with adipose
   tissue and macrophages. Among the DMR genes, the DNA methylation ratio
   of Jazf1 promoter reached 55% in the OVX group compared with 21% in the
   control group (Figure [93]4G). Subsequently, we validated the
   differential expression of the Jazf1 gene in macrophages from Sham and
   OVX mice using qPCR. As anticipated, the mRNA expression level of Jazf1
   was reduced in the OVX group (Figure [94]4H). These findings suggest
   that estrogen deficiency alters DNA methylation levels in macrophages,
   and JAZF1 may serve as a potential therapeutic target.

Figure 4.

   Figure 4
   [95]Open in a new tab

   DNA methylation sequencing revealed differential methylation levels in
   BMDMs between Sham and OVX mice. A) Procedures of DNA methylation level
   profiling. B) The proportion of every type of mC in BMDMs from Sham and
   OVX mice. C) The proportion of different types of distribution
   methylation C in BMDMs from the groups above (Black line: mC, Red line:
   mCG, Blue line: mCHG, Green line: mCHH). D) The distribution of
   methylation levels in different functional elements in genome of BMDMs
   from the groups above (a: upstream, b:first exon, c:first intron,
   d:internal exon, e:internal intron, f:last exon, g: downstream). E) The
   methylation distribution of the promoter region of the genome, and the
   corresponding CpG density. In the heat map at left below, CpG density
   (X‐axis) is defined as the number of CPG in a 200 bp window, and the
   vertical axis represents the average methylation level of CPG. The thin
   black line shows the median methylation level for a given CpG density
   in such a window. The red areas, from light to dark, indicate the
   number of CPG at a particular methylation level and CpG density. The
   blue bar at the top shows the distribution of CpG density, mapped on
   the horizontal axis, and the green bar at the right shows the
   distribution of methylation levels, mapped on the vertical axis. F)
   KEGG pathway analysis of genes having hyper‐DMRs in the promoter of OVX
   BMDMs. The top 20 KEGG pathways were shown. G) The ratio of DNA
   methylation in the promoter of the Jazf1 gene. H) Differential
   expression of selected Jazf1 gene in macrophages derived from Sham or
   OVX mice was validated by qPCR. Data represent mean ± SD. **P <0.01
   (Student's t test).

2.5. Hypermethylation of the Jazf1 Promoter Enhances Pro‐Inflammatory Changes
in Macrophages

   To corroborate the hypermethylation of the Jazf1 promoter in
   macrophages derived from ovariectomized (OVX) mice, as indicated by
   whole‐genome bisulfite sequencing (WGBS) methylation array results, we
   conducted an analysis of eight CpG sites within the Jazf1 promoter
   using pyrosequencing. This method is considered the gold standard for
   assessing methylation levels due to its precision and reproducibility.
   Among these sites, seven CpG positions (cg53069459, cg53069412,
   cg53069374, cg53068940, cg53068938, cg53068914, cg53068891, cg53068871)
   exhibited significant hypermethylation in the Jazf1 promoter of
   macrophages from the OVX group (Figure [96] 5A). Immunofluorescence
   analysis revealed that JAZF1 expression was markedly reduced in
   macrophages cultured under estrogen‐deficient conditions compared to
   the control group (Figure [97]5B). To further elucidate the impact of
   Jazf1 promoter methylation on its expression, we examined the mRNA
   expression levels of Jazf1 in estrogen‐deficient macrophages, both with
   and without treatment with 5‐Aza‐2′‐deoxycytidine. The results
   demonstrated a significant increase in Jazf1 expression following
   demethylation treatment in estrogen‐deficient macrophages
   (Figure [98]5C). These findings suggest that hypermethylation of the
   Jazf1 promoter may serve as a potential mechanism underlying the
   aberrantly low expression of Jazf1 in macrophages under conditions of
   estrogen deficiency.

Figure 5.

   Figure 5
   [99]Open in a new tab

   DNA methylation of Jazf1 regulates macrophage polarization. A)
   Representative images of pyrosequencing for Jazf1 gene in macrophages
   derived from Sham or OVX mice. Eight methylation sites were identified
   using pyrosequencing for target‐specific analysis in the JAZF1 promoter
   region. B) Representative immunofluorescence of the macrophages
   cultured in control or estrogen deficient condition, stained for JAZF1
   (green) and Hoechst (blue). C) The mRNA expression levels of Jazf1 in
   macrophages with (+) or without (‐) the 5‐Aza‐2′‐deoxycytidine
   treatment. Aza (‐) and Aza (+) were abbreviated by
   5‐Aza‐2′‐deoxycytidine (‐) and 5‐Aza‐2′‐deoxycytidine (+),
   respectively. D) The mRNA expression levels of Tnfα, Il6,Il1β genes in
   macrophages with (+) or without (‐) the 5‐Aza‐2′‐deoxycytidine
   treatment. E) Levels of JAZF1 protein in macrophages following JAZF1
   knockdown were detected by western blot. F) Immunofluorescence
   detection of iNOS and JAZF1 expression following transfection with
   JAZF1 siRNA or NC control. G) The mRNA expression levels of Il1β and
   Arg1 genes in macrophages after JAZF1‐siRNA or NC transfection. Data
   represent mean ± SD. *P <0.05, **P <0.001 (Student t‐test). H) The
   expression of iNOS in periodontal tissue of mice injected with
   JAZF1‐siRNA or NC was detected using immunofluorescence staining.

