Demethylation in promoter region of severely damaged hepatocytes enhances chemokine receptor CXCR4 gene expression

The liver is known to possess remarkable regenerative potential, but persistent inflammation or severe acute injury can lead to liver fibrosis and incomplete regeneration, ultimately resulting in liver failure. Recent studies have shown that the axis of two types of CXCL12 receptors, CXCR4 and CXCR7, plays a crucial role in liver fibrosis and regeneration. The present study aimed to investigate the regulatory factors involved in CXCR4 expression in injured liver. Immunohistochemical screening of liver tissue samples collected during liver transplantation revealed a reciprocal expression pattern between CXCR4 and MeCP2. An in vitro system involving cultured cell lines and H2O2 treatment was established to study the impact of oxidative stress on signaling pathways and epigenetic alterations that affect CXCR4 mRNA expression. Operating through distinct signaling pathways, H2O2 treatment induced a dose-dependent increase in CXCR4 expression in both hepatocyte- and intrahepatic cholangiocyte-derived cells. Treatment of the cells with trichostatin and azacytidine modulated CXCR4 expression in hepatocytes by modifying the methylation status of CpG dinucleotides located in a pair of TA repeats adjacent to the TATA box of the CXCR4 gene promoter. Only MeCP2 bound to oligonucleotides representing the TATA box region when the cytosine residues within the sequence were methylated, as revealed by electrophoretic mobility shift assay (EMSA). Methylation-specific PCR analysis of microdissected samples revealed a correlation between the loss of CpG methylation and the upregulation of CXCR4 in injured hepatocytes, replicating the findings from the in vitro study. Besides the conventional MEK/ERK and NF-κB signaling pathways that activate CXCR4 in intrahepatic cholangiocytes, the unique epigenetic modifications observed in hepatocytes might also contribute to a shift in the CXCR4–CXCR7 balance towards CXCR4, leading to irreversible liver injury and fibrosis. This study highlights the importance of epigenetic modifications in regulating CXCR4 expression in liver injury and fibrosis.


Introduction
The liver is one of the distinctive organs with a remarkable regenerative capacity.Even in cases of acute and severe injury, compensatory hypertrophy and proliferation of hepatocytes usually contribute to almost complete liver repair (Miyaoka et al. 2012).Despite these highly regenerative capabilities, excessive acute injury or prolonged insult may induce liver fibrosis and hinder hepatocyte regeneration.When the normal program of hepatic regeneration is limited, hepatocytes are surrounded by fibrous septa, forming regenerative nodules, and resulting in an irreversible liver cirrhotic state.Proliferation of pseudo(bile)ducts, termed "ductular reaction" usually occurs around the regenerative nodules as a degraded regeneration process (Nishikawa et al. 2013).
The fibrotic process in the liver is considered to result from the accumulation of extracellular matrices and collagen that are produced mainly by activated hepatic stellate cells (HSCs) located between parenchymal cells and sinusoidal endothelial cells in the spaces of Disse (Nishikawa et al. 1996).Upon liver injury, quiescent vitamin A-storing HSCs are activated by oxidative stress (Nieto 2006) and inflammation (Kisseleva and Brenner 2007), thus transdifferentiating into profibrogenic myofibroblasts (Tsuchida and Friedman 2017).Besides, the presence of both type I collagen (Nishikawa et al. 1996) and tumor necrosis factor (TNF)-α (Nishikawa et al. 2013), an inflammatory cytokine, promotes a ductular reaction.The progression of hepatic fibrosis also promotes imbalanced and altered liver regeneration.
Recently, the direction between regeneration and fibrosis in the liver has been described as being dependent on the balance between two types of CXCL12 receptors: CXCR4 and CXCR7.In response to injury, while the CXCR7 pathway in liver sinusoidal endothelial cells plays an important role toward hepatic regeneration, the predominance of the CXCR4 pathway affects HSC activation and proliferation, leading to fibrosis (Ding et al. 2014;Hong et al. 2009).
In the present study, we carried out immunohistochemical analysis of CXCR4, CXCR7, and their related factors including oxidative stress in resected liver specimens, at liver transplantation for hepatic failure.We also examined epigenetic and biochemical mechanisms activating CXCR4 expression of oxidative stress in hepatocytes and intrahepatic cholangiocytes.

