1. Cell Biology
  2. Microbiology and Infectious Disease

Intracellular growth of Chlamydia trachomatis leads to global histone hypermethylation by impairing demethylation

  1. Chloé I Charendoff
  2. Félix V Louchez
  3. Yongzheng Wu
  4. Lee Dolat
  5. Guillaume Velasco
  6. Stéphanie Perrinet
  7. Adrian Gabriel Torres
  8. Laure Blanchet
  9. Magalie Duchateau
  10. Quentin Giai Gianetto
  11. Mariette Matondo
  12. Laurence Del Maestro
  13. Slimane Ait-Si-Ali
  14. Frédéric Bonhomme
  15. Gaël A Millot
  16. Vannary Meas-Yedid
  17. Lluís Ribas de Pouplana
  18. Elisabeth D Martinez
  19. Raphael H Valdivia
  20. Agathe Subtil  Is a corresponding author
  1. Institut Pasteur, Université Paris Cité, UMR3691 CNRS, Cellular biology of microbial infection, France
  2. Sorbonne Université, Collège Doctoral, France
  3. Department of Integrative Immunobiology, Duke University School of Medicine, United States
  4. Université Paris Cité, CNRS, Epigenetics and Cell Fate, UMR7216, France
  5. Institut de Recerca Biomèdica (IRB Barcelona), Spain
  6. Barcelona Institute of Science and Technology (BIST), Spain
  7. Institut Pasteur, Université Paris Cité, CNRS UAR 2024, Proteomics Platform, Mass Spectrometry for Biology, France
  8. Institut Pasteur, Université Paris Cité, Bioinformatics and Biostatistics HUB, France
  9. Institut Pasteur, Université Paris Cité, UMR 3523 CNRS, Epigenetic Chemical Biology, France
  10. Institut Pasteur, Université Paris Cité, UMR3691 CNRS, BioImage Analysis, France
  11. Institució Catalana de Recerca i Estudis Avançats (ICREA), Spain
  12. Department of Pharmacology, UT Southwestern Medical Center, United States
https://doi.org/10.7554/eLife.110111
Share this article
Cite this article
  1. Chloé I Charendoff
  2. Félix V Louchez
  3. Yongzheng Wu
  4. Lee Dolat
  5. Guillaume Velasco
  6. Stéphanie Perrinet
  7. Adrian Gabriel Torres
  8. Laure Blanchet
  9. Magalie Duchateau
  10. Quentin Giai Gianetto
  11. Mariette Matondo
  12. Laurence Del Maestro
  13. Slimane Ait-Si-Ali
  14. Frédéric Bonhomme
  15. Gaël A Millot
  16. Vannary Meas-Yedid
  17. Lluís Ribas de Pouplana
  18. Elisabeth D Martinez
  19. Raphael H Valdivia
  20. Agathe Subtil
(2026)
Intracellular growth of Chlamydia trachomatis leads to global histone hypermethylation by impairing demethylation
eLife 15:e110111.
https://doi.org/10.7554/eLife.110111
  2. Download BibTeX
  3. Download .RIS

Abstract

Chlamydia trachomatis, an intracellular bacterium, hijacks metabolites from the host cell for its own proliferation. We provide evidence of global hypermethylation of the host proteome, including histones during the late stages of infection. Single-cell analyses revealed co-occurrence of several methylated residues on histones while infection did not alter S-adenosylmethionine levels. Histone hypermethylation correlated positively with bacterial load and was prevented by antibiotic treatment. Mapping of trimethylation of histone 3 at residues K4 and K9 revealed a broad distribution throughout chromatin. Nuclear fractions of infected cells exhibited a fourfold decrease in demethylase activity against H3K4me3 and a twofold increase in succinate concentration, a competitive inhibitor for the demethylase co-factor α-ketoglutarate. Supplementation of the culture medium with dimethyl-ketoglutarate (DMKG) or with iron, a second co-factor of histone lysine demethylases, reduced histone hypermethylation. DMKG supplementation modified the transcription of about one third of the infection-responsive genes, indicating that histone hypermethylation contributes to modulating the transcriptional response of the host to infection. Finally, chemical inhibition of histone demethylases in a mouse model of infection showed a moderate benefit regarding the outcome of infection. Overall, our data show that the metabolic pressure exerted by a pathogen with an intracellular lifestyle drives epigenetic changes in infected cells.

Editor's evaluation

This is a valuable study of changes in host genome histone methylation and transcription changes associated with Chlamydia infection. The data presented are solid and collectively support the model that histone hypermethylation is due to inhibition of demethylases. The findings provide insight into how pathogen-induced metabolic changes shape host epigenetic and transcriptional responses.

https://doi.org/10.7554/eLife.110111.sa0

Introduction

Chemical modifications of histones or DNA determine chromatin condensation, DNA accessibility and recruitment of the transcription machinery. Methylation of DNA and histones are the best characterized chromatin modifications. Generally, DNA methylation is associated with chromatin compaction and inhibition of transcription, while histone methylation can either promote or repress transcription depending on the specific histone and lysine residues modified (Luger et al., 2012).

Chromatin dynamics are regulated by cellular metabolism, as several metabolites support chromatin remodeling. All methylation reactions rely on the availability of S-adenosylmethionine (SAM) (Su et al., 2016) and numerous chromatin modifiers use metabolites as cofactors (Boon et al., 2020). The enzymatic activity of many histone lysine demethylases requires the 2-oxoglutarate oxygenase activity of a Jumonji C (JmjC) domain with 2-oxoglutarate (also known as α-ketoglutarate; aKG) and iron II (Fe(II)) needed for the demethylation reaction (Accari and Fisher, 2015). The activities of the histone demethylases are sensitive to physiological variations in the concentration of these metabolites (or of competitive metabolites), and thereby metabolic changes can influence epigenetic modifications (Jiang et al., 2019; Shapiro et al., 2023; Xiao et al., 2012). Overall, the effect of metabolites on the epigenetic status of chromatin are well documented in metabolic diseases and in cancer (Boon et al., 2020; Reid et al., 2017). Metabolites are also at the heart of the influence of the mammalian gut microbiota on host developmental, immunologic, and metabolic outcomes (Abu Kwaik, 2015; Krautkramer et al., 2016; Nichols and Davenport, 2021; Zhao et al., 2021).

The obligate intracellular bacterium Chlamydia trachomatis infects epithelial cells of the genital tract and of the eye conjunctiva (Brunham and Rey-Ladino, 2005). The bacteria undergo a biphasic developmental cycle, which takes place entirely within a vacuolar compartment, called an inclusion, and lasts about 48 h in vitro (Abdelrahman and Belland, 2005). Their metabolic capacity is particularly reduced and several essential metabolites required for bacterial growth are captured from the host cytoplasm (Mehlitz et al., 2017; Stephens et al., 1998). A decade ago, a seminal study revealed that several histone methylation marks were increased in epithelial cells infected with C. trachomatis (Chumduri et al., 2013) but the mechanism underlying this shift in chromatin dynamics during infection is unknown. Furthermore, histone modifications normally associated with opposite outcomes in terms of gene expression, such as H3K4me3 and H3K9me3, were similarly upregulated at the population level, so that the global outcome of these changes on host transcription was not predictable.

Here, we characterized the nature of histone modification during C. trachomatis infection at the single-cell level. We observed that the level of histone methylation was variable between infected cells, and that the proportion of cells that displayed a high level of histone methylation increased with infection time. Strikingly, at the individual cell level, the levels of methylation of different lysine residues strongly correlated with each other. At the proteome level, we saw an overall increase in protein methylation in infected cells. We then undertook to understand the origin of histone hypermethylation in Chlamydia infected cells, the genomic distribution of two of these epigenetic modifications, and whether histone hypermethylation contributed to the transcriptional response of the host to infection.

Results

C. trachomatis infection induces protein hypermethylation including on histones

Several histone methylation modifications (H3K4me3, H3K9me3, H3K27me3, H3K79me2, H4K20me3) were examined by immunofluorescence in cells infected for 48 hr with C. trachomatis serovar LGV L2 constitutively expressing the fluorescent protein mCherry. All modifications displayed a strong signal in several-but not all-infected cells, and not in non-infected cells (Figure 1A, Figure 1—figure supplement 1). Strikingly, most cells positive for one modification were also positive for a second modification. An automated procedure to quantify fluorescence intensities in nuclei confirmed the increase in methylated histone lysine residues in nuclei of infected cells compared to non-infected ones on the same coverslip (Figure 1B). When co-stained residues were examined, a positive correlation between their intensities was observed (Figure 1A). Histone lysine hypermethylation was not restricted to C. trachomatis, as the mouse-adapted strain Chlamydia muridarum also induced histone hypermethylation late in infection in HeLa cells and in mouse embryonic fibroblasts (MEFs) (Figure 1—figure supplement 2). Histone hypermethylation in MEFs was also observed upon infection with C. trachomatis (Figure 1—figure supplement 2).

Figure 1 with 4 supplements see all
Download asset Open asset
C. trachomatis infection induces protein hypermethylation.

(A) Cells were infected at MOI = 1 with mCherry expressing C. trachomatis serovar L2 (blue) for 48 hr before fixation, permeabilization, and staining with the indicated antibodies. DAPI staining (grey) of DNA allows for the detection of host nuclei and bacteria. Bar, 50 μm. Scatter plots display the correlation between mean intensities for the indicated methylated residues in the nuclei of infected cells. a.u., arbitrary unit. R=Pearson’s coefficient of correlation. (B) Mean intensities for the indicated methylated residues were quantified in the nuclei of non-infected and infected cells of the same coverslip. Histograms display the percentage of positive nuclei for the indicated histone methylation mark, defined as nuclei with fluorescence intensity greater than twofold the mean intensity of nuclei of non-infected samples. The p-value of a Welch’s t-test is shown ****p<0.0001, **p<0.01, *p<0.05. (C) Mouse-derived organoids infected for 40 hr with GFP-expressing C. trachomatis serovar L2 (top) or for 30 h with GFP-expressing C. muridarum (bottom) were fixed and stained for H3K9me3 (purple). Arrow heads point to nuclei with increased H3K9me3 intensity. Quantification of the intensity of the signal in nuclei of infected cells and of neighboring non-infected cells is shown. The p-value of a Welch’s t-test is shown. Bar, 20 μm. (D) Volcano plot of the relative enrichment of methylated host proteins in cells infected with C. trachomatis. The site of methylation (on R or K) is indicated for the proteins that are significantly hypomethylated (green quadrant) or hypermethylated (orange quadrant) in infected cells. Proteins or methylated peptides that were recovered in only one condition are absent from this representation (Supplementary file 1). (E) Analysis of the biological pathways associated with proteins differentially methylated late in infection using g:Profiler (Kolberg et al., 2023).

To validate these observations in non-cancer-derived cells, we first used primary epithelial cells derived from the ectocervix of female patients (Tang et al., 2021). The number of nuclei positive for the three histone methylation modifications tested (H3K9me3, H3K27me3, H4K20me3) was higher in the infected cells compared to non-infected cells (Figure 1—figure supplement 3). Finally, we used a mouse endometrial organoid model that recapitulates epithelial cell diversity, polarity and ensuing responses to Chlamydia infection (Dolat and Valdivia, 2021). We observed an increase in H3K9me3 intensity in nuclei of infected cells, compared to the surrounding non-infected cells, when organoids were infected with C. trachomatis or C. muridarum (Figure 1C). These observations confirm the ubiquity of global histone hypermethylation in the late stages of infection.

To understand whether histone post-translational modifications (PTMs) late in infection were specific to histones or the result of a general increase in protein methylation, we applied mass spectrometry-based bottom-up proteomics to assess the level of methylation of proteins in infected versus uninfected cells. Statistical analyses were performed on confidently identified methylated peptides encompassing mono-, di-, or tri-methylations. To control for variations in the relative abundance of proteins because of infection, the changes in quantified methylation values were normalized to the abundance of their cognate proteins, quantified from non-modified peptides. These analyses revealed an increase in protein methylation in infection on arginine and lysine residues (Figure 1D and Supplementary file 1). Despite the inherent challenges associated with the combinatorial nature of post-translational modifications on histones, and the fact that a bulk analysis is less sensitive than single-cell measurements when only a fraction of the infected cells display histone hypermethylation (Figure 1B), we detected hypermethylation of H3K79 in infected samples. In addition, we observed increased methylation in several proteins, including proteins implicated in mRNA synthesis (e.g. SUPT5H, TFIP11, HNRNPAB, HNRNPA2B1) and the cytoskeleton (TUBA3D and KRT18). Overall, gene ontology analysis indicated that mRNA metabolism was the biological pathway that counted the highest proportion of proteins differentially methylated late in infection (Figure 1E). In addition, although the abundance of some of the known methyl transferases and demethylases recovered in the proteomic data varied in abundance between infected and not-infected samples, there was no global trend that could account for the global increase in methylation (Supplementary file 2).