   Juxtaposed with another zinc finger (Znf) protein 1 (JAZF1, also known
   as TIP27, ZNF802) encodes a 27 kDa protein with three C2H2‐type ZnFs
   and functions as a repressor of DNA response element 1 (DR1)‐dependent
   transcription of NR2C2, TR4.^[ [100]^29 ^] JAZF1 is integral to glucose
   homeostasis and the pathophysiology of metabolic diseases. It is
   ubiquitously expressed across various tissues, with particularly high
   expression levels in adipose tissues (AT). JAZF1 mitigates inflammation
   and immune cell infiltration by attenuating the NF‐κB pathway.^[
   [101]^25 ^] Studies have demonstrated that JAZF1 reduces CD11c+ adipose
   tissue macrophages and suppresses the secretion of TNF‐α and IL‐1β,
   thereby diminishing adipose tissue inflammation in diabetic mice.^[
   [102]^26 ^] To evaluate our hypothesis, we analyzed the mRNA expression
   levels of Tnf‐α, Il6, and Il1β in estrogen‐deficient macrophages
   treated with (+) or without (‐) 5‐Aza‐2′‐deoxycytidine. The expression
   levels were significantly reduced following the inhibition of DNA
   methylation by 5‐Aza‐2′‐deoxycytidine (5‐Aza) (Figure [103]5D).
   Subsequently, we assessed the role of JAZF1 in macrophage inflammation
   using JAZF1‐siRNA or negative control. Western blot analysis was
   employed to confirm the knockdown efficiency of JAZF1 (Figure [104]5E).
   Immunofluorescence analysis revealed a significant reduction in JAZF1
   expression in macrophages treated with JAZF1‐siRNA compared to the
   control group. Conversely, JAZF1‐knockdown led to a notable increase in
   iNOS expression (Figure [105]5F). Subsequent qPCR analysis confirmed
   that JAZF1‐knockdown resulted in elevated Il1β expression and reduced
   Arg1 expression (Figure [106]5G). Furthermore, in a mouse model, iNOS
   expression in periodontal tissues was observed to increase following
   periodontal injection of JAZF1‐siRNA, as compared to the negative
   control injection group (Figure [107]5H). Collectively, these findings
   suggest that hypermethylation of the JAZF1 promoter suppresses its
   expression and it is likely to promote pro‐inflammatory changes in
   macrophages.

2.6. M1 Macrophage‐Derived sEVs Promoted M1 Macrophage Polarization

   Previous studies have demonstrated that small extracellular vesicles
   (sEVs) are critical in cellular communication and regulation of distant
   organs. Our prior research successfully isolated, extracted, and
   characterized macrophage‐derived extracellular vesicles.^[ [108]^3 ^]
   The characteriazation of small extracellular vesicles was also
   conducted in this study via the Transmission electron microscopy (TEM),
   Nanoflow cytometry (nanoFCM) and Western Blot analysis of CD63 and CD81
   marker (Figure [109]S1, Supporting Information). To investigate the
   potential connection between adipose tissue and periodontitis, small
   extracellular vesicles were isolated from visceral adipose tissue
   (AD‐sEVs) and administered intravenously into mice following
   pretreatment with DiR fluorescent dye (Figure [110]S2A, Supporting
   Information). The in vivo imaging results implied that AD‐sEVs could
   reach periodontium via circulation (Figure [111]S2B, Supporting
   Information). Further vesicle‐tracing experiments revealed that
   DiO‐labeled M1‐sEVs could be effectively endocytosed by macrophages, as
   shown by immunofluorescence and flow cytometry (Figure [112]S2C,
   Supporting Information). Moreover, M1‐sEVs derived from M1‐polarized
   macrophages cultured in estrogen deficiency or in control conditions
   presented similar uptake efficiency (Figure [113]S2D,E, Supporting
   Information).

   To explore the relationship between viseral adipose tissue and
   circulating M1‐sEVs, we further conducted VAT‐resection or control
   treatment on ovariectomized (OVX) mice, subsequently isolating small
   extracellular vesicles (sEVs) from the serum. The expression of the M1
   macrophage marker CD86 in serum sEVs was quantified via nano‐flow
   cytometry to assess the levels of M1‐sEVs. The experiments demonstrate
   a significant reduction in circulating M1‐sEV levels following the
   excision of visceral adipose tissue in OVX mice (Figure [114] 6A). To
   further investigate the effect of M1‐sEVs, macrophages in resting state
   (M0) were co‐cultured with M1‐sEVs (Figure [115]6B,C). The data of qPCR
   analysis showed that M1‐sEVs up‐regulated the mRNA expression level of
   M1 characteristic genes, including Il1b, Tnfα, Il6 and Nos2
   (Figure [116]6D–G). Consistently, FACS detection also revealed that
   M1‐sEVs induced M1‐like macrophage polarization (Figure [117]6H,I). In
   vivo experiments also verified the iNOS‐positive M1‐polarized
   macropahges increased in periodontal tissues following M1‐sEVs
   injection (Figure [118]6J). Taken together, the above results indicated
   that M1 macrophage‐derived sEVs might potentiate the M1
   pro‐inflammatory phenotype after being endocytosed by quiescent
   macrophages.