Quantitative RT-PCR
One microgram of the isolated RNA per sample was reverse transcribed to synthesize cDNA, amplified and quantified by the ABI Prism 7300 Sequence Detection Real Time PCR System (Applied Biosystems) with the use of sets of primers and probes (assay ID; CXCR4; Hs00607978_s1, CXCR7; Hs00664172_s1 and GAPDH; Hs00266705_g1).mRNA was quantified relative to that of GAPDH in each sample, according to the manufacturer's protocol.

Methylation-specific PCR (MSP) for CXCR4 promotor region
Histopathological specimens were microdissected and enclosed in low-melting agarose (1.6%) beads.The beads were then treated with proteinase K, followed by bisulfite conversion, as previously described (Nakagawa et al. 2013).Agarose bead fragments were sliced and used directly as a template for methylation-specific PCR (MSP) (Darwanto et al. 2008).The CXCR4 primers used for nested PCR are shown in Supplementary Table 1.The first PCR was carried out under the following conditions: 95 °C for 5 min; 30 cycles at 94 °C for 30 s, at 55 °C for 30 s, at 72 °C for 60 s, and a final extension for 5 min at 72 °C.The second PCR was done at 95 °C for 5 min; 35 cycles at 94 °C for 30 s, at 58 °C for 30 s, at 72 °C for 60 s, and a final extension for 5 min at 72 °C.Each PCR product was separated on a 2% agarose gel by electrophoresis and imaged with Printgraph (ATTO).

Bisulfite sequencing of CXCR4 promotor region
With the use of a T-vector pMD20 cloning kit (cat.no.3270, TaKaRa), MSP products were transfected into DH10B competent cells (cat.no.EC0113, Thermo Fisher Scientific).Single colonies of transformed cells were randomly selected and grown overnight in enrichment broth with ampicillin.Selected recombinant plasmid DNA was extracted for DNA sequencing with an M13F or M13R primer.The sequencing reactions for the cloned PCR products were carried out with a DNA sequencing kit (cat.no.433745, Applied Biosystems) and the reaction products were analyzed on a 310 Genetic Analyzer (cat.no.ABI310, Applied Biosystems).CpG dinucleotides within the sequences were mapped and labeled as methylated when cytosine remained unconverted, and unmethylated when a thymine was in a cytosine position.

Electrophoretic mobility shift assay (EMSA)
Ready-to-use nuclear protein extracts of human cervical cancer cell line HeLa were purchased from Promega (cat.no.E352A).EMSA was conducted with the use of double-stranded oligonucleotides spanning the TA repeat in the human CXCR4 basic promoter region (− 46/− 7).The probes used for EMSA are shown in Supplementary Table 2.The sense probes were 3′-end labeled with biotin (customordered oligonucleotides from Thermo Fisher Scientific).Nuclear extract with or without recombinant TATA binding protein (rhTBP, cat.no.81897, Abcam) was incubated in EMSA binding buffer (cat.no.20148A, Thermo Fisher Scientific) for 10 min on ice.After incubation with specific antibodies against MeCP2 (1:20, cat.no.ab2828, Abcam) for 10 min at room temperature, the binding reaction was carried out by preincubating labeled and annealed oligonucleotides at room temperature for 30 min.For the competition assay, tenfold to 250-fold excess molar amounts of double-stranded oligonucleotides, containing the TA repeat sequence, were added to the binding reaction.Samples were electrophoresed on a 6% DNA Retardation Gel (cat.no.EC63652BOX, invitrogen) at 4 ℃ for 5 min at 150 V, then for 70 min at 100 V.The gel was visualized and analyzed as described for the western blotting experiments.

Statistical analysis
Each value was represented as the mean ± standard deviation.Statistical analyses were carried out by Student's t test.P < 0.05 was considered statistically significant.

