Histone H3 Lysine 9 Acetylation is Downregulated in GDM Placentas and Calcitriol Supplementation Enhanced This Effect

Despite the ever-rising incidence of Gestational Diabetes Mellitus (GDM) and its implications for long-term health of mothers and offspring, the underlying molecular mechanisms remain to be elucidated. To contribute to this, the present study’s objectives are to conduct a sex-specific analysis of active histone modifications in placentas affected by GDM and to investigate the effect of calcitriol on trophoblast cell’s transcriptional status. The expression of Histone H3 lysine 9 acetylation (H3K9ac) and Histone H3 lysine 4 trimethylation (H3K4me3) was evaluated in 40 control and 40 GDM (20 male and 20 female each) placentas using immunohistochemistry and immunofluorescence. The choriocarcinoma cell line BeWo and primary human villous trophoblast cells were treated with calcitriol (48 h). Thereafter, western blots were used to quantify concentrations of H3K9ac and the transcription factor FOXO1. H3K9ac expression was downregulated in GDM placentas, while H3K4me3 expression was not significantly different. Cell culture experiments showed a slight downregulation of H3K9ac after calcitriol stimulation at the highest concentration. FOXO1 expression showed a dose-dependent increase. Our data supports previous research suggesting that epigenetic dysregulations play a key role in gestational diabetes mellitus. Insufficient transcriptional activity may be part of its pathophysiology and this cannot be rescued by calcitriol.


Introduction
Gestational Diabetes Mellitus (GDM) is defined as glucose intolerance firstly detected during pregnancy [1]. It is the most common pregnancy-related metabolic disorder, affecting up to 14% of pregnancies [2]. Contrary to other obstetric complications, its prevalence has increased steadily over the past decades, making it an urgent issue in pre-and perinatal care [3,4].
The short-term consequences of GDM include a higher risk for preeclampsia (Odds Ratio (OR) 1.81), large for gestational age infants (LGA) (OR 3.43) and C-section (OR 1.46) [5,6]. However, GDM also affects long-term health outcomes of both mothers and offspring. Women who suffered from GDM stand a substantially higher risk of developing diabetes mellitus type 2 (DM2) later in life (relative risk (RR) 7.43) [7]. Their offspring also suffer from a higher risk for developing metabolic syndrome, DM2, adiposity and cardiovascular disease [8][9][10].
The epidemiological connection between vitamin D deficiency and GDM has been discussed, but remains controversial, with several studies suggesting that there is a positive correlation between low vitamin D and insulin resistance or GDM respectively [36][37][38], while others did not find a significant connection [39][40][41]. Recent metanalyses suggest, however, that vitamin D deficiency is in fact a risk factor (OR) for the development of GDM [42,43]. Investigations on GDM placentas have shown an increased expression of VDR compared to control placentas and hypothesised this to be a reaction to low vitamin D levels [44]. Recently, ChiP-seq analyses in cell culture experiments by Meyer, et al. [45] showed that stimulation with Calcitriol (1,25(OH) 2 D 3 ) lead to an enrichment of H3K9ac at promotor regions of VDR regulated genes, suggesting a link between epigenetic modifications and vitamin D.
The aim of this study was to conduct a systematic, sex-specific immunohistochemical and immunofluorescence analysis of active histone modifications in GDM placentas. Furthermore, cell culture experiments were carried out to test the hypothesis that vitamin D stimulation could increase H3K9ac and correspondingly decrease FOXO1 expression in the choriocarcinoma cell line BeWo. Cell culture results were verified using primary human villous trophoblast (HVT) cultures.
We found H3K9ac to be upregulated in syncytiotrophoblast, extra villous trophoblast (EVT), as well as foetal endothelial cells in GDM placentas. The analysis of H3K4me3 did not reveal any significant differences in expression between GDM and control placentas. The stimulation of BeWo cells with human calcitriol resulted in a decrease of H3K9ac at high concentrations and no significant changes at low concentrations. This corresponded with an increase in FOXO1 expression after stimulation. This indicates that the dysregulated H3K9ac expression in GDM cannot be salvaged by vitamin D.