Figure 6.

   Figure 6
   [119]Open in a new tab

   Extracellular vesicles derived from M1 macrophages promoted M1
   polarization. A) Small extracellular vesicles isolated from the serum
   of OVX mice with VATectomy (n = 6) or control treatment (n = 5) were
   analyzed the ratio of CD86‐positive via nanoflow cytometry (nanoFCM).
   Data represent mean ± SD. B,C) A schematic illustration shows bone
   marrow isolation, macrophage induction, stimulation and EV‐isolation.
   D–G) After 6 h of incubation with 1, 5,10 µg M1‐sEVs or an equal volume
   of PBS, the mRNA expression levels of the M1 characteristic gene
   profile, including Il‐1β, Il6, Tnf‐α and Nos2, were measured. H,I)
   Representative gates and corresponding group data for identifying
   F4/80+ CD86+ cells in BMDMs after treatment of 10 µg (low) or 50 µg
   (high) M1‐sEVs, 10 µg M0‐sEVs, or PBS control. Data represent mean ±
   SD. *P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant. (One‐way
   ANOVA). J) The expression of iNOS in periodontal tissue of mice
   injected with M1‐sEVs or PBS was detected using immunofluorescence
   staining.

2.7. miR‐30e‐5p Contained in M1‐sEVs Induced Pro‐Inflammatory Polarization

   Further miRNA sequencing was applied to find out the mechanism of the
   pro‐inflammatory role of M1‐sEVs. Using miRNA expression levels as a
   basis for coloring, a heatmap is created to present the different miRNA
   profiles between M1‐sEVs and M2‐sEVs (Figure [120] 7A). The Venn
   diagram showed differences and overlaps of detected miRNAs between
   M1‐sEVs and M2‐sEVs (Figure [121]7B). We confirmed the validity of
   miRNA profiling via further qPCR detection (Figure [122]7C). Among
   them, miR‐30e‐5p, which is highly expressed in M1‐sEVs, was also found
   to largely increase in serum derived sEVs (Figure [123]7D) and AD‐sEVs
   (Figure [124]7E) from OVX+EP mice. The TargetScan database provided
   target gene predictions for miR‐30e‐5p. According to the data, two
   immune‐related genes, Irf4 and Socs1 (Figure [125]7F), were potential
   targets of miR‐30e‐5p. Further qPCR analysis showed that the mRNA
   expression levels of Socs1 and Irf4 decreased after treatment of
   M1‐sEVs (Figure [126]7G,H).

Figure 7.

   Figure 7
   [127]Open in a new tab

   Profiling of miRNAs derived from small extracellular vesicles of M1/M2
   macrophages. A) A heatmap of differential miRNA expression between
   M1‐sEVs and M2‐sEVs, ranging from red (upregulated) to green
   (downregulated). B) Venn diagram of the miRNAs’ expression in M1‐sEVs
   and M2‐sEVs. C) qPCR analysis of the level of miRNA expression in
   M1‐sEVs and M2‐sEVs. Data represent mean ± SD, ns: not significant,
   ***P <0.001 (Student's t test). D) qPCR analysis of different
   miR‐30e‐5p expression in serum‐derived‐sEVs from Sham, Sham+EP, OVX and
   OVX+EP mice. E) qPCR analysis of miR‐30e‐5p expression in sEVs derived
   from abdominal adipose tissue in the groups above. F) Schema of
   predicted miRNA‐binding sites at the 3′ UTR region of Irf4 and Socs1
   genes. G,H) qPCR was used to detect the mRNA levels of the Socs1 and
   Irf4 genes in macrophages after 6 h of incubation with 1, 5, and 10 µg
   M1‐sEVs or an equal volume of PBS. Data represent mean ± SD, ns: not
   significant, *P <0.05, **P <0.01, ***P <0.001 (One‐way ANOVA).

   M1‐sEVs were electroporated with either miR‐30e‐5p mimics or inhibitors
   and subsequently introduced to macrophages to assess the mRNA
   expression levels of miRNA targets and pro‐inflammatory genes. The
   expression levels of the miRNA target, Socs1, were found to decrease
   following the overexpression of miR‐30e‐5p (Figure [128]8A).
   Conversely, its expression in recipient cells was restored upon
   treatment with M1‐sEVs electroporated with the miR‐30e‐5p inhibitor
   (Figure [129]8B). Immunofluorescence analysis revealed a decrease in
   SOCS1 expression and an increase in iNOS expression subsequent to
   miR‐30e‐5p overexpression (Figure [130]8C). Furthermore, in recipient
   macrophages, M1‐associated genes, including Il1b, Il6, Tnfα, and Nos2,
   were upregulated following incubation with sEVs pre‐treated with
   miR‐30e‐5p mimics (Figure [131]8D–G). These genes were subsequently
   downregulated when co‐cultured with inhibitor‐loaded M1‐sEVs
   (Figure [132]8H–K). Collectively, these findings suggest that
   miR‐30e‐5p in M1‐sEVs may facilitate M1‐like macrophage polarization by
   modulating Socs1 expression. This study highlights the role of adipose
   tissue macrophages as initiators of exacerbated periodontitis in
   estrogen‐deficient environments through the action of amplifier
   extracellular vesicles and its loading miR‐30e‐5p.