Representative histochemical demonstration of liver tissues from normal and hepatic failure
In normal liver tissue, hepatocytes and a central vein are seen in the polygonal-shaped lobule (Fig. 1A a).IHC with the use of Hep par 1 and CK7 clearly distinguished between hepatocytes and bile ducts.A few normal bile ducts at the corners of the hexagon stained positive for CK7 (Fig. 1A b,  c).Here, in a typical hepatic failure, the loss of the nucleus in some hepatocytes was attributed to irreversible cell injury in the regenerative nodule (Fig. 1A d).CK7-positive pseudoducts were also observed proliferating around the Hep par 1-positive regenerative nodule (Fig. 1A e, f).Since IHC examinations confirmed that this was typical hepatic failure, it was used in further IHC and microdissection analyses.
In normal liver tissue, CXCR4 IHC expression was weakly positive in hepatocytes (Fig. 1B a) and a mature bile duct (Fig. 1B b, arrow).On the other hand, tissue from hepatic failure demonstrated that CXCR4 expression was slightly higher than in normal liver tissue in both hepatocytes (Fig. 1B c) and proliferating pseudoducts (Fig. 1B d).In terms of CXCR4, the following antibodies were used to screen factors associated with liver injury, regeneration, and oxidative stress: 8-OHdG, αSMA, AXIN2, CD31, CD34, CD68, CK19, CXCR7, CyclinD1, Endomucin, ERG, Factor-VIII, HNF-1β, Ki-67, MeCP2, MUC1, SOX9, and β-catenin.Only MeCP2, a transcriptional repressor, and CXCR4 showed a reciprocal expression pattern (Fig. 1B e, f), while none of the other antibodies displayed any expression pattern related to CXCR4.IHC findings of only CXCR7, 8-OHdG, and αSMA are depicted in Supplementary Fig. 1a-c (other IHC data are not shown).MeCP2 was observed in the nuclei of normal hepatocytes (Fig. 1B e) and normal bile ducts (Fig. 1B f, arrowheads), and its expression was invariably observed in the background of lymphocytes in both cases, but was selectively diminished in injured hepatocytes (Fig. 1B g) and pseudoducts (Fig. 1B h) in hepatic failure.

Effects of H 2 O 2 treatment on Huh-7 and TKKK cells
After stimulation with 1 mM of H 2 O 2 for 24 h, hepatocytederived Huh-7 and intrahepatic cholangiocyte-derived TKKK cells were detached through morphological alterations (Fig. 2A).To evaluate the effect of oxidative stress, cells from both cell lines were incubated with increasing concentrations of H 2 O 2 .As shown in Fig. 2B, H 2 O 2 enhanced mRNA expression of CXCR4 in a dose-response manner.After 24 h of H 2 O 2 treatment, CXCR4 mRNA expression increased in Huh-7 cells 3.6-and 3.9-fold after stimulation with 500 µM and 1 mM H 2 O 2 , respectively (dark gray bars).In TKKK cells, CXCR4 mRNA also increased 1.9-and 5.4-fold after stimulation with 500 µM and 1 mM H 2 O 2 , respectively, with high variability (light gray bars).Immunocytochemical staining confirmed the upregulation of CXCR4 expression by exposure of Huh-7 and TKKK cells to H 2 O 2 (Supplementary Fig. 3).

Roles of ERK pathway in Huh-7 and TKKK cells
Western blotting revealed that ERK activators induced rapid ERK1/2 phosphorylation.In pERK the increase was transient with a subsequent decrease to near baseline levels by 6 to 24 h (Fig. 4A).After 24 h of serum starvation (0.1% FBS) followed by 6 or 24 h treatment with 1 mM of 1 3 H 2 O 2 , CXCR4 mRNA expression in Huh-7 cells increased 3.1-and 7.3-fold, respectively.Nonetheless, treatment with two ERK activators, C6-ceramide and tBHQ, at any activator dose, did not affect CXCR4 expression.Oxidative stress induced by 1 mM of H 2 O 2 treatment for 6 h did not increase CXCR4 mRNA expression in TKKK cells; after 24 h of treatment; however, the expression increased 3.3-fold.On the other hand, albeit statistically not significant, CXCR4 mRNA increased 1.5-fold after 6 h treatment with C6-ceramide (Fig. 4B, *p < 0.05, n = 3).