Results
The expression of specific post-translational modifications of histone protein 3 were analysed in placental tissue from 40 heathy (20 of female, 20 of male foetuses) and 40 GDM pregnancies (20 of female, 20 of male foetuses) using immunohistochemistry and double immunofluorescence. The modifications under investigation were H3 lysine 9 acetylation (H3K9ac) and H3 lysine 4 trimethylation (H3K4me3).

H3K9ac Expression is Downregulated in GDM Placentas
There was no sex-specific difference of H3K9ac expression within the control group, nor within the GDM group. Therefore, we grouped male and female data of control and GDM group for the following data analysis. A very strong H3K9ac expression (median immuno reactive score (IRS): 12) was detected in nuclei of villous syncytiothrophoblast cells (SCT), extra villous trophoblast cells (EVT) as well as foetal endothelial cells in control placentas. The boxplot in Figure 1A illustrates that the expression of H3K9ac was significantly downregulated in the syncytiotrophoblast cells of GDM placentas (p < 0.001), with a median IRS of 12 in the control compared to 8 in the GDM placentas. The same pattern was found in EVTs ( Figure 1D-F) and foetal endothelial cells ( Figure 1G-I), where statistical analysis proved the downregulation to be highly significant (p < 0.001). showing the IRS for H3K9ac expression in syncytiotrophoblast (SCT), decidua and foetal endothelial to be highly significantly lower in GDM placentas (p < 0.001). The range between the 25th and 75th percentiles is represented by the boxes with the horizontal line showing median. The bars indicate the 5th and 95th percentiles. Circles indicate values more than 1.5times the boxes' lengths. Pictures showing representative slides for immunohistochemical staining of H3K9ac in the SCT (control (B,b); GDM (C,c)), decidua (control (E,e); GDM (F,f)) and foetal endothelial cells (control (H,h); GDM (I,i)) of according placentas. Arrowheads indicate foetal endothelial cells. Pictures were taken with a 100× lens (capital letters) and 250× lens (lower-case letters) respectively.

No Difference between H3K4me3 Expression in GDM and Control Placentas
In contrast to H3K9ac expression, no statistically significant differences were found in the expression of H3K4me3 (see Figure S1). The median IRS for H3K4me3 expression in the syncytiotrophoblast was 8 in control, as well as GDM placentas (p = 0.853). Similar results were obtained for EVTs. There was however a significant difference in the H3K4me3 expression between male and female control placentas (p = 0.040). showing the IRS for H3K9ac expression in syncytiotrophoblast (SCT), decidua and foetal endothelial to be highly significantly lower in GDM placentas (p < 0.001). The range between the 25th and 75th percentiles is represented by the boxes with the horizontal line showing median. The bars indicate the 5th and 95th percentiles. Circles indicate values more than 1.5-times the boxes' lengths. Pictures showing representative slides for immunohistochemical staining of H3K9ac in the SCT (control (B,b); GDM (C,c)), decidua (control (E,e); GDM (F,f)) and foetal endothelial cells (control (H,h); GDM (I,i)) of according placentas. Arrowheads indicate foetal endothelial cells. Pictures were taken with a 100× lens (capital letters) and 250× lens (lower-case letters) respectively.

No Difference between H3K4me3 Expression in GDM and Control Placentas
In contrast to H3K9ac expression, no statistically significant differences were found in the expression of H3K4me3 (see Figure S1). The median IRS for H3K4me3 expression in the syncytiotrophoblast was 8 in control, as well as GDM placentas (p = 0.853). Similar results were obtained for EVTs. There was however a significant difference in the H3K4me3 expression between male and female control placentas (p = 0.040).