Figure 8.

   Figure 8
   [133]Open in a new tab

   miR‐30e‐5p promoted M1 polarization via targeting Socs1. A) The mRNA
   levels of Socs1 gene in macrophages were detected after 6 h incubation
   of NC, miR‐30e‐5p mimics or equal volume of PBS by qPCR. B) The mRNA
   levels of Socs1 genes in macrophages were detected after 6 h of
   incubation. C) Immunofluorescence analysis of iNOS and SOCS1 expression
   following transfection with miR‐30e‐5p mimics or NC. D–G) The mRNA
   levels of Il1β, Il6, Tnfα and Nos2 genes in macrophages were detected
   after 6 h of incubation with NC, miR‐30e‐5p mimics, or an equal volume
   of PBS by qPCR. H–K) The mRNA levels of Il1β, Il6, Tnfα and Nos2 genes
   in macrophages were detected after 6 h incubation of M1‐sEVs, M1‐sEVs +
   NC, M1‐sEVs + miR‐30e‐5p inhibitor, or an equal volume of PBS by qPCR.
   Data represents mean ± SD. ns, not significant, *P <0.05, **P <0.01,
   ***P <0.001 (One‐way ANOVA).

3. Discussion

   During our experiments, we observed that OVX mice exhibited a phenotype
   characterized by the accumulation of visceral adipose tissue.
   Furthermore, we identified abdominal obesity, a marker of visceral fat
   accumulation,^[ [134]^24 ^] as a prominent characteristic in
   postmenopausal women^[ [135]^25 , [136]^26 , [137]^27 ^] and correlated
   with periodontitis. Utilizing the OVX mouse model for periodontitis, we
   discovered that estrogen deficiency led to an increase in M1‐polarized
   macrophages within the visceral adipose tissue of OVX mice with
   periodontitis. Additionally, small extracellular vesicles derived from
   M1‐polarized macrophages facilitated systemic M1‐like pro‐inflammatory
   polarization and induced excessive M1‐macrophage infiltration, through
   delivering pro‐inflammatory miRNAs, including miR‐30e‐5p. The upstream
   mechanism of the dysfunctional macrophage status might result from the
   abnormal adipose microenvironment caused by hyper DNA methylation of
   the promoter of the Jazf1 gene in macrophages under estrogen
   deficiency.

   Previous epidemiological evidence demonstrates that periodontal
   diseases promote inflammation in adipose tissue, leading to diabetes
   and obesity.^[ [138]^30 ^] However, the majority of prior research has
   predominantly concentrated on the association between obesity and
   inflammation. Distinct from these earlier studies, our research
   introduces a novel perspective by examining sex‐specific obesity,
   particularly abdominal obesity induced by estrogen deficiency. We
   explore the epigenetic alterations in adipose tissue macrophages within
   this unique environmental context. Furthermore, we investigate the
   mechanisms underlying the transition from systemic pro‐inflammatory
   states to sub‐health conditions, integrating findings from both animal
   experiments and clinical sample analyses. This study suggests that
   periodontal pathogens serve merely as the initial trigger for a cascade
   of inflammatory responses. The exacerbation of inflammation and the
   progression of periodontitis are primarily attributed to a “sub‐health”
   state of chronic inflammation throughout the body, which is exacerbated
   by estrogen deficiency. This “sub‐health state” is analogous to the
   heightened susceptibility to acne and inflammation observed in women
   during puberty,^[ [139]^31 ^] when elevated estrogen levels render the
   skin more reactive to bacterial irritation, representing a “pre‐disease
   state”.^[ [140]^32 , [141]^33 ^] Consequently, this condition
   elucidates our findings that not all patients with estrogen deficiency
   will inevitably experience exacerbation of periodontitis. Although
   there is a general positive correlation, some individuals maintain a
   healthy periodontal status. This discrepancy may be attributed to
   effective oral hygiene practices in certain patients, which prevent the
   activation of periodontal pathogens as a critical trigger.

   Prior research has primarily concentrated on the direct impact of
   estrogen deficiency on osteoclasts and osteoblasts. However, clinical
   investigations have revealed phenomena, such as heightened inflammation
   in the periodontal tissues of menopausal women, that cannot be solely
   attributed to osteoclast activity.^[ [142]^34 , [143]^35 , [144]^36 ^]
   Recent studies have demonstrated that estrogen deficiency can modify
   the immune‐inflammatory response, resulting in heightened inflammation
   and an increased susceptibility to infections.^[ [145]^37 ^] Our study
   identified an elevation in macrophage inflammation in
   estrogen‐deficient mice, alongside an increased pro‐inflammatory
   phenotype of macrophages in both adipose and periodontal tissues.
   Consistently, evidence suggests that estrogen promotes M2 macrophage
   polarization and facilitates browning in adipose tissue.^[ [146]^38 ^]
   In addition, a study found that estrogen deficiency caused M1
   macrophage polarization and promoted steatohepatitis in obese OVX
   mice.^[ [147]^39 ^]