Finally, to assess whether hypermethylation extended to host DNA, we quantified the methylation level of cytosine residues in nuclei from infected vs. non-infected cells by LC-MS/MS analysis. In contrast to what we had observed on histones, host DNA was slightly hypomethylated in infected cells (Figure 1—figure supplement 4). Altogether, we concluded that the hypermethylation of histones during C. trachomatis infection occurred in the context of global protein, but not DNA hypermethylation.

Histone hypermethylation is linked to bacterial metabolic activity but not to the bacterial effector Nue

We next analyzed the kinetics of occurrence of histone hypermethylation using two representative marks, e.g., H3K9me3 and H4K20me3. Histone hypermethylation was rarely observed in the first 24 hr of the infectious cycle, indicating that it occurs in the second half of the cycle, when bacterial load is high (Figure 2A). To further link the level of histone modification with bacterial burden, we used the level of fluorescence of mCherry as a readout for bacterial load. We observed that, within the population of infected cells, the sum intensity of the mCherry signal was higher in cells that displayed hypermethylation of H3K9me3 than in cells with a low level of H3K9me3, indicating that histone hypermethylation correlated with bacterial load (Figure 2B). If high bacterial replication is required for histone hypermethylation, slowing down bacterial growth with the addition of an antibiotic at mid-course should reduce the methylation phenomenon late in infection. We chose to use the fluoroquinolone ciprofloxacin as it does not affect mitochondrial activities in the host, while an antibiotic, such as doxycycline, which impairs the TCA cycle, could have indirect metabolic effects (Riesbeck et al., 1990). Addition of ciprofloxacin in the last 24 hr of infection reduced bacterial load and the level of histone lysine methylation. A similar effect was obtained when treating the cultures with the bacterial protein synthesis inhibitor chloramphenicol (Figure 2C). Altogether, we concluded that histone hypermethylation required sustained bacterial proliferation.

Figure 2 with 1 supplement see all
Download asset Open asset
Histone hypermethylation correlates with bacterial proliferation.

(A) HeLa cells were infected with mCherry-expressing C. trachomatis serovar L2 (MOI = 1) for the indicated time before fixation, permeabilization, and staining for H3K9me3 and H4K20me3. The mCherry signal is displayed in blue, DNA in white. Bar, 50 μm. Dot plots display the mean intensities for the indicated methylated residues in one representative experiment. a.u., arbitrary unit. Histograms display the percentage of positive nuclei for the indicated histone methylation mark, defined as nuclei with fluorescence intensity greater than twofold the mean intensity of nuclei of non-infected cells. Data from three independent experiments and the p-values of Welch’s t-tests are shown ****p<0.0001, **p<0.01, *p<0.05. (B) HeLa cells were infected with C. trachomatis L2 expressing mCherry for 48 h before fixation, permeabilization, and staining for H3K9me3. The mean fluorescence intensity of mCherry of each individual inclusion is shown for infected cells with a low level of H3K9me3 signal (signal < twofold mean H3K9me3 signal in nuclei of non-infected cells) and with a high level of H3K9me3 signal (signal > twofold mean H3K9me3 signal in nuclei of non-infected cells). The p-value of a Welch’s t-test is shown. (C). Cells were infected with mCherry-expressing C. trachomatis serovar L2 and 15 μg/ml ciprofloxacin or 10 μg/ml chloramphenicol was added to the culture medium 24 hpi. Cells were fixed 48 hpi, permeabilized and stained for H3K4me3 and H3K27me3. The dot plots display the mean mCherry intensities in inclusions of control and antibiotic-treated cultures (left) and the relative mean intensities of H3K4me3 (top) and H3K27me3 (bottom) in the nuclei from infected cells and non-infected cells in one representative experiment. Histograms display the percentage of positive nuclei, each dot represents an independent experiment. The p-values of Welch’s t-tests are shown.

We next asked whether Nue, a protein secreted by Chlamydia endowed with histone methyltransferase activity (Pennini et al., 2010), was implicated in this process. We obtained a C. trachomatis strain with a nonsense mutation in nue (C to T mutation at position 129,786 leading to a W71* truncation), designated as CTL2M340 (Kokes et al., 2015), and confirmed the absence of Nue expression by western blot analysis (Figure 2—figure supplement 1). Histone hypermethylation was still observed in cells infected for 48 hr with CTL2M340, demonstrating that Nue is not required for histone hypermethylation observed late in infection (Figure 2—figure supplement 1).

Histone hypermethylation is the result of reduced demethylase activity late in infection

The histone methylation modifications we observed in infected cells are deposited by several types of histone methyltransferases and removed by an equally diverse set of dedicated demethylases (Accari and Fisher, 2015; Husmann and Gozani, 2019; Maiques-Diaz and Somervaille, 2016). Because some of these enzymes use common co-factors, we reasoned that changes in the concentrations of these co-factors due to infection could affect several histone modifications simultaneously. Hypermethylation could be the result of increased methylation or decreased demethylase activity. Since all methylation reactions use SAM as a co-factor, we determined SAM concentrations from isolated host nuclei using a fluorescence transfer-based assay. We observed a 50% decrease in the amount of nuclear SAM in cells infected by C. trachomatis for 48 hr, suggesting that histone hypermethylation is not a consequence of SAM accumulation during infection (Figure 3A). To gain single-cell resolution, we used a ratiometric biosensor designed to measure relative SAM concentration in living cells (Dey et al., 2022). Briefly, cells were transfected with a plasmid coding for a SAM-binding aptamer that produces orange fluorescence in proportion to SAM levels, and an aptamer that produces green fluorescence for normalization purposes (Figure 3B). Transfected cells were infected or not and imaged live 40 hpi. As a control, duplicate wells were treated with cycloleucine, an inhibitor of SAM biogenesis (Dey et al., 2022). Plotting the ratio of orange-to-green fluorescence in transfected cells revealed a strong heterogeneity in SAM levels in HeLa cells, both in the cytoplasm and in the nucleus. Cycloleucine addition decreased the orange-to-green fluorescence ratio significantly, confirming the sensitivity of the probe to SAM concentration. Infected and non-infected cells displayed similar levels of SAM (Figure 3C). Consistently, no correlation between SAM level and bacterial load was observed among infected cells (Figure 3D). Thus, infected cells appear to maintain normal SAM concentration during infection. The decrease in SAM concentration observed in isolated nuclei (Figure 1A) may result from difference in SAM stability and/or consumption between non-infected and infected cells during the isolation procedure. In any case, these data strongly suggest that protein hypermethylation during infection does not result from increased SAM availability.

Figure 3
Download asset Open asset
S-adenosyl methionine (SAM) levels display cell-to-cell heterogeneity and do not increase in infected cells.

(A) Cells were infected with C. trachomatis serovar L2 for 48 hr before isolation of the nuclei. SAM concentrations were normalized to protein content. Each dot represents an independent experiment; the mean ± SD is shown. The p-value of a ratio paired t-test is indicated. (B) Cells transfected with the SAM sensor 5–1 and infected or not for 40 hr with mCherry-expressing C. trachomatis serovar L2 (MOI = 1) were imaged live after addition of BI dihydrochloride and DFHO to the culture medium. SAM binding activates the orange fluorescence by promoting Squash folding and subsequent binding with DFHO. The constitutive green fluorescence of Broccoli–BI is used for normalization. Bar, 30 μm. (C) The orange to green fluorescence ratio was measured in the cytoplasm (left) and in the nucleus (right) of infected and non-infected cells, with or without cycloleucine. The experiment is representative of three; black bars indicate the mean. The p-value of Welch’s t-test is shown. (D) Dot plot of the orange to green fluorescence ratio against the mCherry signal, used as a read-out of bacterial load, in the cytoplasm (left) or nucleus (right) of infected cells. R=Pearson’s coefficient of correlation, showing an absence of correlation between SAM concentration and bacterial load in both compartments.

We thus considered the alternative hypothesis that histone demethylase activity may decrease late in infection. To test it, we first looked at whether a demethylase inhibitor could reproduce the histone hypermethylation phenotype. JIB-04 is a membrane-permeable small inhibitor of all JmjC demethylases (Wang et al., 2013). When the inhibitor was added to the culture medium 24 hr after infection by C. trachomatis, and the cells were fixed and stained 48 hpi, we observed a dose-dependent increase in the H3K9me3-positive cells among the infected cells (Figure 4A). A similar result was obtained for H4K20me3 (Figure 4—figure supplement 1). The increase in histone methylation induced by JIB-04 treatment was much more pronounced in the infected than in the non-infected cells. Note that hypermethylation occurred despite a negative effect of JIB-04 treatment on bacterial development (Figure 4—figure supplement 2). Other demethylase inhibitors (SD70 and TACH101) also increased the methylation of their target histone lysine H3K9, while impairing bacterial replication (Figure 4—figure supplement 3). These data indicate that infected cells are sensitized to the action of inhibitors of JmjC demethylases, consistent with the hypothesis that lysine demethylation by JmjC demethylases is low late in infection. We next tested whether histone hypermethylation could be prevented by overexpression of histone demethylases in a mark-specific manner. KDM4A (also known as JMJD2A) is a demethylase specific for H3K9me3 and H3K36me3, while KDM5A (JARID1A) demethylates H3K4me3 to di- and mono-methylated forms (Iwase et al., 2007; Whetstine et al., 2006). Cells were transfected with HA-tagged demethylases, or catalytically dead enzymes as negative controls and infected 3 hr later with mCherry-expressing C. trachomatis L2. We observed a 60% reduction in the proportion of infected cells that displayed H3K4me3 hypermethylation in cells expressing HA-KDM5A, but not the catalytically dead version of the enzyme. In contrast, the proportion of infected cells positive for H3K9me3 was unchanged upon overexpression of KDM4A. Overexpression of KDM4A had the opposite effect: it reduced the proportion of cells displaying H3K9me3 hypermethylation without affecting H3K4me3 levels. Overexpression of the catalytically dead KDM4A mutant affected neither H3K4me3 nor H3K9me3 levels (Figure 4B). Thus, overexpression of a demethylase was sufficient to partially prevent hypermethylation of its specific target. Finally, we directly measured the demethylase activity in nuclear extracts of infected and non-infected cells against one specific substrate, i.e., trimethylated H3K4. The demethylase activity in nuclear extracts from infected cells was reduced to about 25% of its level in nuclear extracts of non-infected cells, supporting the hypothesis of a global decrease in histone demethylase activity in the host nucleus late in infection (Figure 4C).

Figure 4 with 3 supplements see all
Download asset Open asset
The activities of histone demethylases are low in late C. trachomatis infected cells.

(A) Cells were infected with mCherry-expressing C. trachomatis serovar L2. After 24 hr, the culture medium was refreshed with or without the indicated concentration of JIB-04. Cells were fixed 48 hpi, permeabilized, and stained for H3K9me3 (green). DNA was labeled with DAPI (white), mCherry is displayed in blue. Bar, 50 μm. The dot plot displays H3K9me3 mean fluorescence intensity in nuclei of non-infected or infected cells. a.u., arbitrary unit. Black bars indicate the mean. The p-value of Welch’s t-test are shown. The histogram displays the percentage of positive nuclei in five independent experiments. The p-value of Welch’s t-tests are shown ****p<0.0001; ***p<0.001; **p<0.01; *p<0.05; n.s., not significant. Nuclei were scored positive when H3K9me3 intensity was greater than two times its mean intensity in nuclei of non-infected cells from the same culture condition. (B) Cells were transfected with the indicated HA-tagged constructs (dKDM: catalytically dead mutant) and infected 3 h later with mCherry-expressing C. trachomatis L2. Cells were fixed 48 hr later, permeabilized, and stained for HA and for the indicated methylated residues. DNA was labeled with DAPI (white), mCherry is displayed in blue. Bar, 50 μm. The dot plots display mean fluorescence intensities of the indicated marks in one representative experiment, each dot representing one infected cell. The histograms display the percentages of positive nuclei in the transfected infected cells in four independent experiments. Cells were considered transfected when the HA signal in the nucleus was greater than the twofold value of the mean intensity in nuclei in non-transfected cells of duplicate wells. Nuclei were scored positive as in B. (C) Demethylase activity against H3K4me3 was measured in nuclear extracts from cells infected for 40 h, normalized to protein concentration and expressed relative to the activity measured in nuclear extracts from non-infected cells. The p-value of a Welch’s t-test is indicated.