Identification of H3K9ac Expressing Cells by Immunofluorescence Double Staining
Immunofluorescence double staining was carried out using Cytokeratin 7 (CK7) and Cluster of differentiation 31 (CD31) as identifying markers for EVTs and foetal endothelial cells respectively. By using triple filter excitation microscopy, cell phenotypes could be identified ( Figure 2).
Double filter excitation showed CK7 and H3K9ac expression in the same cell, thus confirming the expression of H3K9ac by EVTs. H3K9ac was also expressed by some CK7-negative decidual stroma cells. The H3K9ac expression was more intense in EVT cells of control placentas than GDM placentas. Similarly, double filter excitation showed co-expression of CD31 and H3K9ac, confirming H3K9ac expression by endothelial cells. Again, the H3K9ac expression was more intense in endothelial cells of control placentas than GDM placentas. Immunofluorescence double staining was carried out using Cytokeratin 7 (CK7) and Cluster of differentiation 31 (CD31) as identifying markers for EVTs and foetal endothelial cells respectively. By using triple filter excitation microscopy, cell phenotypes could be identified ( Figure 2).
Double filter excitation showed CK7 and H3K9ac expression in the same cell, thus confirming the expression of H3K9ac by EVTs. H3K9ac was also expressed by some CK7-negative decidual stroma cells. The H3K9ac expression was more intense in EVT cells of control placentas than GDM placentas. Similarly, double filter excitation showed co-expression of CD31 and H3K9ac, confirming H3K9ac expression by endothelial cells. Again, the H3K9ac expression was more intense in endothelial cells of control placentas than GDM placentas.

Figure 2.
Double immunofluorescence phenotyping of placenta cells. H3K9ac, marked with Cy-3labled secondary antibody, is stained red in both rows. CK7 is stained green in the first row, marking EVT cells. CD31 is stained green in the second row, marking endothelial cells. Pictures were taken with a 400× lens.

Downregulation of H3K9ac in Trophoblast Tumour Cells BeWo by Human Calcitriol (Vitamin D)
To gain insight into the possible functional relationship between H3K9ac and FOXO1 expression and human calcitriol (Vit. D) Western Blot analysis was used to quantified H3K9ac expression in BeWo cells after Vit. D stimulation.
Quantitative Western Blot analysis showed that the 48 h in vitro culture with 1.0 µM human calcitriol resulted in a significantly lower H3K9ac expression in BeWo cells ( Figure 3B,C; p = 0.008). However, the stimulation with lower concentrations of human calcitriol (0.1 and 0.01 µM), though lowering H3K9ac expression slightly, did not have a significant influence on the expression of H3K9ac (p = 0.515 for both concentrations).

Upregulation of FOXO1 in Trophoblast Tumour Cells BeWo by Human Calcitriol (Vitamin D)
The expression of FOXO1 in BeWo cells after vitamin D stimulation was found to increase in a dose-dependent manner ( Figure 3A,C). Statistical analysis showed the increased expression to be significant (p = 0.011) even after stimulation with the lowest concentration (0.01 µM). FOXO1 expression increased further with increasing concentrations of the stimulant, still showing statistical significance (p = 0.021).

Downregulation of H3K9ac in Primary Human Villous Trophoblast Cells HVT by Human Calcitriol (Vitamin D)
To solidify results concerning the downregulation of H3K9ac in BeWo cells, the experiment was repeated using human villous trophoblast cells. The stimulation of primary culture human villous trophoblast cells (HVT) with 1.0 µM human calcitriol showed an effect similar to BeWo cells. As seen in Figure 4, the expression of H3K9ac was significantly decreased (p = 0.03) after a 48 h incubation in comparison with the control culture.

Upregulation of FOXO1 in Trophoblast Tumour Cells BeWo by Human Calcitriol (Vitamin D)
The expression of FOXO1 in BeWo cells after vitamin D stimulation was found to increase in a dose-dependent manner ( Figure 3A,C). Statistical analysis showed the increased expression to be significant (p = 0.011) even after stimulation with the lowest concentration (0.01 µM). FOXO1 expression increased further with increasing concentrations of the stimulant, still showing statistical significance (p = 0.021).