   Given that hormonal changes belong to environmental stimuli, might
   regulate gene mainly through an epigenetic way, we subsequently
   investigated the upstream mechanism of how estrogen deficiency
   initially altered adipose tissue macrophages from the perspective of
   epigenetics and conducted DNA methylation sequencing. Notably, among
   the differentially methylated region (DMR) genes, we discovered that
   the promoter region of Jazf1 exhibited hypermethylation in estrogen
   deficiency. JAZF1 (juxtaposed with another zinc finger protein 1), also
   known as TIP27 (TAK1‐interacting protein 27) or ZNF802 (zinc finger
   protein 802), is a 243 amino acid protein that localizes to the nucleus
   and contains three C2H2‐type zinc fingers. It interacts with the
   nuclear orphan receptor TR4 and is thought to function as a
   transcriptional repressor. Previous investigations have established
   that JAZF1 suppresses inflammation (including IL‐1β, IL‐4, IL‐6, IL‐8,
   IL‐10, TNFα, IFN‐γ, IAR‐20, COL3A1, laminin, and MCP‐1) by reducing
   NF‐κB pathway activation and AT immune cell infiltration.^[ [148]^40 ,
   [149]^41 ^]  Meng et al. found that overexpression of JAZF1 decreased
   ATM numbers and pro‐inflammatory cytokines secretion, indicating that
   JAZF1 reduces adipose tissue inflammation via regulating macrophages.^[
   [150]^42 ^] Furthermore, JAZF1, commonly referred to as the diabetes
   gene, plays a critical role in regulating metabolism and inhibiting
   lipid accumulation. Dysregulated expression of JAZF1 has been
   implicated in the development of type 2 diabetes.^[ [151]^29 , [152]^43
   ^] Among the eight CpG sites validated by pyrosequencing, one site in
   the OVX group showed lower methylation than in the Sham group, which
   contradicts the general trend of increased methylation in the gene's
   promoter region identified by WGBS. This discrepancy highlights the
   regional and site‐specific characteristics of DNA methylation. WGBS
   provides an average methylation level across all CpG sites in a region,
   indicating a general increase in methylation in the JAZF1 promoter in
   the OVX group, while pyrosequencing offers precise measurements at
   specific sites, revealing that methylation can vary significantly
   within a region. Additionally, treatment of macrophages in the OVX
   group with the methylation inhibitor 5‐Aza revealed changes in the
   expression of JAZF1 and macrophage polarization genes, indirectly
   supporting the role of JAZF1 methylation in regulating macrophage
   polarization. Our research has additionally demonstrated that DNA
   hypermethylation of JAZF1 might contribute to inflammation and promote
   M1 macrophage polarization.

   Our study further demonstrated that the accumulation of visceral
   adipose tissue and macrophage inflammation under estrogen‐deficient
   conditions may exacerbate systemic chronic inflammation via small
   extracellular vesicles (sEVs). Consistent with previous research,
   extracellular vesicles derived from macrophages have been described to
   regulate inflammation and affect distant tissues.^[ [153]^44 , [154]^45
   ^] Exosomes from IL‐4 polarized macrophages, for example, have been
   linked to the control of cardiometabolic inflammation and diabetes.
   Reprogramming circulating monocytes reduced systematic inflammation in
   the circulation, aorta, adipose tissue, and the liver.^[ [155]^46 ^]
   Furthermore, the macrophage‐derived sEVs would target macrophages more
   efficiently. It has been reported that exosomes from
   visceral‐fat‐tissues express lower CD47 and thus facilitate their
   endocytosis by macrophages.^[ [156]^23 ^]

   To investigate the mechanism, we newly characterized a specific miRNA,
   miR‐30e‐5p, that could promote pro‐inflammatory M1 polarization.
   Notably, miR‐30e‐5p was predominantly expressed and abundant in
   macrophages.^[ [157]^3 , [158]^47 , [159]^48 ^] In addition, we
   demonstrated that miR‐30e‐5p suppresses the expression of Socs1, a gene
   recognized for its anti‐inflammatory regulatory functions. It has been
   reported that downregulated expression of SOCS1 enhances M1‐type
   polarization, probably via activating the JAK1/STAT1 pathway.^[
   [160]^49 ^] Furthermore, exosomal miR‐221 promotes M1 macrophage
   polarization via SOCS1/STATs to promote inflammatory response.^[
   [161]^50 ^] Meanwhile, IRF4 is considered a key regulator of
   anti‐inflammation and M2 macrophage polarization.^[ [162]^51 , [163]^52
   ^] Richa et al. identified that the upregulated miR‐30e‐5p promote
   inflammation by targeting key negative regulators such
   as SOCS1 and SOCS3 of innate immune signaling pathways.^[ [164]^48 ^]

   In conclusion, our study highlights that during non‐resting states,
   such as pathogen invasion, adipose tissue serves as a primary site for
   macrophage accumulation, second only to the bone marrow. Macrophages
   derived from mature monocytes, characterized by F4/80 positivity,
   predominate in this context. This study found that in an
   estrogen‐deficient environment, macrophages become pro‐inflammatory,
   increasing M1‐sEVs in circulation. These M1‐sEVs could further
   transform macrophages, and when periodontal pathogens stimulate and
   motivate monocytes‐macrophages, it worsens local periodontal macrophage
   polarization, prolonging inflammation. In summary, we propose that
   while periodontal pathogens trigger inflammation, the chronic
   “sub‐health” state from low estrogen post‐menopause amplifies
   inflammation, explaining why some women develop aggravated
   periodontitis while others remain healthy.