Supply of demethylase cofactors reduces histone hypermethylation

JmjC demethylases require oxygen, iron (FeII), and aKG, and a limitation of one or more of these three co-factors could impair their activity. An increase in succinate and fumarate levels could also account for the defect in histone demethylase activity, as those are competitive inhibitors of the 2-oxoglutarate oxygenase activity carried out by Jumonji C domains (Xiao et al., 2012). Oxygen can be excluded, as the experiments were performed under normoxic conditions. Interpretation of the measurements of metabolite concentrations in infected cells is obscured by the contribution of metabolites located in the bacteria, which, late in infection, represent a significant fraction of the total metabolites. To circumvent this difficulty and focus on the compartment in which the demethylase activity occurs, we measured succinate concentrations in nuclear fractions, devoid of bacteria. We observed a twofold increase in the concentration of succinate in nuclei isolated 40 hpi compared to control nuclei (Figure 5A). Comforting this finding, we found nearly twice as many succinylated peptides in infected samples as in non-infected samples (Supplementary file 1). This result supports the hypothesis that a decrease in the aKG to succinate ratio during infection contributes to the loss in histone demethylase activity. If iron and/or aKG became limited during infection, supplementation of the culture medium with these metabolites might restrain this phenomenon, provided they are transported to the host cytoplasm. Iron was supplemented with addition of 100 μM ferric ammonium citrate to the culture medium, a concentration sufficient to reverse iron deficiency in vitro (Jiang et al., 2019). For aKG, we supplemented the culture medium with 5 mM dimethyl-ketoglutarate (DMKG), which is membrane permeable and converted into aKG by cytosolic esterases. Interestingly, we observed that changing the medium at mid-course during infection (24 hpi) was sufficient to partially reduce H3K4 and H4K20 tri-methylation at 48 hpi, possibly by replacing rate-limiting metabolites depleted during infection (Figure 5B–C). Supplementation of iron or DMKG reduced the level of histone methylation even further. To control for the possible effect of these treatments on bacterial growth, we measured their impact on the infectious forming units (IFUs) produced within 48 hr. Iron supplementation had no effect on IFUs while DMKG supplementation reduced the infectious progeny by about 25%, and combining iron and DMKG reduced progeny by about 50% (Figure 5—figure supplement 1). The effect of the combination on histone methylation was not tested, as the reduction in bacterial growth could be sufficient on its own to prevent histone methylation. Finally, since infected cells are sensitized to the action of the demethylase inhibitor JIB-04 (Figure 4A) and JIB-04 is a competitor for iron (Wang et al., 2013), addition of iron should counteract the effect of JIB-04, and conversely. Indeed, we observed that addition of iron prevented the methylation of H3K9 upon JIB-04 treatment (Figure 5—figure supplement 2). Addition of iron also abrogated the negative effect of JIB-04 on bacterial growth (Figure 5—figure supplement 2). Altogether, these experiments show that DMKG or iron supplementation of the culture medium diminishes hypermethylation of histone lysine residues late in infection. Together with the observation that succinate levels increase in infection, the data support the hypothesis that the reduction in JmjC histone demethylase activities is driven by an imbalance in the concentration of their co-factors.

Figure 5 with 2 supplements see all
Download asset Open asset
Supply of demethylase cofactors reduces histone hypermethylation.

(A) Succinate concentrations were measured in nuclear extracts from cells infected for 40 hr or not infected and normalized to protein concentrations. The p-value of a ratio paired t-test is indicated. (B) Cells were infected with mCherry-expressing C. trachomatis serovar L2 (MOI = 2). After 24 hr, the culture medium was left untouched (control) or refreshed with the addition of 5 mM DMKG, solvent alone (DMSO), or 100 μM ferric ammonium citrate (iron). Cells were fixed 48 hpi, permeabilized, and stained for H3K4me3 (green) and H4K20me3 (red). DNA was stained with DAPI (displayed in white). Bar, 50 μm. (C) The dot plots show the intensities for the indicated methylated residues in nuclei of infected and non-infected cells for one representative experiment. The second condition (-) corresponds to refreshment of the medium without addition of DMSO. The p-value of Welch’s t-tests are shown. The histograms display the percentages of positive nuclei in five independent experiments. The p-value of Welch’s t-tests are shown.

Hypermethylation on H3K4 and H3K9 spreads throughout chromatin

Does histone hypermethylation occur randomly across chromatin, or are some genomic regions more exposed? To answer this question, we investigated the pattern of chromosomal deposition of two well-studied histone methylation modifications, H3K4me3 and H3K9me3, associated with active and repressed chromatin, respectively, in infected cells by chromatin immunoprecipitation followed by DNA sequencing (ChIP-seq). We observed an increase in the amount of material immunoprecipitated from infected samples compared to non-infected samples, consistent with the increase of these methylation modifications during infection (Figure 6—figure supplement 1). Samples were adjusted to equal DNA concentration before sequencing. Peak calling showed that the number of peaks for H3K4me3 or H3K9me3 were not markedly changed (Figure 6—figure supplement 2). H3K4me3 was enriched in active promoters, whereas H3K9me3 was enriched in heterochromatin and regions of low activity, as expected. The patterns of the distribution of the marks across the different genomic regions were globally unchanged between infected and non-infected samples, suggesting that hypermethylation did not accumulate preferentially in specific chromatin regions for these two marks (Figure 6A). Since the two independent experiments showed the expected hierarchical clustering (Figure 6—figure supplement 3) we merged the reads from biological replicate pairs for H3K4me3 and H3K9me3 for further analysis. The density of reads in H3K4me3 peaks diminished in infected samples (Figure 5B, Supplementary file 3) and was accompanied by a slight but significant increase in the median of read coverage all over the genome in the immunoprecipitated fractions and not in the input (Figure 6C). These data are consistent with H3K4 hypermethylation in infected cells occurring outside H3K4me3-enriched chromatin rather than reinforcing existing peaks. In other words, the inhibition of H3K4me3 demethylation increases the proportion of methylated H3K4 outside H3K4me3-enriched areas of chromatin. The median of read coverage all over the genome also increased slightly in the H3K9me3 immunoprecipitated fractions (Figure 6C). Only 20 sites were significantly differentially enriched for H3K9me3 between the two conditions (Supplementary file 3). These data suggest that, while globally increasing late in infection, the pattern of deposition of H3K9me3 remained similar to that of non-infected cells. The similarity between the two samples might also be due to the caveat that ChIP-seq was performed on the whole population, in which only a fraction of cells displays strong histone hypermethylation. Altogether, histone hypermethylation did not appear preferentially in specific chromatin regions, at least for the two marks investigated. The data were consistent with a random reduction of demethylase activity across host chromatin.

Figure 6 with 3 supplements see all
Download asset Open asset
H3K4me3 and H3K9me3 chromatin landscape.

(A) Distribution of genomic sites enriched in H3K4me3 and H3K9me3 marks. The functional annotation was based on the chromatin states established from ChromHMM (ChromHMM track from ENCODE/Broad Institute). The relative abundance in HeLa cells of each chromatin state is shown on the right. (B) Profile of the H3K4me3 ChIP-seq signal (read density) from merged biological replicates, i.e., pileups of the H3K4me3 enrichment relative to chromatin input (subtraction mode, input from IP of BAM files) in non-infected (NI) and infected HeLa cells on genomic loci defined as 3 kb upstream and downstream of annotated Transcriptional Start Site (TSS) of genes (NCBI RefSeq). (C) Boxplots showing read coverage (log2(CPM + 1); CPM, counts per million; bin = 500 bp) from H3K4me3, H3K9me3 ChIP-seq experiments and input in non-infected (NI), and infected HeLa cells. The median and the p-value (Wilcoxon test) is indicated within each boxplot. Outliers are not shown for clarity.

DMKG supplementation modifies the host cell’s transcriptional response to infection

To investigate the consequences of histone hypermethylation on the host transcriptional response to infection, cells were infected with or without supplementing the culture medium with DMKG at 24 hpi to restrain histone hypermethylation. Non-infected cells were treated in parallel. RNA was extracted from these cells 48 hpi and processed for analysis by microarray. The differentially expressed genes (DEGs) were defined with a cut-off of |Fold-increase|>2 and a p-value <0.05. The number of elements differentially up- or down-regulated between any conditions are split evenly (Figure 7A). In non-treated samples, we observed that 1055 genes were differentially regulated by infection (Figure 7A and Supplementary file 4). About one third of those (364) were no longer differentially expressed when DMKG was added to the culture medium, indicating that increasing intracellular levels of aKG during infection impacts the host cell’s transcriptional response to infection (Figure 7B).

Figure 7
Download asset Open asset
Dimethyl-ketoglutarate (DMKG) supplementation modifies the host transcriptional response to infection.

(A) Transcripts from infected (INF) and non-infected (NI) samples, supplemented or not with 5 mM DMKG 24 hpi, and extracted 48 hpi were analyzed by microarray. Two-by-two comparisons of differentially expressed genes are shown (cut-off: twofold difference). (B) Venn diagram showing about one third (364/1055) of the differentially regulated genes upon infection were no longer differentially regulated with DMKG supplementation. (C) For each of the 10 genes tested, the histogram on the left displays the microarray result (p-values were extracted from Supplementary file 5) and the histogram on the right shows the result of the reverse transcription quantitative PCR (RT-qPCR) validation. For RT-qPCR validation, samples were prepared in the same conditions as the microarray samples. Specific sets of primers for the indicated genes were used to measure transcript levels by RT-qPCR normalized to actin transcripts following the ΔΔCt method. Each experiment was performed at least four times (dots). The data are presented as relative mRNA levels compared to uninfected cells and shown as the mean± SD. Statistical comparisons were obtained using contrast analysis after linear modeling, 2-by-2 multiple testing and p-value adjustment (See Methods for details). ****p<0.0001; ***p<0.001; **p<0.01. (D) KEGG pathways associated to the 55 infection responsive genes differentially regulated by DMKG supplementation (Supplementary file 5, analysis with g:Profiler). (E) Cells were infected with C. trachomatis L2 for 48 hpi. Levels of tRNAGly-GCC and tRNAArg-UCG measured by northern blot are expressed relative to 5 S RNA and U6 snRNA. See the Method section for statistical analysis of the data.

Figure 7—source data 1

pdf file containing original northern blots for Figure 7E.

https://cdn.elifesciences.org/articles/110111/elife-110111-fig7-data1-v3.pdf
Figure 7—source data 2

Original files of Figure 7E.

https://cdn.elifesciences.org/articles/110111/elife-110111-fig7-data2-v3.zip

To validate these microarray data, we tested by reverse transcription quantitative PCR (RT-qPCR) the mRNA levels of the 10 coding genes that displayed the strongest response to infection (> twofold difference between infected and non-infected samples) and to DMKG addition (>2.5 fold difference between DMKG-treated and non-treated in infected samples) (Supplementary file 5). RNA was extracted at 48 hpi from infected vs non-infected cells with and without DMKG supplementation at 24 hpi and compared to uninfected controls via RT-qPCR. The RT-qPCR results recapitulated the microarray data with the exception of STEAP4, that was not significantly induced by infection, and SVOPL and RIMKLB, that did not respond significantly to DMKG supplementation (Figure 7C). Thus, the microarray data were validated at 90% for the response to infection (9 genes out of 10) and 78% for the DMKG sensitivity (7 genes out of 9).

Given that the histone methylation can enhance or repress transcription, depending on the methylated residue (Luger et al., 2012), we next analyzed whether histone hypermethylation during infection globally tilted the balance of transcription regulation in one direction or the other. In non-infected cells, DMKG supplementation resulted in differential expression of 267 genes, evenly distributed between transcripts that were upregulated (133) and downregulated (134) (Figure 7A). In infected cells, a similar range of changes was observed, with 314 DEGs, 184 being up, and 130 being down (Figure 7A, Supplementary file 5). Among the 184 DEGs that were upregulated with DMKG supplementation in infection, 32 corresponded to transcripts that were upregulated by infection, and 62 to transcripts that were downregulated by infection (the remaining 90 elements showing less than a twofold difference in transcription upon infection). Therefore, the shift in transcription when comparing infection versus infection with DMKG supplementation does not have a global trend. In other words, DMKG supplementation enhances the transcription of some of the genes that were already upregulated by infection (e.g. DUSP6), while it dampens the transcription of others (e.g. CCL20). The same observations were made on genes downregulated by infection. Thus, overall, the effect of DMKG supplementation on the expression of a given transcript depends on the transcript.