Downregulation of H3K9ac in Primary Human Villous Trophoblast Cells HVT by Human Calcitriol (Vitamin D)
To solidify results concerning the downregulation of H3K9ac in BeWo cells, the experiment was repeated using human villous trophoblast cells. The stimulation of primary culture human villous trophoblast cells (HVT) with 1.0 µM human calcitriol showed an effect similar to BeWo cells. As seen in Figure 4, the expression of H3K9ac was significantly decreased (p = 0.03) after a 48 h incubation in comparison with the control culture.

Discussion
Dysregulation of histone modifications is an important factor in the pathophysiology of metabolic diseases and foetal programming, including GDM. The present study provides further evidence of this, as we identified a significant downregulation of H3K9ac in syncytiotrophoblast, EVT and foetal endothelial cells in GDM cases. Additional investigations of H3K4me3 did not show any dysregulation in GDM placentas.
While studies on insulin resistance were able to identify changes in histone modifications on a gene-specific level [46,47], this is the first study to show a pan-placental downregulation of H3K9ac in gestational diabetes mellitus. H3K9ac is an important modification for transcription activity in general and especially relevant for intrauterine development, synzytialisation and angiogenesis [26,48,49]. Animal studies have shown that treatment of somatic cell nuclear transfer embryos with histone deacetylase (HDAC)-inhibitors and the corresponding increase in global H3K9ac levels lead to improved embryo development and blastocyst quality [50]. Thus, the global H3K9ac downregulation found in GDM may indicate insufficient capacity of gene expression. This reduction in transcriptional activity could in turn be linked to foetal complications such as organ immaturity.
By now it has been well established that pregnancy-related diseases-including GDM-show sex-specific differences in terms of pathophysiology and outcome [51,52]. Very recently, Alexander, et al. [53] were able to show sex-specific differences in GDM on an epigenetic level. Thus, we considered a sex-disaggregated collection of data to be of great importance for the present study. The analysis of our data, however, did not show any sex-specific differences in H3K9ac expression.
Research investigating the "cross-talk" between H3K9ac and H3K4me3 indicates that acetylation of H3 tails is dependent on the recognition of and binding to H3K4me3. Thus, H3K4me3 could be seen as a docking basis for further activation of chromatin through H3 acetylation [54]. These findings correspond with the PTM's localizations. As mentioned before, H3K4me3 is an essential part of the pre-initiation complex, forming the basis of transcription start, with transcription factor IID (TFIID) selectively binding to H3K4me3. H3K9ac is part of active enhancers, and as such potentiates transcription activity [55]. Taking this into account our findings may suggest that, while the basis for gene transcription is still intact, the activation of chromatin is in fact impaired in GDM placentas.
Investigating the role of histone modifications in insulin resistance, studies on beta-cells showed that treatment with incretin hormones, substances commonly used in treatment of DM2, leads to a global increase in H3K9ac. This in turn leads to an increase in cAMP response element binding