4. Conclusion

   This study found that most ovariectomized (OVX) mice developed
   significant visceral fat, similar to the abdominal fat increase seen in
   postmenopausal women linked to periodontitis progression. Removing
   visceral adipose tissues (VATectomy) reduced periodontitis symptoms in
   OVX mice. The research revealed that dysfunctional macrophages in VAT,
   caused by hyper DNA methylation of the Jazf1 gene promoter under
   estrogen deficiency, could probably lead to a pro‐inflammatory state.
   JAZF1‐knockdown resulted in M1‐polarization of macrophages.
   Additionally, small extracellular vesicles from these M1‐polarized
   macrophages promoted systemic inflammation by delivering
   pro‐inflammatory miRNAs like miR‐30e‐5p. These M1‐sEVs could further
   transform macrophages, and when periodontal pathogens stimulate and
   motivate monocytes‐macrophages, it worsens local periodontal macrophage
   polarization, prolonging inflammation. Generally, our study proposes a
   novel perspective: periodontal pathogens act as a trigger for
   inflammatory responses, but the real issue is chronic systemic
   inflammation, worsened by estrogen deficiency, which exacerbates
   periodontitis. M1‐sEVs can transform macrophages, and when periodontal
   pathogens activate these cells, it leads to prolonged local
   inflammation. Pro‐inflammatory macrophages in adipose tissue release
   sEVs, which activate quiescent macrophages at infection sites, causing
   persistent inflammation. Addressing the root causes of this
   dysregulation may offer novel therapeutic strategies for the treatment
   of periodontitis and other systemic inflammatory conditions,
   particularly in postmenopausal women.

5. Experimental Section

Ethical Consent

   The study  was approved by the Medical Ethics Committee of Hospital of
   Stomatology Sun Yat‐sen University (No. KQEC‐2023‐13‐0) and
   the Clinical Research Center of Hospital of Stomatology Sun Yat‐sen
   University (No. GHKQ‐2002301‐L2).

Participants and Criteria Design

   Women who visited the Hospital of Stomatology at Sun Yat‐sen University
   and consented to complete the study were eligible for participation. A
   total of 58 women were recruited, with data collected on the timing of
   menopause and periodontal assessment. Eligibility required the presence
   of amenorrhea for 12 consecutive months without any evident
   pathological changes or physiological causes. The cohort consisted of
   26 premenopausal women (aged >36 years) and 33 postmenopausal women
   (aged ≤65 years). The following conditions were grounds for exclusion:
   1) Pregnancy or a history of abdominal tumors, liver disease, kidney
   disease, colon disease, retroperitoneal disease, or other conditions
   causing abdominal enlargement. 2) A history of periodontitis treatment
   within the past year. 3) Diabetes, smoking, leukemia, dental trauma,
   and the use of nitrobenzene and phenytoin, which may exacerbate
   periodontitis. 4) Diseases affecting bone metabolism, rheumatism, and
   immune‐related disorders. The number of subjects enrolled in each
   group, as well as their average age, height, and weight were listed in
   Appendix Table [165]S6 (Supporting Information).

Mice

   Eight‐week‐old wild‐type female C57BL/6 mice were used and acquired
   from the laboratory animal center. Experimental periodontitis (EP) and
   ovariectomy (OVX) models were induced as described in the previous
   study.^[ [166]^3 ^] These animal experiments were certified by the
   Institutional Animal Care and Use Committee.

micro‐CT Analysis

   For micro‐CT analysis, 4% paraformaldehyde (PFA) was used to fixed the
   sampled maxillae and Scanco µCT50 scanner (Scanco Medical AG,
   Brutishly, Switzerland) was further applied for scanning as reported
   before.^[ [167]^53 ^] To assess alveolar bone loss, the distances
   between the cemento‐enamel junction and the alveolar bone crest were
   measured at four sites around each maxillary molar using 3D images. The
   average of 12 site measurements per sample was reported in millimeters.
   Evaluations included both buccal and palatal sides to comprehensively
   assess ligature‐induced bone loss.

Differentiation and Cell Culture of Bone Marrow Derived Macrophages (BMDMs)

   Macrophages were differentiated using 20 ng mL^−1 macrophage
   colony‐stimulating factor (M‐CSF) after being isolated from bone marrow
   and cultured with DMEM medium (including 10% FBS and 1% Penicillin and
   Streptomycin solution). For M1‐polarization, 100 ng mL^−1 LPS (Sigma,
   USA) was added, and cells were co‐cultured for 24 h. For
   estrogen‐deficient culturing, phenol red‐free DMEM medium was used with
   activated carbon‐adsorbed FBS, adding 17‐β estradiol (10^−9 m) as a
   control.

Macrophage Depletion

   Clodronate liposomes or control liposomes (Clophosome, Cat#F70101C‐N,
   FormuMax) in equal volume were intraperitoneally injected (0.2 mL per
   time) to clear macrophages. Flow cytometry and immunofluorescence were
   conducted to detect the depletion efficiency 24 h after injection.

Adipose Tissue Isolation

   Visceral adipose tissue was isolated and cut into small pieces using
   ophthalmic scissors and placed in DMEM culture medium with 10% BSA and
   1% Penicillin–streptomycin. Hemolysin was added to remove the
   erythrocytes.