To minimize the intrinsic effect DMKG has on transcription, we excluded from the 314 DEG list (comparison of infected and infected with DMKG supplementation, Figure 7A) those that showed over twofold differential expression upon DMKG treatment alone, reducing the list to 246 DEGs. Among those, 55 correspond to genes that are differentially regulated by infection (Supplementary file 5). Analysis of this set of genes with g:Profiler pointed to a weak enrichment in genes involved in the MAPK signaling pathway (CACNB2, HSPA1A, EFNA1, FGF2, DUSP6, EPHA2) and in cytokine signaling (IL1RL1, CCL20, TNFRSF12A, TNFRSF10D, BMPR2) (Figure 7D).

Finally, besides coding genes, the 246 DEGs contained many non-coding RNAs like pseudogenes, lncRNAs and piRNAs. Non-coding DEGs particularly abundant in the data set were transfer tRNAs (70 DEGs). All tRNAs in this data set were decreased in infected samples compared to non-infected samples, an effect prevented by DMKG supplementation. To confirm the depletion of tRNA late in Chlamydia infection by an independent approach, we extracted total RNAs from uninfected cells and from cells infected for 48 h. Samples were normalized to the 18 S rRNA content and analyzed by northern blot using probes specific for tRNAGly-GCC and tRNAArg-GCU. Probes for U6snRNA and 5 S rRNA were used to normalize the signals (Figure 7E). The quantity of tRNAGly-GCC and tRNAArg-GCU were reduced by about 30% in infected cells compared with non-infected cells, confirming that infection leads to a decrease in tRNA levels. Thus, our data revealed a decrease in tRNA abundance late in infection, which is antagonized by DMKG supplementation.

Administration of JIB-04 shows limited benefit on bacterial load and tissue damage in a mouse model of Chlamydia infection

We have observed that exposure to JIB-04 was detrimental to bacterial development in vitro (Figure 4—figure supplement 2). The same was true for two KDM4 inhibitors tested (Figure 4—figure supplement 3), suggesting that chemical enhancement of histone methylation had a negative impact on bacterial proliferation. To challenge the hypothesis that demethylase inhibitors might impair bacterial development, we used a vaginal Chlamydia infection model with the mouse-adapted strain, C. muridarum. JIB-04 is amenable to in vivo studies, as it showed activity in a mouse model of cancer (Wang et al., 2013). We performed peritoneal injection of JIB-04 at the onset of infection (Figure 8A). Vaginal lavages were collected on day 4 post-infection to measure bacterial burden. In line with our in vitro observations, JIB-04 administration tended to lower bacterial burden at day 4 of infection (Figure 8B). The mice were sacrificed on day 35, the genital tracts excised, and their morphology examined. Consistent with a decrease in the bacterial load, we observed a trend towards a reduction of the pathological manifestations of infection in JIB-04 treated animals, with fewer cases of hydrosalpinx (Figure 8C) and less dilatation in the uterine horns (Figure 8D). However, the benefit from JIB-04 treatment was limited, since severe hydrosalpinx was still observed in two of the six JIB-04 treated mice.

Figure 8
Download asset Open asset
JIB-04 administration shows limited benefit in vivo.

(A) Schematic view of animal infections. JIB-04 treatment was administered intraperitoneally (i.p.) and infections with C. muridarum were intravaginal. (B) Bacterial load was measured on DNA extracted from vaginal lavages at D4 and normalized to host DNA. Each dot represents one mouse; the p-value of Welch’s t-test are shown. (C) The genital tracts were excised at D35 and the presence and severity of hydrosalpinx was scored. (D) Example of the segmentation and detection of the skeletons of uterine horns. The radius along the length of the uterine horn was computed starting from the point of convergence of the right and left skeletons (yellow and red lines). Each color trace corresponds to one animal; measures of the left and right horns are displayed in left and right panels, respectively.

Discussion

Infection with an intracellular pathogen is expected to challenge the metabolic capacity of its host cell. Here, we showed that infection of epithelial cells by C. trachomatis caused global protein hypermethylation late in infection, including hypermethylation of histone lysine residues. We provided evidence that histone hypermethylation resulted from the proliferation of bacteria and that low level of activity of host demethylases accounted for this phenomenon. Further investigation of the consequences of histone hypermethylation on the host response to infection drew a complex gene-dependent picture, in line with the observation that hypermethylation led to histone modifications associated with active transcription as much as to histone modifications associated with chromatin silencing. Finally, we provide some initial evidence that drugs targeting the epigenome of the host can alter the outcome of infection in vivo.

Hypermethylation of several histone residues during C. trachomatis infection was previously documented at the cell population level (Chumduri et al., 2013). Our analysis by immunofluorescence showed cell-to-cell variability and combinatorial staining for different modifications revealed that cells that displayed a high methylation level for a given lysine also displayed hypermethylation for other residues. Hypermethylation occurred late in infection, was most prominent in cells with a high bacterial load, and was prevented by antibiotic treatment, indicating that it was linked to bacterial proliferation. We excluded a role for Nue, the predicted histone methyltransferase of C. trachomatis, as hypermethylation was also observed in cells infected by a mutant strain that did not express Nue. A proteomic analysis of infected cells showed that, in addition to histones, several proteins were hypermethylated, consistent with the observation that histone demethylases have multiple substrates (Kuwik et al., 2023). Hypermethylated proteins include several implicated in mRNA synthesis and processing, with potential consequences on host gene expression. Hypermethylation also occurred on cytoplasmic proteins, such as tubulin or keratin 18. For the latter, the possibility of methylation at the site we identified had already been reported (Jang et al., 2019).

The hypermethylation of histones and its correlation with bacterial burden suggested that infection might create an environment favoring methylation reaction over demethylation. Consistent with this premise, we observed that infected cells were sensitized to the histone demethylase inhibitor JIB-04, indicating that demethylase activity is low late in infection. This was confirmed by directly measuring one such activity (on H3K4me3) in nuclei extracts isolated from infected cells. The strong loss of activity (75% reduction at 40 hpi) suggests that the reduction in KDM enzymatic activity may affect a larger proportion of cells than detected with antibodies against histone methylation marks, possibly because of low SAM availability (see below). Several observations support the hypothesis that the metabolic burden of infection is responsible for the global histone hypermethylation (i) histone hypermethylation increased with infection time and positively correlated with bacterial load; (ii) antibiotic treatment reduced histone hypermethylation (iii) histone hypermethylation was partially abolished by providing aKG or Fe(II), cofactors for demethylases of the Jumonji family; and (iv) succinate concentration, a competitor for aKG, doubled in the nuclei of infected cells. Altogether, these data strongly suggest that aKG and Fe(II) become limiting during late infection. Indeed, Fe(II) is captured from the host cytoplasm by the bacteria (Pokorzynski et al., 2017). The host responds by enhancing expression of the transferrin receptor (3.3-fold increase of the transcript, Supplementary files 4 fivefold increase of the protein, Supplementary file 2) but cytosolic and nuclear Fe(II) concentrations might remain at a sub-optimal concentration for demethylase activities. The second metabolite, aKG, represents an entry point into C. trachomatis incomplete Krebs cycle (Mehlitz et al., 2017; Stephens et al., 1998). A prior metabolomic study also detected a global increase in succinate levels in C. trachomatis-infected cells (Rother et al., 2018), in agreement with the twofold increase we measured in purified nuclei. The same study showed that aKG levels remained globally stable while fumarate, another competitive metabolite for aKG-dependent 2-oxoglutarate oxygenase (Xiao et al., 2012) also increased over the course of infection. Thus, an increase in the ratio of succinate and fumarate over aKG, together with low iron level, can account for the decrease in demethylase activity. In agreement with the hypothesis of limiting metabolites, replenishment of the culture medium was sufficient to partially prevent histone hypermethylation.

Is protein hypermethylation also a consequence of increased SAM availability? Measurement of SAM levels at single-cell resolution using a ratiometric probe indicated that SAM levels remained stable throughout infection. Measurement of SAM concentration in isolated nuclei suggested, if anything, a decrease in this compartment. Thus, while Chlamydia hijack SAM (Binet et al., 2011), the host appears to adapt its metabolism to maintain normal SAM levels. Importantly, our measurements revealed the heterogeneity in SAM levels in HeLa cells. Such heterogeneity was also observed in other cell lines (Dey et al., 2022). While we lack single-cell resolution for the abundance of other key metabolites (aKG, succinate/fumarate, iron) we can hypothesize that a certain level of SAM is required to sustain HMT activity, below which proteins would escape hypermethylation. In addition, more SAM might become available late in the infectious cycle, once the bacterial consumption slows down, and be immediately consumed in methylation reactions, without accumulating at detectable levels. Thus, late increase in SAM availability, or decrease in the by-product, S-adenosylhomocysteine (SAH), might also contribute to protein hypermethylation.

Consistent with low SAM levels in the nuclei of infected cells, we observed that DNA was slightly hypomethylated late in infection, as previously reported (Bryan et al., 2021). How can protein hypermethylation and DNA hypomethylation be reconciled, when Ten-eleven-translocation (TET) oxidative demethylases, which initiate DNA demethylation, also use aKG and iron as cofactors? Enhanced expression of GADD45A and GADD45B (3.8-fold and 2.4-fold increase in the transcripts, respectively, Supplementary file 4), which promote DNA demethylation downstream of TET enzymes, might compensate for reduced TET enzyme activity (Li et al., 2015). Also, methylation and demethylation reactions depend on the relative affinity of each enzyme for their co-factors and on their respective rates of activities (Islam et al., 2018). This, together with the difference in turnover between DNA and histone methylation/demethylation events, might contribute to the different outcomes when comparing methylation levels of DNA and histones in infection.

Epigenetic modifications govern the accessibility of chromatin. Formaldehyde-assisted isolation of regulatory elements enrichment followed by sequencing (FAIRE-Seq) experiments had shown that the proportion of all differentially accessible regions mapping to promoter regions was at its highest late in Hep2 cells infected with C. trachomatis, with the majority of regions showing a reduction in chromatin accessibility (Hayward et al., 2020). Keeping in mind the caveat that ChIP-seq data were obtained on the whole cell population, when only a fraction of infected cells display histone hypermethylation, our analyzes show that, at least for H3K4 and H3K9, methylation modifications spread throughout chromatin, suggesting that hypermethylation occurs randomly. Since histone methylation can repress or activate gene transcription depending on the residue modified and its position relative to the genetic element, the overall effect histone hypermethylation might have on host gene expression was unclear. To answer this question, we examined the transcriptional changes in the host response to infection upon DMKG supplementation. Interpretation of the data calls for caution, since DMKG supply only partially prevents histone hypermethylation. In addition, hypermethylation affects several proteins involved in mRNA processing (Figure 1D) and DMKG supplementation might affect transcript levels independently of its effect on histone methylation. DMKG supplementation modified the transcription of about one third of the infection-responsive transcripts, indicating that histone hypermethylation contributes to modulating the transcriptional response of the host to infection. DMKG supply did not globally enhance or dampen the host response to infection, since a similar proportion of transcripts were differentially up- and downregulated. A discrete set of infection-regulated genes responded strongly (>2.5 fold) to DMKG supplementation, including genes involved in the MAPK signaling pathway and in cytokine signaling (Supplementary file 5). Finally, several non-coding RNAs were differentially regulated during infection, in a DMKG-sensitive manner, including several tRNAs. Modification in the activity of Fe(II)/aKG-dependent RNA demethylases might be implicated in this phenomenon, as methylations are among the post-transcriptional modifications regulating tRNA stability (Adami and Bottai, 2020). Interestingly, a drop in host protein synthesis was previously reported in Chlamydia infected cells (Ohmer et al., 2019). The decrease in tRNA levels could contribute to this phenomenon.