Role of epigenetic regulation in expression of CXCR4 gene in Huh-7 and TKKK cells
Quantitative RT-PCR demonstrated a significant increase of CXCR4 mRNA expression in Huh-7 cells treated with 5-Aza and/or TSA but not in TKKK cells (Fig. 5).Immunocytochemical staining confirmed the upregulation of CXCR4 expression induced by 5-Aza and/or TSA treatment in Huh-7 cells (Supplementary Fig. 4).These results suggest a possible involvement of epigenetic regulation in the expression of CXCR4 mRNA.staining in the cytoplasm of viable hepatocytes (e).CK7 is strongly positive in the cytoplasm of proliferated pseudoducts (f).These specimens were used for further immunohistochemical and microdissection analyses.B CXCR4 and methylcytosine binding protein 2 (MeCP2) immunostainings.While the cytoplasm of hepatocytes (a) and bile duct (b, arrow) in the normal case are weakly positive for CXCR4, one of the CXCL12 receptors associated with liver fibrogenesis, positive reaction is seen in the cytoplasm of hepatocytes (c) and pseudoducts (d) in the hepatic failure.CXCR4 staining is also observed in some nuclei of damaged hepatocytes.While MeCP2 immunostaining is constantly observed in the background of lymphocytes in both normal liver and the hepatic failure, its expression is selectively negative in injured hepatocytes (g) and pseudoducts (h) of the hepatic failure (right panel).Scale bars indicate 100 µm (A) and 50 µm (B)

Bisulfite mapping of CXCR4 promoter region in normal liver and hepatic failure
Sections of hepatocytes and pseudoducts in normal liver and hepatic failure were microdissected and enclosed in agarose beads, subjected to bisulfite modification, and used directly as a template for MSP.CpG dinucleotides adjacent to the TA repeat in the CXCR4 gene promoter region were amplified with either unmethylated or methylated specific primers.MSP revealed methylated Fig. 2 Effects of H 2 O 2 treatment on Huh-7 and TKKK cells.A After incubation with 1 mM of H 2 O 2 for 24 h, hepatocyte-derived Huh-7 and intrahepatic cholangiocyte-derived TKKK cells were detached through morphological alterations.B At this stage, quantitative RT-PCR revealed that CXCR4 mRNA increased in Huh-7 cells 3.6-and 3.9-fold by treatment with 500 µM and 1 mM H 2 O 2 , respectively (dark gray bars, *p < 0.05, n = 3).In TKKK, CXCR4 mRNA also increased 1.9-and 5.4-fold by treatment with 500 µM and 1 mM H 2 O 2 , respectively, with high variability (light gray bars).Scale bars indicate 500 µm (A) Fig. 3 Effect of inhibitors MEK/ERK, p38 MAPK, NF-κB, and PI3K/Akt pathways on Huh-7 and TKKK cells.A After 24 h pretreatment with various inhibitors to suppress oxidative stress pathways in both cell lines under low serum concentration (0.1% FBS), CXCR4 mRNA expression at 6 h of H 2 O 2 exposure was measured by quantitative RT-PCR (*p < 0.05, n = 3-6).In Huh-7 cells, inhibitors did not influence the CXCR4 expression 6 h after treatment with 1 mM H 2 O 2 .In TKKK cells, each inhibitor except PD98059 and Akt inhibitor MK-2206 significantly suppressed CXCR4 expression after 6 h.Notably, MEK1/2 inhibitor U0126 reduced CXCR4 expression to one-third (*p < 0.05, n = 3-6).B Western blotting showed that U0126 treatment did not significantly affect ERK1/2 phosphorylation in Huh-7 cells, indicating that H 2 O 2 -induced ERK1/2 phosphorylation occurred MEK1/2-independent manner.In contrast, U0126 nullifies H 2 O 2 -induced upregulation of pERK1/2 in TKKK cells, indicating that H 2 O 2 -induced activation of pERK1/2 is regulated mainly through the MEK/ERK pathway cytosines in normal hepatocytes but not in either the damaged hepatocytes or the pseudoducts in hepatic failure (Fig. 6A).Bisulfite mapping revealed two methylated CpGs within the 5′-flanking regions of the TA repeat in normal hepatocytes (Fig. 6B).The schematic diagram of MSP and bisulfite sequencing data are shown in Supplementary Figs. 5 and 6.In Huh-7 cells, after 24 h of serum starvation (0.1% FBS) followed by 6 or 24 h of 1 mM of H 2 O 2 treatment, CXCR4 mRNA expression increased 3.1-and 7.3-fold, respectively.However, treatment with two ERK activators, C6-ceramide and tertbutylhydroquinone (tBHQ), at any dose, did not affect the CXCR4 expression, indicating that ERK pathway do not play a major role in oxidative stress-induced CXCR4 expression in Huh-7 cells.In TKKK cells, unlike Huh-7 cells, oxidative stress induced by 1 mM of H 2 O 2 treatment did not increase CXCR4 mRNA expression by 6 h, and 3.3fold increase of CXCR4 mRNA expression was observed only after 24 h of treatment.On the other hand, albeit statistically not significant, the CXCR4 expression induced by C6-ceramide (arrow) tended to be higher than that induced by H 2 O 2 at 6 h (*p < 0.05, n = 3) Fig. 5 Roles of epigenetic regulation in the expression of CXCR4 gene in Huh-7 and TKKK cells.After treatment with a DNA methyltransferase inhibitor 5-aza-2′-deoxycytidine (5-Aza) and/or with a histone deacetylase inhibitor trichostatin A (TSA), quantitative RT-PCR demonstrated upregulated expression of CXCR4 in Huh-7 cells.On the other hand, CXCR4 expression was not affected by these treatments in TKKK cells.Taken together, CXCR4 expression in Huh-7 cells is, at least in part, mediated by epigenetic pathways