Flow Cytometry Analysis

   Visceral adipose tissues and maxillary periodontal tissues were
   harvested from mice and digested by 2% collagenase I and 1% DNase for
   30 min. The above digests were filtered through nylon mesh filters and
   centrifuged at 800 rpm for 5 min. Precipitation was collected and
   resuspended in 5% fetal bovine serum (FBS) containing PBS. Then the
   cells were dyed with Alexa Fluor 488 anti‐mouse F4/80 (Cat# 123120,
   BioLegend, US), APC anti‐mouse CD11b (Cat# 101212, BioLegend, US) and
   PE anti‐mouse CD86 (Cat# 105007, BioLegend, US) on the ice for 20 min.
   After washing with PBS and centrifugation, 300 µL of 7‐AAD Viability
   Staining Solution (Cat# 420404, BioLegend, US) was added to per tube to
   detect cell activity. Following washing, centrifugation, and
   resuspension, a flow cytometer (BD Biosciences) was applied to detect
   the fluorescent signal of the above samples.

Immunofluorescence

   The visceral adipose tissues were fixed using a fatty tissue fixative
   solution and embedded in OCT at −80 °C. Meanwhile, the periodontal
   tissues were fixed with 4% PFA and decalcified with 0.5 m EDTA for 3
   days. They were then sliced at a thickness of 10 µm at temperatures
   below −30 °C. Before staining, the antigen was repaired with an antigen
   retrieval solution and blocked for 1 h at room temperature in 1% BSA.
   Then sections were incubated overnight with primary antibodies,
   including rat‐anti‐mouse anti‐F4/80 antibody (ab16911, Abcam, USA),
   anti‐CD3 antibody (ab5690, Abcam, USA), anti‐Myeloperoxidase (MPO)
   antibody (ab9535, Abcam, USA) and rabbit‐anti‐mouse anti‐MHC antibody
   (ab180779, Abcam, USA). On the next day, they were washed with PBS
   three times and stained with goat‐anti‐rabbit Alexa Fluor 488 or
   goat‐anti‐rat Alexa Fluor 488 second antibodies. Then the slices were
   counterstained with Hoechst for 10 min at 25 °C. And laser‐scanning
   confocal microscopy was used to observe the distribution of
   fluorescence.

Isolation of Small Extracellular Vesicles

   Visceral adipose tissues were collected and cultured in FBS‐free DMEM
   medium for 48 h. The supernatant was harvested for further isolation.
   Furthermore, macrophages were cultured in FBS‐free DMEM medium for 2
   days. The supernatants were centrifuged at 500 g for 10 min, 10 000 g
   for 20 min, and 120 000 g for 60 min. The precipitate was resuspended
   in PBS and centrifugated at 120 000 g for 30 min. The pellets were
   stored at −80 °C. The isolation and characterization of sEVs was
   described in our previous study.^[ [168]^3 ^]

Fluorescent Tracing of sEVs

   BMDMs were co‐cultured with 5 µg DiO‐labeled sEVs derived from M1
   macrophages cultured in estrogen deficiency (M1‐E sEVs) or in control
   conditions (M1+E sEVs) for 6 h. After washing with sterile PBS, 4% PFA
   was added for fixation. Then Hoechst dye (1:1000, Hoechst 33342,
   Beyotime Biotechnology, China) was applied to stain the cell nuclei for
   10 min at room temperature. Following thorough washing, the cells were
   observed under a laser confocal microscope. For in vivo tracing, mice
   were intravenously injected with 10 µg of DiR‐labeled AD‐sEVs. The
   distribution of the AD‐sEVs in periodontal tissue was observed via the
   IVIS Lumina II in vivo imaging system (Caliper Life Sciences, US).

qPCR Analysis

   Total RNA was extracted and reverse‐transcribed into cDNA as
   reported.^[ [169]^3 ^] Then SYBR Green (FastStart Essential DNA Green
   Master, 06924204001, Roche, USA) was used for qPCR analysis. The Roche
   Real‐Time PCR System was used to detect mRNA expression levels, which
   were normalized to β‐actin or Gapdh mRNA levels. To measure miRNA
   expression, the miRcute Plus miRNA First‐Strand cDNA kit (TIANGEN,
   Beijing, China) was applied to reverse‐transcribe RNA samples. The
   reverse primer (Cat# CD109, TIANGEN, Beijing, China) was used for
   further qPCR analysis. The primers used are listed in Table [170]S4
   (Supporting Information).

miRNA Profiling

   Total RNA samples were qualified and further generated for sequencing
   libraries. Using HiSeq2500 (Illumina; San Diego, CA) for parallel
   sequencing, the expression level of miRNAs were detected as
   described.^[ [171]^3 ^] A heatmap of miRNA sequencing data was
   generated via the software Cluster 3.0 and TreeView. TargetScan
   ([172]http://www.targetscan.org/) was used for miRNA target prediction.