Our data converge to link protein hypermethylation in infection to the metabolic pressure exerted by bacterial growth. Does it benefit the host or the bacteria? The question is difficult to answer in the absence of a complex model in which protein methylation levels could be controlled experimentally. In vitro, iron supplementation reduced histone hypermethylation without affecting progeny production, indicating that hypermethylation is not required for the generation of infectious bacteria. Furthermore, it is likely that the bacteria have adapted their growth rate to limit possible deleterious consequences of the metabolic pressure they exert on the host. Interestingly, several demethylase inhibitors hindered bacterial proliferation in vitro, indicating that increasing histone methylation beyond its physiological range was deleterious for the bacteria. We tested the effect of one broad demethylase inhibitor, JIB-04, on the outcome of infection in an animal model. A reduction of bacterial burden upon exposure to JIB-04 was observed, although below statistical significance with only 6 animals per group. A trend towards reduced tissue damage was also observed, but the effect remained modest. Interpretation of these observations calls for caution, as exposition to JIB-04 might affect non-infected cells implicated in bacterial clearance. Thus, the trend to a reduction in bacterial burden and pathology might not be related to the decrease in bacterial proliferation upon treatment with demethylase inhibitors observed in vitro. Still, these data suggest that chemical modifications of the epigenome of the host can impact the outcome of infection.

In conclusion, the effect of metabolites on the epigenetic status of chromatin are documented in metabolic diseases, in cancer, and in the influence exerted by the gut microbiota on host physiology. We show here that the metabolic activity of an intracellular bacterium has the potential to compromise metabolic homeostasis of the host and to weigh on the host response to infection through epigenetic alterations.

Materials and methods

Cells

HeLa cells (ATCC) and mouse embryonic fibroblasts (MEFs) were grown in Dulbecco’s modified Eagle’s medium with Glutamax (DMEM, Invitrogen), supplemented with 10% (v/v) heat-inactivated fetal bovine serum (FBS). Primary epithelial cells were isolated from ectocervix biopsies of female patients after approval by the French Ethical Committee ‘CPP Ile de France 1’ on May 9, 2016. Approval and authorization of the National Data Protection Authority (‘Commission Nationale de l’Informatique et des Libertés’, CNIL) have been obtained for the research protocol, in compliance with the Helsinki principles. Informed consent was obtained from all participants. Primary cells were cultivated in keratinocyte serum-free medium (KSFM, Thermo Fisher Scientific) containing 50 mg/L of bovine pituitary extract (Thermo Fisher Scientific) and 5 μg/L of human recombinant epidermal growth factor (EGF) (Thermo Fisher Scientific) (Tang et al., 2021). They were used between passages 3 and 6. All cell cultures were maintained at 37 °C, in a 5% CO2 atmosphere, and were routinely tested for mycoplasma by PCR.

Bacteria

The C. trachomatis serovar LGV L2 strain 434 was from ATCC. For IFU measurements and for image analysis, bacteria stably transformed with a plasmid expressing the green fluorescent protein (GFP) or mCherry under the control of the incD promoter (L2IncDGFP and L2IncDmCherry) were used, respectively (Agaisse and Derré, 2013). For the experiments with organoids C. trachomatis L2 or C. muridarum CM006 expressing GFP in p2TK2 or pNigg-derived vectors, respectively, were used (Dolat and Valdivia, 2021). C. muridarum was the MoPn strain provided by Scot Ouellette (University of Nebraska). The C. trachomatis serovar LGV L2 strain 434 with an allele in nuE leading to a W71* truncation (C to T mutation at position 129786) was obtained by chemical mutagenesis as previously described (Kokes et al., 2015). The resulting rifamycin-resistant mutant CTL2M340 was back-crossed with a spectinomycin-resistant wild-type L2 strain to give rise to the CTL2M340rs59 recombinant strain (referred to here as Nue*). Nue* contains 16 single nucleotide polymorphisms (SNPs) compared to the parental strain, which are listed in Supplementary file 8. Bacteria were propagated on HeLa cells using standard procedures (Scidmore, 2005).

Immunofluorescence

Request a detailed protocol

Cells (50×103 per well) were seeded in a 24-well-plate on a glass coverslip and infected at MOI = 1 the following day. In some experiments, the medium was changed 24 hpi for complete medium supplemented with ferric ammonium citrate (MP Biomedicals #158040 100 mM stock), or dimethyl α-ketoglutarate (DMKG, Sigma #349631, the 7.18 M preparation was adjusted to 2.5 M stock with DMSO, aliquoted and frozen), or JIB-04 (Sigma #SML0808 5 mM frozen stock in DMSO, intermediate dilution 40 μM in culture medium), or chloramphenicol (Sigma #C0378 34 mg/ml stock in ethanol), or ciprofloxacin (Sigma 17850 0.3 mM stock in culture medium) or SD70 (Xcess Biosciences #M60194-2S, 40 mM stock in DMSO) or TACH101 (MedChemExpress #HY-O0297 and #HY-O0298 1:1, 150 μM stock in DMSO). Cells were fixed 48 hpi unless otherwise indicated in 4% paraformaldehyde (PFA) in 4% sucrose in phosphate-buffer saline (PBS) for 20 min, followed by 10 min in 50 mM NH4Cl in PBS, and then washed in PBS. Coverslips were permeabilized with 0.3% TritonX-100 in PBS for 5 min and washed in PBS. Samples were incubated for 1 hr in primary antibodies diluted in 1 mg/ml bovine serum albumin (BSA) and 0.01% Tween-20 in PBS, washed three times in PBS, and then incubated for 1 hr in secondary antibodies with 0.5 μg/ml 4,6-diamidino-2-phenylindole (DAPI) (Invitrogen). Antibodies used are listed in Supplementary file 6. Samples were washed in PBS and then mounted on slides in Mowiol. Microscopy was performed on an Axio Observer Z1 microscope equipped with an ApoTome module (Zeiss, Germany) and a 20 X Apochromat lens. Images were taken with an ORCA-Flash4.OLT camera (Hamamatsu, Japan) using the ZEN software from Zeiss.

Image analysis and quantification

Request a detailed protocol

Image analysis was performed with Fiji (Schindelin et al., 2012), the different steps described below were automatically performed using a Fiji macro script. Cell nuclei and inclusions appeared in the uv/DAPI channel (DNA stain) while only inclusions appeared in the red/mCherry channel (mCherry-expressing bacteria). A median filter of radius 2 was applied to both channels for noise removal. ‘Triangle’ and ‘Huang’ automatic threshold methods were, respectively, used on mCherry and DAPI channels to create binary masks. The mCherry mask was subtracted from the DAPI mask, resulting in a binary mask corresponding to cell nuclei. The FIJI functions ‘Erode,’ ‘Dilate,’ ‘Fill Holes,’ and ‘Watershed’ were successively applied to the resulting mask for nuclei individualization. The FIJI functions ‘Erode,’ ‘Dilate’ two times and ‘Watershed’ were successively applied to the inclusion mask to individualize inclusions and dilate them by one pixel to create an overlap with the nuclei next to inclusions. Regions of interest (ROI) corresponding to inclusions were defined on the inclusions mask using ‘Analyze Particles’ function with ‘Size’ parameter set to ‘30-Inf.’ ROI corresponding to nuclei were defined on the nuclei mask using ‘Analyze Particles’ function with ‘Size’ parameter set to ‘50–500’ and ‘Exclude on edges’ option on. For each nuclei ROI, the macro checked each inclusion ROI one by one to see if there was an overlap with the nuclei ROI. This was done by selecting the intersection of both ROI (‘AND’ function of the ROI Manager) and by checking if it was not empty (selectionType() > –1). Nuclei were considered as from an infected cell if they were overlapping with at least one inclusion. For Figure 2B, the overlap area was measured, and nuclei were associated with the inclusion they were overlapping the most with. The mean intensity in each channel was measured for each nucleus and inclusion. Finally, for each field, three composite pictures of DAPI and mCherry channels were saved, displaying either inclusions, infected, or non-infected nuclei segmentation for post-processing verification. The mean value of intensity for a given histone mark in non-infected samples was calculated. Nuclei whose intensity was greater than two times this value were counted as positive. Positive nuclei from non-infected cells were manually double-checked and corrected as nuclei from infected cells if they had been misclassified by the automated procedure.

Organoid culture and imaging

Request a detailed protocol

Murine endometrial organoids were generated as previously described from C57BL/6 J (Jackson Laboratory, strain no. 000664) (Dolat and Valdivia, 2021). Bacteria were diluted in phosphate-buffered saline to a final concentration of 5×106 IFU, vortexed for 15 s and pipetted into a borosilicate capillary glass needle. Organoids were microinjected using an Eppendorf FemtoJet 4 x at a steep vertical angle and only punctured once and refed with media containing 10 μg/mL gentamicin 1 hr after injection. At the indicated timepoint, organoids were washed with warm PBS, fixed with 3% warm paraformaldehyde for 20 min, and resuspended in 0.5 mL 0.25% ammonium chloride. Organoids were centrifuged at 600 × g for 5 min, resuspended in PBS containing 2% BSA (Fraction V; Equitech-Bio) and 0.1% Triton X-100 (Sigma), and incubated on a rocker at 25 °C. The organoids were centrifuged at 600 × g for 5 min and incubated with rabbit anti-H3K9me3 (1:50) overnight at 4 °C. Organoids were washed once with 2% BSA, centrifuged at 600 × g for 5 min, and incubated with anti-rabbit conjugated to Alex Fluor 647 (1:1000) and Hoechst (2 μg/mL) for 1.5 hr at 25 °C. Organoids were centrifuged at 600 × g for 5 min, resuspended in Vectashield (Vector Labs; H-1000), pipetted onto a glass slide, and overlaid with a coverslip. Organoids (nine organoids for infections with C. trachomatis and seven for C. muridarum) were imaged using an inverted microscope (TI2-E; Nikon Instruments) equipped with an ORCA Flash 4.0 V3 sCMOS camera (Hamamatsu), SOLA solid-state white illumination (Lumencor), and acquired using a 20 x Plan Apochromatic NA 0.75 air objective with the NIS-Elements Software (Nikon). 3D reconstructions were generated in ImageJ (NIH) and the nuclei were segmented using auto local thresholding and the watershed algorithm. The mean intensity in infected cells and neighboring, uninfected cells was measured, plotted, and analyzed in RStudio. Statistical analysis was performed using a Welch’s t-test.

Proteomic analyses

35 mm dishes were plated with 400,000 HeLa cells and infected or not the following day with C. trachomatis L2 at MOI = 2. Forty-eight hours later, the cells were collected and resuspended in Tris 100 mM, 8 M urea pH 7.5. Biological triplicates were prepared. Proteins were reduced using 5 mM TCEP for 30 min at room temperature. Alkylation of the reduced disulfide bridges was performed using 10 mM iodoacetamide for 30 min at room temperature in the dark. Proteins were then digested in two steps: first with 1 µg r-LysC Mass Spec Grade (Promega) for 4 hr at 30 °C and then samples were diluted to 1.8 M urea with 100 mM Tris HCl pH 8.5 and 1 µg Sequencing Grade Modified Trypsin was added for the second digestion overnight at 37 °C. Proteolysis was stopped by adding formic acid (FA) at a final concentration of 5%. The resulting peptides were desalted on a Stage Tip (Rappsilber et al., 2007) prepared with Empore 3M C18 material (Fisher Scientific). Peptides were eluted using 50% acetonitrile (ACN), and 0.1% formic acid (FA). Peptides were concentrated to dryness and resuspended in 2% ACN/0.1% FA just prior to LC-MS injection.

LC-MS/MS analysis was performed on a Q Exactive Plus Mass Spectrometer (Thermo Fisher Scientific) coupled with a Proxeon EASY-nLC 1200 (Thermo Fisher Scientific). One µg of peptides was injected onto a home-made 32 cm C18 column (1.9 μm particles, 100 Å pore size, ReproSil-Pur Basic C18, Dr. Maisch GmbH, Ammerbuch-Entringen, Germany). Column equilibration and peptide loading were done at 900 bars in buffer A (0.1% FA). Peptides were separated with a multi-step gradient from 3 to 6% buffer B (80% ACN, 0.1% FA) in 5 min, 6 to 31% buffer B in 130 min, and 31 to 62% buffer B in 30 min at a flow rate of 250 nL/min. Column temperature was set to 60 °C. MS data were acquired using Xcalibur software using a data-dependent method. MS scans were acquired at a resolution of 70,000 and MS/MS scans (fixed first mass 100 m/z) at a resolution of 17,500. The AGC target and maximum injection time for the survey scans and the MS/MS scans were set to 3E6, 20 ms and 1E6, 60 ms, respectively. An automatic selection of the 10 most intense precursor ions was activated (Top 10) with a 30 s dynamic exclusion. The isolation window was set to 1.6 m/z and normalized collision energy was fixed at 27 for HCD fragmentation. We used an underfill ratio of 1.0% corresponding to an intensity threshold of 1.7E5. Unassigned precursor ion charge states as well as 1, 7, 8, and >8 charged states were rejected, and peptide match was disabled.