Discussion
Dysregulation of histone modifications is an important factor in the pathophysiology of metabolic diseases and foetal programming, including GDM. The present study provides further evidence of this, as we identified a significant downregulation of H3K9ac in syncytiotrophoblast, EVT and foetal endothelial cells in GDM cases. Additional investigations of H3K4me3 did not show any dysregulation in GDM placentas.
While studies on insulin resistance were able to identify changes in histone modifications on a gene-specific level [46,47], this is the first study to show a pan-placental downregulation of H3K9ac in gestational diabetes mellitus. H3K9ac is an important modification for transcription activity in general and especially relevant for intrauterine development, synzytialisation and angiogenesis [26,48,49]. Animal studies have shown that treatment of somatic cell nuclear transfer embryos with histone deacetylase (HDAC)-inhibitors and the corresponding increase in global H3K9ac levels lead to improved embryo development and blastocyst quality [50]. Thus, the global H3K9ac downregulation found in GDM may indicate insufficient capacity of gene expression. This reduction in transcriptional activity could in turn be linked to foetal complications such as organ immaturity.
By now it has been well established that pregnancy-related diseases-including GDM-show sex-specific differences in terms of pathophysiology and outcome [51,52].
Very recently, Alexander, et al. [53] were able to show sex-specific differences in GDM on an epigenetic level. Thus, we considered a sex-disaggregated collection of data to be of great importance for the present study. The analysis of our data, however, did not show any sex-specific differences in H3K9ac expression.
Research investigating the "cross-talk" between H3K9ac and H3K4me3 indicates that acetylation of H3 tails is dependent on the recognition of and binding to H3K4me3. Thus, H3K4me3 could be seen as a docking basis for further activation of chromatin through H3 acetylation [54]. These findings correspond with the PTM's localizations. As mentioned before, H3K4me3 is an essential part of the pre-initiation complex, forming the basis of transcription start, with transcription factor IID (TFIID) selectively binding to H3K4me3. H3K9ac is part of active enhancers, and as such potentiates transcription activity [55]. Taking this into account our findings may suggest that, while the basis for gene transcription is still intact, the activation of chromatin is in fact impaired in GDM placentas.
Investigating the role of histone modifications in insulin resistance, studies on beta-cells showed that treatment with incretin hormones, substances commonly used in treatment of DM2, leads to a global increase in H3K9ac. This in turn leads to an increase in cAMP response element binding protein/CREB regulated transcription coactivator 2 (CREB/CRTC2) transcription factor activity [56], indicating the functional relevance of H3K9ac. Since incretin generally improves insulin sensitivity, these findings suggest that low H3K9ac levels may contribute to insulin resistance.
As outlined in the introduction, epidemiological as well as biochemical findings suggest a functional relationship between vitamin D, H3K9ac and FOXO1 with potential relevance for GDM. Thus, we hypothesized that treatment with vitamin D will positively affect H3K9ac (upregulate) and FOXO1 (downregulate). This hypothesis was tested through cell culture experiments using BeWo cells as a trophoblast model and HVT primary cultures to confirm (only H3K9ac). However, the results refuted our hypothesis. H3K9ac expression was not affected by low doses of calcitriol and decreased slightly at the highest concentration. FOXO1 expression increased after stimulation with calcitriol. While these results are consistent within each other, they resemble a state of glucose resistance rather than an improvement. Here it is important to note that stimulation of C2C12 muscle cells with calcitriol did result in a downregulation of FOXO1, suggesting that its role in metabolic control is highly cell type-specific [35]. We have summarized the immunohistochemical and cell culture findings graphically in Figure 5. protein/CREB regulated transcription coactivator 2 (CREB/CRTC2) transcription factor activity [56], indicating the functional relevance of H3K9ac. Since incretin generally improves insulin sensitivity, these findings suggest that low H3K9ac levels may contribute to insulin resistance. As outlined in the introduction, epidemiological as well as biochemical findings suggest a functional relationship between vitamin D, H3K9ac and FOXO1 with potential relevance for GDM. Thus, we hypothesized that treatment with vitamin D will positively affect H3K9ac (upregulate) and FOXO1 (downregulate). This hypothesis was tested through cell culture experiments using BeWo cells as a trophoblast model and HVT primary cultures to confirm (only H3K9ac). However, the results refuted our hypothesis. H3K9ac expression was not affected by low doses of calcitriol and decreased slightly at the highest concentration. FOXO1 expression increased after stimulation with calcitriol. While these results are consistent within each other, they resemble a state of glucose resistance rather than an improvement. Here it is important to note that stimulation of C2C12 muscle cells with calcitriol did result in a downregulation of FOXO1, suggesting that its role in metabolic control is highly cell type-specific [35]. We have summarized the immunohistochemical and cell culture findings graphically in Figure 5. Immunohistochemical analysis showed a downregulation of H3K9ac (red triangle) in GDM placentas, while H3K4me3 (purple hexagons) showed no significant differences (solid pink arrow tail). The H3K9ac reduction may also be part of the aetiology of GDM (pink arrow broken line). Stimulation of trophoblast cells with calcitriol also lead to a reduction of H3K9ac, as well as an increase in FOXO1 (solid orange arrows). This may in turn be linked to the reduction of H3K9ac (yellow arrow broken line), ultimately leading to a downregulation of transcription (green arrows).
Extensive cell culture experiments using a dexamethasone as well as a TNF-induced model of insulin resistance were able to show that both agents induced VDR expression. Furthermore, they were able to show that VDR overexpression resulted in a reduction of insulin-mediated glucose uptake. These findings suggest that VDR is itself a mediator of different pathways of insulin resistance [57]. This may explain why treatment of BeWo cells with calcitriol, leading to an increased VDR activity, was unable to salvage H3K9ac and FOXO1 expression profiles, but rather mimicked a state of GDM.
The fact that we found a dysregulation of H3K9ac not only in SCT and EVT, but also in foetal endothelial cells, may indicate long-lasting effects for the offspring in general and their vascular system in particular. There is evidence that even short-term exposure to hyperglycaemia induces lasting epigenetic changes in vascular cells [58]. These epigenetic changes were causally linked to changes in gene expression, leading to an inflammatory and proatherogenic state [59]. Taking this into account we hypothesis that the downregulation of H3K9ac in foetal endothelial cells may Figure 5. Two potential mechanisms leading to a reduction of H3K9ac in trophoblast cells. Immunohistochemical analysis showed a downregulation of H3K9ac (red triangle) in GDM placentas, while H3K4me3 (purple hexagons) showed no significant differences (solid pink arrow tail). The H3K9ac reduction may also be part of the aetiology of GDM (pink arrow broken line). Stimulation of trophoblast cells with calcitriol also lead to a reduction of H3K9ac, as well as an increase in FOXO1 (solid orange arrows). This may in turn be linked to the reduction of H3K9ac (yellow arrow broken line), ultimately leading to a downregulation of transcription (green arrows).
Extensive cell culture experiments using a dexamethasone as well as a TNF-induced model of insulin resistance were able to show that both agents induced VDR expression. Furthermore, they were able to show that VDR overexpression resulted in a reduction of insulin-mediated glucose uptake. These findings suggest that VDR is itself a mediator of different pathways of insulin resistance [57]. This may explain why treatment of BeWo cells with calcitriol, leading to an increased VDR activity, was unable to salvage H3K9ac and FOXO1 expression profiles, but rather mimicked a state of GDM.
The fact that we found a dysregulation of H3K9ac not only in SCT and EVT, but also in foetal endothelial cells, may indicate long-lasting effects for the offspring in general and their vascular system in particular. There is evidence that even short-term exposure to hyperglycaemia induces lasting epigenetic changes in vascular cells [58]. These epigenetic changes were causally linked to changes in gene expression, leading to an inflammatory and proatherogenic state [59]. Taking this into account we hypothesis that the downregulation of H3K9ac in foetal endothelial cells may contribute to the foetal programming of cardiovascular disease associated with GDM. The role of H3K9ac in "metabolic memory" of the vascular system is controversial. Chen, et al. [60] showed an increase of non-specific H3 acetylation in human umbilical vein endothelial cells after glucose exposure. Furthermore, they found that the overexpression of p300, a histone acetyltransferase (HAT), results in similar expression profiles (e.g., transcription of VEGF-A and fibronectin) as glucose exposure. A global downregulation of H3K9ac was, however, found in endothelium and, in general, in a state of hypoxia [61,62], which in GDM may interact with the state of hyperglycaemia. Thus, further research is needed, concentrating on functional pathways associated with the global downregulation of H3K9ac.
In conclusion, our findings corroborate the growing evidence that epigenetic dysregulations play a key role in gestational diabetes mellitus. Due to the tremendous complexity of epigenetic mechanisms, conclusions concerning functional implications are to be drawn with extreme caution. The pan-placental downregulation of H3K9ac in GDM cases may, however, reflect a downregulation of transcriptional activity. Whether this is cause or effect of the metabolic disorder needs to be investigated further. Our cell culture experiments suggest that treatment with vitamin D is not sufficient to rescue the epigenetic and transcriptional changes in GDM, making this another area where further research is needed urgently.