Genome‐Wide DNA Methylation Sequencing

   For the construction of a whole‐genome bisulfite sequencing library,
   BMDMs from OVX or Sham mice were lysed in lysis buffer and digested by
   proteinase K. Three samples for replicates were used for DNA
   methylation sequencing in each group. Genomic DNA was isolated with the
   instruction of the DNA extraction kit (TIANGEN, Beijing, China). The
   purity and concentration of DNA were confirmed by NanoDrop. For library
   construction, DNA was sonicated to an average length of 300 bp. Then,
   the fragmented DNA was end‐repaired, tailed with a single “A” base, and
   ligated with adaptors. Bisulfite treatment was conducted using the
   ZYMOEZ DNA Methylation‐Gold kit. DNA fragments were recovered and
   amplified to set up the sequencing library, followed by library size
   selection. The qualified libraries were further sequenced and analyzed
   by the Chengdu Life Baseline Technology Company. The Kyoto Encyclopedia
   of Genes and Genomes (KEGG) pathway enrichment analysis
   ([173]https://david.ncifcrf.gov/summary.jsp) was used to analyze the
   DMR gene related signaling pathways.

Pyrosequencing

   To validate promoter hypermethylation in macrophages derived from
   ovariectomized (OVX) mice, pyrosequencing was employed to quantify
   methylation levels. Initially, DNA was extracted as previously
   described, followed by bisulfite conversion. The bisulfite‐converted
   DNA was purified using the Zymo‐Spin IC Column and subsequently
   amplified via PCR with specific primers targeting the CpG sites. For
   PCR amplification, 10–20 ng of bisulfite‐converted DNA was utilized
   with the TaKaRa EpiTap HS kit. The resulting PCR products were analyzed
   using the PyroMark Q96 pyrosequencing and quantification platform,
   following the manufacturer's protocol (Shanghai Biotechnology
   Corporation). A pyrosequencing assay was specifically designed to
   target the JAZF1 promoter region at positions chr6:53068124‐53069624,
   with two pairs of primers developed to assess eight CpG sites.
   Methylation levels were calculated as the percentage of the methylated
   (C) allele relative to the unmethylated (T) allele background, using
   the formula: Methylation level = (mC / (mC + umC)) × 100%. The primers
   for JAZF1 were as follows:

   1) 3 CpGs in the promoter region (sequence
   AATCGAATTTGCTTTGACTTTCATTTCCAAACCTCTACACTTGCCCCTTTCGTCTTTGGAGCAGCACTGCC
   AGACCTGAGCCACAACACGGGGTTCAGGAGGTG):

   PCR Forward 5′‐ AGAGGTTATTTGGTTAATAGTATGGT‐3′

   PCR Reverse 5′‐ TCAATTTCTTTACCCCAAATTCCC‐3′

   Sequencing 5′‐GGTGAAATTTTTTTTTTAAGTGAA‐3.

   2) 5 CpGs in the promoter region (sequence
   TAGCGCGCATAACCAAGGGAGGCTCCAAGCGGACTGCAGGCACCTCACAGGGCGAAAGGGCAGGCCCAGGG
   ACGTAAAGTTGGAGG):

   PCR Forward 5′‐ GTGTTTTTTTGGGGTGAAAAGTA‐3′

   PCR Reverse 5′‐ACAACTCCCCCCTCCAACTT‐3′

   Sequencing 5′‐ATAGATGTGTATTTATTAAGAATG‐3′.

5‐Aza‐2′‐Deoxycytidine Treatment

   Bone marrow derived macrophages (BMDMs) were cultured in phenol red
   free medium with 10% FBS (after activated carbon adsorption treatment)
   and 1% Penicillin–streptomycin solution to simulate an
   estrogen‐deficient environment. Then, cells were treated with DNMT
   inhibitors 5‐Aza‐2′‐deoxycytidine (5‐Aza, 5 µm) or DMSO as control for
   48 h.

SiRNA Transfection

   JAZF1‐siRNA and Negative control (NC) were designed and produced by
   GenePharma Company. The sequences are listed in Table [174]S5
   (Supporting Information). siRNA transfection was performed using
   siRNA‐mate plus Transfection Reagent. Macrophages were cultured
   overnight on 6 cm plates. JAZF1 siRNA or NC was incubated with 51 µL of
   buffer and 5 µL of siRNA (1 OD siRNA diluted with 125 µL DEPC‐treated
   H2O), and then thoroughly mixed with 9 µL of siRNA transfection reagent
   for 5 min at room temperature. Subsequently, the mixture was applied to
   cells on each plate. After 48 h, cells were harvested, and western blot
   analysis was conducted to verify the knockdown efficiency of JAZF1
   using an anti‐JAZF1 antibody (Cat#sc‐376503, Santa Cruz Biotechnology).

Statistical Analysis

   Statistical analyses were conducted via GraphPad Prism (La Jolla, CA,
   USA). The data were tested for normal distribution. Parametric tests
   were conducted on data that had a normal distribution, whereas
   non‐parametric tests were used on data that did not have a normal
   distribution. Comparisons between two groups were calculated by the
   Student's t test. Differences at p <0.05 were considered statistically
   significant.

Conflict of Interest

   The authors declare no conflict of interest.

Author Contributions

   J.T. performed conceptualization; D.L., T.Y., and Y.L. methodology;
   D.L., X.L., C.H., and J.Y. performed investigation; J.T. performed
   supervision; D.L. wrote the original manuscript; J.T. reviewed and
   edited the final manuscript.

Supporting information

   Supporting Information
   [175]ADVS-12-e06121-s001.docx^ (1.6MB, docx)

Acknowledgements