Protein identification

Request a detailed protocol

Acquired Raw data were analyzed using MaxQuant software version 1.6.6.0 using the Andromeda search engine (Tyanova et al., 2016). The MS/MS spectra were searched against the Chlamydia trachomatis serovar L2-434-Bu and the Homo sapiens UniProt reference proteome database (884 and 79,057 entries, respectively). All searches were performed with oxidation of methionine, mono- and di-methylation of lysine and arginine, trimethylation of lysine, and acetylation of lysine and protein N-terminus as variable modifications and cysteine carbamidomethylation as fixed modification. Trypsin was selected as protease allowing for up to two missed cleavages. The minimum peptide length was set to five amino acids and the peptide mass was limited to a maximum of 8000 Da. One unique peptide to the protein group was required for the protein identification. The main search peptide tolerance was set to 4.5 ppm and to 20 ppm for the MS/MS match tolerance. Second peptides were enabled to identify co-fragmentation events and match between runs option was selected with a match time window of 0.7 min over an alignment time window of 20 min. The false discovery rate (FDR) for peptide and protein identification was set to 0.01. The mass spectrometry proteomics data have been deposited in the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD051857.

Statistical analysis

Request a detailed protocol

Statistical analysis of proteomics data was based on the comparison of intensities between biological conditions. Proteins identified as reverse hits or potential contaminants were first removed from the analysis. Intensities associated with at least one unique peptide were kept for further statistics. Additionally, only proteins with at least two intensity values in one of both conditions were kept to ensure a minimum of replicability. Next, intensities were log2-transformed. They were normalized so that the intensities of human proteins are centered in each sample on the mean of the medians of the human protein intensities across all samples. This normalization was carried out to ensure that the intensities in each sample corresponded to a same global level of the host proteins. Proteins without any intensity value in one of both conditions have been considered as proteins quantitatively present in a condition and absent in the other. They have, therefore, been set aside and considered as differentially abundant proteins. Next, missing values of the remaining proteins were imputed using the impute.mle function of the R package imp4p (Gianetto et al., 2020). Statistical testing was conducted using limma t-tests thanks to the R package limma (Ritchie et al., 2015). The FDR control was performed using an adaptive Benjamini–Hochberg procedure on the resulting p-values thanks to the function adjust.p of R package cp4p using the robust method to estimate the proportion of true null hypotheses among the set of statistical tests (Giai Gianetto et al., 2016). The proteins associated to an absolute log2(fold-change) superior to 1 and adjusted p-values inferior to a FDR level of 1% have been considered as significantly differentially abundant proteins.

Statistical analysis of modified peptides

Request a detailed protocol

We searched for mono-, di-, or tri-methylations on lysine or arginine residues, or for succinylation on lysine residues. Some modifications were identified through mass differences in MS1 spectra, without precise localization on the peptide. Others have been pinpointed on their respective peptides using MS/MS spectra with localization probabilities exceeding 0.75 as determined by MaxQuant. For statistical analyses, raw intensities of all identified modified peptides in the experiments were log2-transformed. Subsequently, they were normalized by median centering within the biological conditions (infected or not infected) to account for potential shifts in peptide median signals between conditions. Missing values were then imputed by condition using the impute.mle function from the R package imp4p. Standard t-tests were employed to examine whether the fold-changes of modified peptides significantly differed from 0 between conditions. Furthermore, ANOVA contrast analyses were conducted to identify modified peptides with fold-changes significantly diffcell analyses revealed erent from those of the unmodified proteins to which they belong as explained in Giai Gianetto, 2023.

Western blot

Request a detailed protocol

Wild-type and Nue* bacteria were lysed in urea buffer (8 M urea, 1% SDS (v/v), 150 mM NaCl, 30 mM Tris-HCl pH 8.0), normalized for protein content (BCA Protein assay, Pierce) and lysates were subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to a polyvinylidene difluoride (PVDF) membrane. The membrane was blocked with 1 x PBS containing 5% milk and 0.01% Tween 20. The membranes were then immunoblotted with primary antibodies diluted in the same buffer. Primary antibodies against Nue were previously described (Pennini et al., 2010). Primary antibodies against Hsp60 (GroEL) were obtained by immunizing rabbits with GroEL purified from C. muridarum, kindly provided by Lingling Chen (Illingworth et al., 2017). Secondary antibodies were anti-rabbit-HRP (Invitrogen) conjugated antibodies. Blots were developed using the Western Lightning Chemiluminescence Reagent (GE Healthcare) and imaged using Fusion FX (Vilber).

Nuclei isolation and measure of DNA methylation

Request a detailed protocol

35 mm dishes were seeded with 4×105 HeLa cells. Cells were infected or not at MOI = 2 with C. trachomatis LGV L2. Forty-eight hpi, cells were washed in PBS and detached in PBS-EDTA, centrifuged for 5 min at 400 × g and resuspended in 344 ml of hypotonic buffer (10 mM Tris pH 7.65, 1.5 mM MgCl2, 10 mM KCl, 0.2% NP-40, 0.6 mM phenylmethanesulfonyl (PMSF), 1 X Protease inhibitor cocktail (Sigma)). Volume was adjusted to about 1 ml by adding PBS, cells were then lysed using a Dounce homogenizer (20 up and down), and then 96 μl of sucrose buffer (20 mM Tris-HCl pH 7.65, 15 mM KCl, 60 mM NaCl, 0.34 M sucrose, 150 mM spermine (Merck S4264), 50 mM spermidine (Merck S0266), 0.6 mM PMSF, 1 X protease inhibitor cocktail) was added to the samples. Samples were centrifuged for 5 min at 500 × g, supernatants were discarded, and the pellets were frozen. Genomic DNA was extracted with RNase A and digested with the kit nucleoside digestion mix of NEB (M0649S) according to the manufacturer’s protocol. The reaction was processed for LC-MS/MS analysis. Analysis of total 5-methyl-2′-deoxycytidine (5-mdC) concentrations was performed using a Q Exactive mass spectrometer (Thermo Fisher Scientific), equipped with an electrospray ionization source (H-ESI II Probe) coupled with an Ultimate 3000 RS HPLC (Thermo Fisher Scientific). Digested DNA was injected onto a Thermo Fisher Hypersil Gold aQ chromatography column (100 mm × 2.1 mm, 1.9 µm particle size) heated at 30 °C. The flow rate was set at 0.3 mL/min and run with an isocratic eluent of 1% acetonitrile in water with 0.1% formic acid for 10 min. Parent ions were fragmented in positive ion mode with 10% normalized collision energy in parallel-reaction monitoring (PRM) mode. MS2 resolution was 17,500 with an AGC target of 2e5, a maximum injection time of 50 ms, and an isolation window of 1.0 m/z. The inclusion list contained the following masses: dC (228.1), 5-mdC (242.1). Extracted ion chromatograms of base fragments (±5 ppm) were used for detection and quantification (112.0506 Da for dC; 126.0662 Da for 5-mdC). Calibration curves were previously generated using synthetic standards in the ranges of 0.2–50 pmoles injected for dC and 0.02–10 pmoles for 5-mdC. Results are expressed as a % of total dC.

SAM measurement in nuclei

Request a detailed protocol

For each condition, a 10 cm dish was seeded with 2.5 million HeLa cells. Cells were infected or not at MOI = 2 with C. trachomatis LGV L2. Forty hpi, nuclei were isolated as described above. The samples were divided into eight aliquots and centrifuged at 500 × g for 5 min. The pellet (nuclei) was resuspended in 40 μl of CM buffer from the Mediomics Bridge-It SAM fluorescence assay (Gentaur) (2 aliquots, flash-frozen) or urea buffer (30 mM Tris, 150 mM NaCl, 8 M urea, 1% SDS, pH = 8.0) (1 aliquot) to determine protein concentration using a BCA protein assay kit (Pierce 23225). Aliquots were thawed and normalized to equal protein concentrations in CM buffer, diluted 1:4 in CM buffer, and incubated for 1 hr at 24 °C with vigorous shaking. Samples were centrifuged for 5 min at 16,000 × g. The supernatant was diluted 1:5 in Buffer S, mixed, and 20 μl of the mix was deposited on one Optiplate (Perkin Elmer #6005270) in duplicate. 80 μl of reaction buffer was added to each well. The reaction was incubated for 40 min at room temperature before reading fluorescence resonance energy transfer (FRET) on a FLUOStar Omega with filters (485–12 excitation 540 nm–10 emission, 485–12 excitation 660 nm–10 emission). SAM concentrations in test samples were then determined using a SAM standard curve.

SAM measurement in living cells

Request a detailed protocol

8-well μ-slides (Ibidi #80827) were seeded with 65,000 HeLa cells per well. On the following day, cells were transfected with pAV-U6+27-Tornado-F30-Broccoli-Squash-SAM-sensor 5–1 (Addgene #177920) and infected or not 4 hr later with L2IncDmCherry at MOI = 1. Two days later, 5 μM BI dihydrochloride (Tocris Bioscience #7466, 10 mM stock in DMSO) and 10 μM DFHO (Tocris Bioscience #6434, 10 mM stock in DMSO) were added to the wells. In one control well, 20 mM cycloleucine (Sigma #A48105) was added. Cells were imaged 1 hr later (corresponding to 40 hpi) with a STELLARIS 8 inverted microscope (Leica Microsystems, Mannheim, Germany). A supercontinuum white light laser tunable between 440 and 790 nm was used for excitation and focused through an HC PL APO CS2 40 x, NA 1.2 motCORR water immersion objective. Emission signals were captured with Power HyD S and X detectors. The system was controlled with Leica Application Suite (LAS) X. The excitation wavelength and detection window are the following: DAPI (405 nm; 415–450 nm), Broccoli (460 nm; 475–507 nm); Squash (512 nm; 517–555 nm); and mCherry (587 nm; 598–683 nm). All segmentations were performed using the pretrained Cellpose v2.2.3 model (cyto) with the diameter parameter set to 0. Cells were segmented based on the CellTracker Blue signal, whereas inclusions were segmented from the mCherry channel. The workflow described below was performed semi-automatically using a custom Fiji macro. Regions of interest (ROIs) corresponding to cells and inclusions were generated from the Cellpose masks. Only ROIs corresponding to transfected cells were retained for further analysis and were manually corrected when necessary. Nuclei were segmented as follows. Both the cytoplasm and nuclei were visible in the Broccoli channel, whereas only the cytoplasm was detected in the Squash channel. The Broccoli image was divided by the Squash image to generate a new image in which nuclei were specifically highlighted. A Gaussian blur filter (σ=2) was applied, and a threshold was manually adjusted for each cell to segment its nucleus (or nuclei). To identify infected cells, each cell ROI was sequentially compared with all inclusion ROIs to determine overlap. When a cell overlapped multiple inclusions, it was assigned to the inclusion with the largest overlap area. For each cell, the integrated fluorescence intensities in the Broccoli and Squash channels were measured separately in the nuclei and cytoplasm. For each inclusion, the integrated mCherry fluorescence intensity and area were measured. Broccoli-to-Squash fluorescence intensity ratios were calculated and correlated with inclusion-associated mCherry fluorescence intensities. Data visualization and statistical analyses were performed in RStudio. Differences in fluorescence ratios were assessed using Welch’s t-test, and correlations were evaluated using Pearson’s correlation coefficient (r).

H3K4me3 demethylation activity and succinate measurement

Request a detailed protocol

Ten centimeter dishes were seeded with 1.5 million HeLa cells (2 dishes were used for the infected plates, to compensate for a lower nucleus harvest). Cells were infected or not at MOI = 2 with C. trachomatis LGV L2. Forty hpi, nuclei were isolated as described above and resuspended in 0.3 ml of ice-cold PBS with 1 X Protease inhibitor cocktail. Samples were sonicated three times for 5 s with cooling on ice between sonication steps and centrifuged for 10 min at 16,000 × g at 4 °C. The supernatants were collected and centrifuged a second time for 5 min. The supernatants were immediately used to measure their ability to demethylate H3K4me3 by immune-detection of the reaction product using the Epigentek kit P-3083. Briefly, 5 μl of nuclear extracts was mixed to 45 ul of the manufacturer’s assay buffer JD2 supplemented with 2 μM ascorbate, loaded onto a plate coated with H3K4me3, and incubated for 120 min at 37 °C. The demethylated products were revealed with a specific antibody following the manufacturer’s instructions and fluorometrically measured on a plate reader (Tecan) at 530 nm excitation and 590 nm emission. Background readings were given by heat inactivation of the nuclear extracts prior to the demethylase reaction (15 min at 85 °C). The remainder of the nuclear extracts were flash-frozen and later used to determine protein and succinate concentrations using BCA and Succinate-Glo JmjC Demethylase/Hydroxylase (Promega V7990) assay kits, respectively. Nuclear extracts were heat-inactivated prior to succinate measurement; samples were diluted 1:10 and 25 μL were used. Protein concentrations were used to normalize H3K4me3 demethylase activity and succinate concentrations in nuclear extracts.