Tissue Samples
The study design was approved by the Ludwig Maximilians University's ethics committee and written consent was obtained from all participating patients. All participants underwent an oral glucose tolerance test (oGTT) between weeks 24 and 28 of their pregnancy. Using the criteria of the German society for Diabetes Mellitus (capillary whole blood; fasting glucose >90 mg/dL, 1 h > 180 mg/dL, and 2 h > 155 mg/dL) the diagnosis GDM was given when two measurements were above limits. For the study design, 40 patients with GDM and 40 healthy patients (control) were chosen to participate. In each group, foetal gender was balanced. For detailed information on clinical and epidemiological data see Table 1. All GDM patients underwent insulin treatment and were monitored at least once a week at the Diabetes Centre of the Department of Internal Medicine LMU Munich. The study was approved by the ethical committee of the University of Munich (project identification code: 337-06) on the 26th of August 2013 and informed consent was obtained from each patient in written form. Samples and clinical information were anonymized for statistical workup.
Tissue samples (2 × 2 × 2 cm 3 ) of the participants' placentas were obtain directly after birth. The pieces were taken from a cotyledon located the central part of placenta, with sufficient blood supply, aiming to contain decidua, villous as well as extra villous trophoblasts and amniotic epithelia. Areas showing signs of calcification, bleeding or ischemia were excluded from tissue collection. A 24 h incubation period in 4% buffered formalin solution served to fixate the tissue samples, after which they were embedded in paraffin for long-term storage.