Progeny assays

Request a detailed protocol

HeLa cells (105) were seeded in a 24-well plate. The next day cells were infected with L2IncDGFP bacteria at an MOI = 0.15. 24 hr post infection, the medium was replaced with fresh medium supplemented with the indicated drug, or left unchanged, with the addition of DMSO in the well if the drug was dissolved in DMSO. Cells were detached 48 hpi, lysed using glass beads, and the supernatant was used to infect new HeLa cells plated the day before (100,000 cells/well in a 24-well plate), in serial dilution. The next day, three wells per condition with an infection lower than 30% (checked by microscopy) were detached and fixed as described above, before analysis by flow cytometry and determination of the bacterial titer. Acquisition was performed using a CytoFLEX S (Beckman Coulter) and 50,000 events per sample were acquired and then analyzed using FlowJo (version 10.0.7).

Transfections

Request a detailed protocol

One day after being seeded in a 24-well-plate (105 cells/well) in 500 μL of complete medium transfection mix was added to cells 50 μL of Jetprime buffer (Polyplus), 0.5 μg of DNA, and 3 μL of Jetprime transfection reagent (Polyplus) per well. After 3 hr, the medium was changed to a new complete medium containing mCherry-bacteria at MOI = 0.5. Plasmids used were coding for HA-tagged KDM4A/JMJD2A, KDM5A, and dKDM5A (H483A mutation) were obtained from Addgene (#24180, #14800 and #14801, respectively). The plasmid coding for dKDM4A was generated from the wild-type construct by restriction-free cloning (van den Ent and Löwe, 2006) using primers ctcgacggatcaattcaccat and aggtagttgatgctgtagaggtccatgtcttcagtgGCccaagcaaaggatgtcttcc. This resulted in His-to-Alanine mutation in the catalytic site (position 188).

Chromatin immunoprecipitation

Request a detailed protocol

Four 15 cm culture dishes were seeded with 6×106 cells and half were infected the day after at MOI = 2. Forty-eight hours later, cells were rinsed twice in PBS and fixed by adding 15 ml of fixation medium prewarmed to 37 °C 15 mM NaCl, 0.15 mM EDTA, 0.075 mM EGTA, 0.015 mM HEPES in DMEM pH 8.0. Formaldehyde (Sigma F-8775) was added to the prewarmed medium just prior to the fixation step to the final concentration of 1%. Cells were rocked gently at room temperature; afterwards, formaldehyde was quenched by adding glycine drop by drop from a 1.06 M stock prepared fresh to a final concentration of 0.125 M. Quenching was pursued for 5 min with gentle rocking, the cells were rinsed twice with PBS, scraped using a soft silicone scraper, and transferred to a 15 ml conical tubes on ice, pooling the two identical plates together. Cells were centrifuged at 400 × g for 10 min at 4 °C, the pellet was resuspended in 10 ml cold PBS, centrifuged again. The pellet was resuspended in 1 ml of cold PBS and centrifuged, the supernatant was discarded and the pellet was frozen at – 80 °C. Pellets were resuspended in lysis buffer (50 mM HEPES (pH 7.5), 140 mM NaCl, 1 mM EDTA (pH 8), 10% glycerol, 0.5% NP-40, 0.25% Triton X-100), and incubated for 10 min at 4 °C. After centrifugation (300 × g, 5 min, 4 °C) the supernatant was discarded, the pellet resuspended in buffer (200 mM NaCl, 1 mM EDTA (pH 8), 0.5 mM EGTA (pH 8), 10 mM Tris-HCl (pH 8)), and FPIC (Fast Protease Inhibitors Cocktail; Sigma-Aldrich; #S8830-20TAB) and incubated 10 min at 4 °C. After centrifugation the supernatant was discarded and the pellet resuspended in low-SDS buffer (50 mM Tris-HCl (pH 8), 0.1% SDS, 0.95% NP-40, 0.1% NaDoC, 10 mM EDTA (pH 8), 150 mM NaCl, and FPIC) for sonication for 4 min (4 x (30 s ON, 30 s OFF)) (Bioruptor, Diagenode), yielding genomic DNA fragments with a bulk size of 150–500 bp. One percent of chromatin extracts were taken aside for inputs. Chromatin corresponding to 5 μg of DNA was incubated with 2 μg of anti-H3K4me3 (Diagenode; Cat#: C15410003) or anti-H3K9me3 (Diagenode; Cat#: C15410193) antibodies overnight at 4°C. The DNA-protein-antibody complexes were captured by DiaMag protein A-coated magnetic beads (Diagenode; Cat#: C03010020) by incubating at 4 °C for 2 hr. Magnetic beads were then washed with low-salt buffer (20 mM Tris-HCl (pH 8), 0.1% SDS, 1% Triton X-100, 2 mM EDTA (pH 8), 150 mM NaCl, and FPIC) and high-salt buffer (20 mM Tris–HCl (pH 8), 0.1% SDS, 1% Triton X-100, 2 mM EDTA (pH 8), 500 mM NaCl, and FPIC). DNA was eluted with TE buffer (100 mM Tris-HCl (pH 8), 1% SDS, 1 mM EDTA (pH 8)) from the beads overnight, and then reverse cross-linked with 1 μl of RNase A (10 μg/ml; Sigma-Aldrich; Cat#: R4642), followed by Proteinase K digestion (5 μl of 20 mg/mL; Sigma-Aldrich; Cat#: P2308) during 2 hr; inputs corresponding to 1% of IP were reverse cross-linked in the same conditions. DNA was purified using MinElute PCR Purification Kit (Qiagen; Cat#: 28004). Two microliters of input and ChIP were loaded on High Sensitivity D5000 ScreenTape (Agilent Technologies #5067–5592) with 2 μl of reagent (#5067–5593) (Figure 6—figure supplement 1). Libraries were prepared using the MicroPlex Library Preparation Kit v3 (Diagenode) according to the manufacturer’s protocol. In brief, 1–25 ng were used as input material. After adapter ligation, fragments were amplified with Illumina primers for 10 cycles. Libraries were purified using AMPure XP beads protocol (Beckman Coulter, Indianapolis, IN) and quantified using the Qubit fluorometer (Thermo Fisher Scientific). Library size distribution was assessed using the Bioanalyzer High Sensitivity DNA chip (Agilent Technologies). Libraries were normalized and pooled at 2 nM and spike-in PhiX Control v3 (Illumina) was added. Sequencing was done using a NextSeq 500 instrument (Illumina) in paired-end mode (2×40 cycles) and a NextSeq 2000 instrument (Illumina) in paired-end mode (2×50 cycles). After sequencing, a primary analysis based on AOZAN software (ENS, Paris, France) (Perrin et al., 2017) was applied to demultiplex and control the quality of the raw data (based on bcl2fastq v2.20, bcl-convert v4.1.5 and FastQC modules/v0.11.5).

ChIP-seq datasets were analyzed under the Galaxy platform (https://usegalaxy.eu) (Afgan et al., 2022). The quality of raw data has been evaluated with FastQC (Galaxy Version 0.74) (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/). Poor-quality sequences and adapters have been trimmed or removed with Trimmomatic (Galaxy Version 0.23.2) software (Bolger et al., 2014) to retain only paired reads of 40 bp with Phred score >20. Reads were mapped to the reference genome (GRCh38) using Bowtie2 (Galaxy Version 2.5.0) (Langmead et al., 2019) to generate BAM files. Duplicates were removed using the Picard tool MarkDuplicates (Galaxy Version 2.18.2.4). Narrow and broad peaks were called with MACS2 (Galaxy Version 2.2.7.1) (Feng et al., 2012) using the input as a control, FDR ≤0.05. Differential binding analysis of ChIP-Seq peak data was performed using DiffBind (Galaxy Version 2.10.0) (Ross-Innes et al., 2012). Peak annotation was performed using the ChIPseeker R package (Yu et al., 2015) and Bedtools intersect (Galaxy Version 2.28.0) (Quinlan and Hall, 2010) to annotate peaks according to their chromatin feature with genome segmentation data from ENCODE (Dunham et al., 2012; Ernst and Kellis, 2012). The Gene Ontology enrichment analysis was performed with the clusterProfiler R package (Yu et al., 2012). The significant GO categories were identified with Benjamini-Hochberg adjusted p-value ≤0.05. Data can be found under the GEO accession number GSE265792.

Microarray

Request a detailed protocol

A 6-well plate was seeded with 200,000 HeLa cells per well. The following day, cells were infected with C. trachomatis LGV L2 at MOI = 2. 24 hpi, the medium was replaced with complete medium supplemented with 5 mM DMKG or left unchanged with the addition of an equal volume of DMSO. Forty-eight hpi RNA was harvested using the RNeasy kit (Qiagen), and their quality was analyzed using the BioAnalyzer (Agilent). Transcriptome analysis was done using the GeneChip Human Transcriptome Array 2.0 in combination with the GeneChip WT PLUS Reagent kit (Applied Biosystems). 33.3 ng total RNA was used as input for the analysis that was done according to the manufacturer’s guidelines. In brief, double-stranded cDNA is first synthesized using random primers and then used as template for in vitro transcription (IVT) that produces amplified amounts of antisense mRNA (cRNA). The cRNA is used as input for a second round of first-strand cDNA synthesis, producing single-stranded sense cDNA. After fragmentation and end-labeling, the targets are hybridized to single-sample GeneChip cartridge arrays that are stained and washed on a GeneChip Fluidics Station 450 and scanned on a GeneChip scanner 3000 7 G. Cell files containing the probe cell intensity data generated by the scanner are imported into the Expression Console Software 4.0.1 (Thermo Fisher Scientific) for data analysis. Expression values correspond to Robust Multiarray Average (RMA) normalized probe set intensities. Data can be found under the GEO accession number GSE265791.

RT-qPCR and qPCR

Request a detailed protocol

Total RNA from cells in culture were isolated 48 hpi with the RNeasy Mini Kit (Qiagen) with DNase treatment (DNase I, Roche). RNA concentrations were determined with a NanoDrop spectrophotometer (Thermo Fisher Scientific) and normalized to equal contents. Reverse transcription (RT) was performed using the M-MLV Reverse Transcriptase (Promega). Quantitative PCR (qPCR) was undertaken on DNA or on complementary DNA (cDNA) with the LightCycler 480 system using LightCycler 480 SYBR Green Master I (Roche). Data were analyzed using the ΔΔCt method with the actin gene as a control gene (Schmittgen and Livak, 2008). Each RT-qPCR experiment was performed in duplicate and repeated at least three times. Primers used are listed in Supplementary file 7.

tRNA quantification

Request a detailed protocol

tRNAs were extracted and analyzed by northern blot as described (Torres et al., 2019) with the exception that in some cases samples were run on a 6% PAGE (instead of 8%) to get better separation between the 5 S rRNA and the U6 snRNA.

Infection in mice

Request a detailed protocol

The experiments were performed in accordance with the French national and European laws regarding the protection of animals used for experimental and other scientific purposes. The study was approved by the ethics committee of Institut Pasteur (approval N° dap210086). Female C57BL/6 J mice aged 6–7 weeks were purchased from Charles River Laboratories (France) and housed in the animal facility at Institut Pasteur. JIB-04 was synthesized as described (Wang et al., 2013). All animals were treated with 2.5 mg of medroxyprogesterone (Depo-Provera-SC, Pfizer) 7 days prior to infection to synchronize their menstrual cycle. Vaginal lavages were conducted the day prior to infection twice with 50 μl PBS each time by pipetting up and down several times, followed by swabbing to remove the mucus. On day 0, the animals were anesthetized by intramuscular injection of Ketamine/Xylazine in PBS before intraperitoneal (i.p.) injection of JIB-04 (100 mg/kg body weight) resuspended in 300 μl sesame oil (Sigma #3547), control animals were injected with sesame oil only. 2×105 IFUs of C. muridarum resuspended in 5 μl of sucrose-phosphate-glutamic acid buffer (SPG: 10 mM sodium phosphate [8 mM Na2HPO4- 2 mM NaH2PO4], 220 mM sucrose, 0.50 mM l-glutamic acid) were introduced in the vagina of mice. Additional doses of JIB04 or sesame oil were injected i.p. at D2 and D4. A vaginal lavage was performed at D4 and two washes were pooled and frozen immediately in liquid nitrogen. Total RNAs and genomic DNA from lavage fluids were extracted using AllPrep DNA/RNA Micro Kit (Qiagen, #80284). At D35, the animals were sacrificed by cervical dislocation. The genital tracts were excised, imaged, and their morphology examined. Each image was segmented using Icy (https://icy.bioimageanalysis.org/) (de Chaumont et al., 2012) based on color quantization. From the binary image, a distance map and skeleton of the uterine horns were computed using the plugin Morphomath. The distance maps, corresponding to the radius of each uterus horn, were computed starting from the point of convergence of the two uterine horn skeletons.

Statistical analyses

Request a detailed protocol

The R environment v4.1.2 was used for analysis of tRNA quantification and RT-qPCR (R Development Core Team, 2019). Data were not averaged prior to analysis. For tRNA quantification, band intensities were measured in each well of northern blot pictures and the log2 of (band of interest intensity/control band intensity) was computed, with band of interest being either GlyGCC or ArgGCU tRNA, and with control band being either U6 or 5 S rRNA. Then, data were fitted to a mixed linear model that included the condition and the technical replicates as random covariates (i.e. band intensity ~condition + (1|replicate)). For RT-qPCR analysis, a variation of the ΔΔCt method (gene of interest versus actin) (Schmittgen and Livak, 2008) that keeps the technical replicate variations was applied. These values were fitted to a mixed linear model that included the treatment (DMSO, DMKG), the infection (Infected, Not infected) and the technical replicates as random covariates (i.e. ΔΔCt2 ~treatment * infection + (1|replicate) in Figure 5C). Contrasts were performed with the emmeans() function of the emmeans package. Statistical significance was set to p≤0.05. In each panel, type I error was controlled by correcting the p values according to the Benjamini & Hochberg method (‘BH’ option in the p.adjust() function of R). The nature of the statistical tests used in other figures are provided in the respective legend.

Data availability

Microarray data and ChIP-seq data have been deposited in GEO under accession codes GSE265791 and GSE265792 respectively. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD051857.

The following data sets were generated
    1. Charendoff C
    2. Subtil A
    (2026) NCBI Gene Expression Omnibus
    ID GSE265791. Intracellular growth of Chlamydia trachomatis leads to global histone hypermethylation by impairing demethylation.
    https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE265791
    1. Velasco G
    2. Del Maestro L
    3. Ait-Si-Ali S
    (2026) NCBI Gene Expression Omnibus
    ID GSE265792. Intracellular growth of Chlamydia trachomatis leads to global histone hypermethylation by impairing demethylation.
    https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE265792
    1. Duchateau M
    2. Giai Gianetto Q
    3. Matondo M
    4. Perrinet S
    5. Subtil A
    (2026) PRIDE
    ID PXD051857. Metabolic imprint of an intracellular pathogen drives histone hypermethylation and tunes the host transcriptional response to infection.
    https://www.ebi.ac.uk/pride/archive/projects/PXD051857

References

  45. Software
    1. R Development Core Team
    (2019) R: a language and environment for statistical computing
    R Foundation for Statistical Computing, Vienna, Austria.
    http://www.r-project.org

Article and author information

Author details

  1. Chloé I Charendoff

    1. Institut Pasteur, Université Paris Cité, UMR3691 CNRS, Cellular biology of microbial infection, Paris, France
    2. Sorbonne Université, Collège Doctoral, Paris, France
    Contribution
    Conceptualization, Investigation, Visualization, Writing – original draft
    Contributed equally with
    Félix V Louchez
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0009-0007-1980-3822

  2. Félix V Louchez

    1. Institut Pasteur, Université Paris Cité, UMR3691 CNRS, Cellular biology of microbial infection, Paris, France
    2. Sorbonne Université, Collège Doctoral, Paris, France
    Contribution
    Conceptualization, Investigation, Visualization, Writing – original draft
    Contributed equally with
    Chloé I Charendoff
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0009-0001-3800-6614

  3. Yongzheng Wu

    Institut Pasteur, Université Paris Cité, UMR3691 CNRS, Cellular biology of microbial infection, Paris, France
    Contribution
    Investigation, Visualization
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0003-0050-8480

  4. Lee Dolat

    Department of Integrative Immunobiology, Duke University School of Medicine, Durham, United States
    Contribution
    Investigation, Visualization, Writing – original draft
    Competing interests
    No competing interests declared

  5. Guillaume Velasco

    Université Paris Cité, CNRS, Epigenetics and Cell Fate, UMR7216, Paris, France
    Contribution
    Investigation, Visualization, Writing – original draft
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-3620-8810

  6. Stéphanie Perrinet

    Institut Pasteur, Université Paris Cité, UMR3691 CNRS, Cellular biology of microbial infection, Paris, France
    Contribution
    Investigation
    Competing interests
    No competing interests declared

  7. Adrian Gabriel Torres

    1. Institut de Recerca Biomèdica (IRB Barcelona), Barcelona, Catalonia, Spain
    2. Barcelona Institute of Science and Technology (BIST), Barcelona, Catalonia, Spain
    Contribution
    Investigation, Visualization, Writing – original draft
    Competing interests
    No competing interests declared

  8. Laure Blanchet

    1. Institut Pasteur, Université Paris Cité, UMR3691 CNRS, Cellular biology of microbial infection, Paris, France
    2. Sorbonne Université, Collège Doctoral, Paris, France
    Contribution
    Investigation
    Competing interests
    No competing interests declared

  9. Magalie Duchateau

    Institut Pasteur, Université Paris Cité, CNRS UAR 2024, Proteomics Platform, Mass Spectrometry for Biology, Paris, France
    Contribution
    Investigation
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-5475-3065

  10. Quentin Giai Gianetto

    1. Institut Pasteur, Université Paris Cité, CNRS UAR 2024, Proteomics Platform, Mass Spectrometry for Biology, Paris, France
    2. Institut Pasteur, Université Paris Cité, Bioinformatics and Biostatistics HUB, Paris, France
    Contribution
    Formal analysis
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-6815-2498

  11. Mariette Matondo

    Institut Pasteur, Université Paris Cité, CNRS UAR 2024, Proteomics Platform, Mass Spectrometry for Biology, Paris, France
    Contribution
    Supervision, Writing – original draft
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0003-3958-7710

  12. Laurence Del Maestro

    Université Paris Cité, CNRS, Epigenetics and Cell Fate, UMR7216, Paris, France
    Contribution
    Investigation
    Competing interests
    No competing interests declared

  13. Slimane Ait-Si-Ali

    Université Paris Cité, CNRS, Epigenetics and Cell Fate, UMR7216, Paris, France
    Contribution
    Supervision, Writing – review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-8211-9256

  14. Frédéric Bonhomme

    Institut Pasteur, Université Paris Cité, UMR 3523 CNRS, Epigenetic Chemical Biology, Paris, France
    Contribution
    Investigation
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-6797-289X

  15. Gaël A Millot

    Institut Pasteur, Université Paris Cité, Bioinformatics and Biostatistics HUB, Paris, France
    Contribution
    Formal analysis, Writing – original draft
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-0591-3509

  16. Vannary Meas-Yedid

    Institut Pasteur, Université Paris Cité, UMR3691 CNRS, BioImage Analysis, Paris, France
    Contribution
    Formal analysis, Visualization
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-0416-864X

  17. Lluís Ribas de Pouplana

    1. Institut de Recerca Biomèdica (IRB Barcelona), Barcelona, Catalonia, Spain
    2. Barcelona Institute of Science and Technology (BIST), Barcelona, Catalonia, Spain
    3. Institució Catalana de Recerca i Estudis Avançats (ICREA), Barcelona, Catalonia, Spain
    Contribution
    Supervision, Writing – review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-2087-0359

  18. Elisabeth D Martinez

    Department of Pharmacology, UT Southwestern Medical Center, Dallas, United States
    Contribution
    Resources, Writing – review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0003-1131-3689

  19. Raphael H Valdivia

    Department of Integrative Immunobiology, Duke University School of Medicine, Durham, United States
    Contribution
    Resources, Supervision, Writing – review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0003-0961-073X

  20. Agathe Subtil

    Institut Pasteur, Université Paris Cité, UMR3691 CNRS, Cellular biology of microbial infection, Paris, France
    Contribution
    Conceptualization, Formal analysis, Supervision, Funding acquisition, Investigation, Visualization, Writing – original draft
    For correspondence
    asubtil@pasteur.fr
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-7481-4846

Funding

Agence Nationale de la Recherche (20-PAMR-0011 TheraEPI)

  • Chloé I Charendoff
  • Félix V Louchez
  • Yongzheng Wu
  • Laure Blanchet
  • Magalie Duchateau
  • Quentin Giai Gianetto
  • Mariette Matondo
  • Frédéric Bonhomme
  • Agathe Subtil

Institut National du Cancer (INCA_16719)

  • Chloé I Charendoff
  • Félix V Louchez
  • Yongzheng Wu
  • Guillaume Velasco
  • Laure Blanchet
  • Laurence Del Maestro
  • Slimane Ait-Si-Ali
  • Agathe Subtil

Fondation ARC pour la Recherche sur le Cancer

  • Chloé I Charendoff
  • Agathe Subtil

Welch Foundation (I-1878)

  • Elisabeth D Martinez

Centre National de la Recherche Scientifique

  • Stéphanie Perrinet
  • Laure Blanchet
  • Laurence Del Maestro
  • Slimane Ait-Si-Ali
  • Frédéric Bonhomme
  • Agathe Subtil

Institut Pasteur

  • Chloé I Charendoff
  • Félix V Louchez
  • Yongzheng Wu
  • Stéphanie Perrinet
  • Laure Blanchet
  • Magalie Duchateau
  • Mariette Matondo
  • Frédéric Bonhomme
  • Gaël A Millot
  • Vannary Meas-Yedid
  • Agathe Subtil

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

We thank Béatrice Niragire, Klementina Borovnik, and Valentin Petit for technical assistance, Drs. MS Phan and JY Tinevez for the pipeline of analysis of images of infected cells with non-fluorescent inclusions, Dr. J Fernandes for live microscopy, Drs. M Hamon, P Arimondo, S Polo, and MD Magnitov for discussion and reagents. CIC was supported by the Pasteur - Paris University (PPU) International PhD Program. This work was supported by the Agence Nationale pour la Recherche 20-PAMR-0011 TheraEPI, the Fondation ARC pour la recherche sur le Cancer, the Institut national du cancer INCA_16719, the Institut Pasteur, the Centre National de la Recherche Scientifique, the Welch foundation (grant I-1878 to EDM).

Ethics

Primary epithelial cells were isolated from ectocervix biopsies of female patients after approval by the French Ethical Committee 'CPP Ile de France 1' on 9 May 2016. Approval and authorization of the National Data Protection authority ('Commission Nationale de l'Informatique et des Libertés', CNIL) have been obtained for the research protocol, in compliance with the Helsinki principles.

The experiments were performed in accordance with the French national and European laws regarding the protection of animals used for experimental and other scientific purposes. The study was approved by the ethics committee of Institut Pasteur (approval N°dap210086). Every effort was made to minimize suffering.

Copyright

© 2026, Charendoff, Louchez et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 267
    views
  • 58
    downloads
  • 0
    citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Chloé I Charendoff
  2. Félix V Louchez
  3. Yongzheng Wu
  4. Lee Dolat
  5. Guillaume Velasco
  6. Stéphanie Perrinet
  7. Adrian Gabriel Torres
  8. Laure Blanchet
  9. Magalie Duchateau
  10. Quentin Giai Gianetto
  11. Mariette Matondo
  12. Laurence Del Maestro
  13. Slimane Ait-Si-Ali
  14. Frédéric Bonhomme
  15. Gaël A Millot
  16. Vannary Meas-Yedid
  17. Lluís Ribas de Pouplana
  18. Elisabeth D Martinez
  19. Raphael H Valdivia
  20. Agathe Subtil
(2026)
Intracellular growth of Chlamydia trachomatis leads to global histone hypermethylation by impairing demethylation
eLife 15:e110111.
https://doi.org/10.7554/eLife.110111

Share this article

https://doi.org/10.7554/eLife.110111

Categories and tags

Research organism