Staining
Immunohistochemical staining was carried out in accordance with the recently published protocol by Hutter, et al. [63]. After blocking the endogenous peroxidase using 3% H 2 O 2 , the slides were treated with sodium citrate (pH 6.0) in a high-pressure cooker in order to de-mask all protein epitopes. To prevent any unspecific antigen-antibody interaction, blocking solution was applied (ZytoChem Plus HRP Polymer System, Zytomed Systems GmBH, Berlin, Germany). The slides were then incubated with the primary antibodies (see Table 2) for 16 h at 4 • C. Antigen-antibody interaction was detected by applying the ZytoChem Plus HRP Polymer System (Zytomed Systems GmBH, Berlin, Germany) in accordance with the manufacturer's instructions. Chromogen 3,3'-diaminobenzidine (DAB; Dako, Glostrup, Denmark) was used for staining followed by haemalaun for counterstaining. After dehydration, the slides were cover-slipped. For each experiment, positive and negative control staining was carried out on human colon tissue ( Figure 6).

Double Immunofluorescence
To determine the phenotype of H3K9ac expressing cells in the placenta tissue, double immunofluorescence staining was performed. Placenta tissue of both the control and the GDM group were stained, using CK7 as a marker for EVTs and CD31 as a marker for foetal endothelial cells.
Tissue samples from the same patient collective were used for double immunofluorescence staining as for immunohistochemistry. Pre-treatment of the slides (deparaffinising, blocking of endogenous peroxidase activity, de-masking of protein epitopes) was identical to that used for immunohistochemistry. Blocking solution (Ultra V-Block, Thermo Scientific, Lab Vision, Fremont, CA, USA) was applied for 15 min in order to prevent any unspecific antigen-antibody binding. Thereafter, the primary antibodies were mixed and applied together (see Table 2). Following this, the slides were incubated with the secondary antibodies (see Table 2) for 30 min. The slides were then cover-slipped with minimal light exposure using mounting buffer (Vector Laboratories, Burlingame, CA, USA), which contains DAPI for nuclear counterstaining. The phenotypes were analysed using the fluorescent Axioskop photomicroscope (Zeiss, Oberkochen, Germany). Images were taken with a digital Axiocam camera system (Zeiss, Oberkochen, German).

Cell Culture and Stimulation
The choriocarcinoma cell line BeWo (ECACC, Salisbury, UK) was used as a trophoblast model. Human villous trophoblast cells (HVT) (ScienCell, Carlsbad, CA, USA), a primary cell culture which was kryoconserved at −196 • C, was used to confirm BeWo results. Both cell lines were cultured in DMEM (3.7 g/L NaHCO 3 , 4.5 g/L D-glucose, 1.028 g/L stable glutamine, and Na-Pyruvate; Biochrom, Berlin, Germany) enriched with 10% foetal bovine serum (FBS) at 37 • C and 5% CO 2 . The BeWo as well as the HVT cells were grown in a 12-well plate at a density of 500,000 cells/mL DMEM for western blotting. To ensure adherence the cells were firstly cultured in DMEM with 10% FBS for 4 h, after which it was replaced with fresh DMEM not containing any supplementation. Following the 24 h incubation, the cells were stimulated with 0.01, 0.1 and 1.0 µM of human calcitriol (Sigma-Aldrich, St. Louis, MO, USA), dissolved in ethanol and diluted in DMEM without FBS. Corresponding amounts of ethanol were added to the control cells (see Table 3 for stimulation scheme). Stimulation lasted 48 h.

Westernblotting of Stimulated BeWo Cells and HVT Cell
The cells were treated with 200 µL of lysis buffer for 30 min, consisting of RIPA (Radioimmunoprecipitation assay buffer, Sigma-Aldrich, St. Louis, MO, USA) and protease inhibitor (Sigma-Aldrich, St. Louis, MO, USA) (dilution 1:100). The obtained lysates were centrifuged. To determine the protein concentration a Bradford assay was carried out. The amount of protein chosen for β-Actin and H3K9ac/FOXO1 detection were 5 and 20-25 µg respectively. Firstly, the samples' proteins were separated by molecular weight through sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) and thereafter transferred onto a polyvinylidene fluoride (PVDF) membrane (Millipore, Billerica, MA, USA). Blocking of unspecific background staining was performed by incubating the membrane in 1× Casein solution (VECTASTAIN ABC-AmP Kit for rabbit IgG, Vector Laboratories, Burlingame, CA, USA) for 1 h. The primary antibodies, anti-H3K9ac (monoclonal rabbit IgG, Abcam, Cambridge, UK), anti-FOXO1 (monoclonal mouse IgG, Novus Biologicals Europe, Abingdon, UK) diluted at 1:500 in Casein and anti-β-Actin (monoclonal mouse IgG, Sigma-Aldrich, St. Louise, MO, USA) diluted at 1:1000 in Casein, were added for 16 h at 4 • C, plus an additional 2 × 15 min incubation at room temperature for anti-H3K9ac and anti-FOXO1. Following washing in 1× Casein solution the membrane was incubated with the respective secondary antibodies, biotinylated anti-rabbit-/mouse-IgG (VECTASTAIN ABC-AmP Kit for rabbit/mouse IgG, Vector Laboratories, Burlingame, MA, USA) as instructed by manufacturer's manual. After 20 min treatment with ABC-AmP-reagent (VECTASTAIN ABC-AmP Kit for rabbit IgG, Vector Laboratories, Burlingame, MA, USA), the blots were developed with 5-bromo-4-chloro-3'-indolyphosphate/nitro-blue tetrazolium (BCIP/NBT)-chromogen substrate solution (VECTASTAIN ABC-AmP Kit for rabbit IgG, Vector Laboratories, Burlingame, MA, USA). Blots were detected using the Bio-Rad Universal Hood II (Bio-Rad Laboratories, Hercules, CA, USA) and quantitative analysis was performed using Bio-Rad Quantity One software (Bio-Rad Laboratories, Hercules, CA, USA).

Statistical Analysis
IBM SPSS Statistics for Windows, Version 22.0. Armonk, NY, USA: IBM Corp was used for data collection, data analysis and charts. More specifically, the non-parametric Mann-Whitney-U test was used for comparison of IRS results and the Wilcoxon-signed rank test for analysis of western blot results. P-values smaller than 0.05 were considered statistically significant.