Effects of CpG methylation of CXCR4 promoter region on binding with MeCP2 and TBP
EMSA showed that the unmethylated oligonucleotides spanning the TA repeat in the CXCR4 gene promoter region exhibited protein-DNA binding (Fig. 7A, black arrow).On the other hand, the methylated oligonucleotides showed another protein-DNA binding (black arrowhead) at a higher position (Fig. 7A, left panel).The protein-DNA binding in unmethylated oligonucleotides (black arrow) was washed out by excess amounts of the cold TA repeat competitor (Fig. 7A, middle panel).Another protein-DNA binding in methylated oligonucleotides (black arrowhead) was blockshifted with an anti-MeCP2 antibody (Fig. 7A, right panel).Figure 7B illustrates MeCP2-mediated epigenetic mechanisms regulating CXCR4 expression.In normal hepatocytes, since the CpG locus adjacent to the TA repeat is methylated and forms a typical MeCP2 binding target, CXCR4 gene expression is suppressed by the disturbed TBP binding to the TA repeat (upper row).Under oxidative stress, however, this CpG locus became unmethylated and restored the TBP binding (lower row).

Assumed regulation mechanism of CXCR4 gene expression under oxidative stress
Among several signaling pathways shown to be activated by oxidative stress, this study suggests that MEK/ERK and NF-κB signaling play central roles in CXCR4 expression in intrahepatic cholangiocytes (Fig. 8).On the other hand, the expression of CXCR4 in hepatocytes is regulated by a unique epigenetic mechanism in that CpG dinucleotides located in the promoter region avoid MeCP2 binding.

Discussion
In this study, to identify factors impeding liver regeneration, the specimens obtained at real-time biopsy from the liver donor's graft were compared through histological analysis with those from the hepatic failure that required liver transplantation.As shown in Fig. 1A a-f, these specimens showed typical histopathological features of normal liver and hepatic failure by HE staining and IHC, including Hep par 1 and CK7.A panel of antibodies was used to screen factors related to liver injury, regeneration, and oxidative stress in the specimens.
Since previous studies have shown that tilting the CXCR4-CXCR7 axis toward CXCR4 interferes with liver regeneration, leading to irreversible liver injury (Ding et al. 2014), the present study focused on factors related to CXCR4 and CXCR7 expression.Chronic liver injury suppresses CXCR7 expression and upregulates CXCR4 expression in liver sinusoidal endothelial cells (Ding et al. 2014); the present study also demonstrated that CXCR7 was strongly positive in normal hepatocytes and mature bile ducts while its expression was considerably attenuated in injured hepatocytes and pseudoducts (Supplementary Fig. 1a).On the other hand, CXCR4 expression was upregulated with an irregular distribution pattern in the injured hepatocytes.Among selected antibodies, the expression of 8-OHdG, a marker of oxidative DNA damage, was observed mainly in the nuclei of injured hepatocytes (Supplementary Fig. 1b).CXCR4 appeared upregulated in hepatocytes damaged by oxidative stress.Conversely, MeCP2 showed a reciprocal expression pattern with CXCR4 (Fig. 1B a-h).Since MeCP2 is an epigenetically relevant factor, the following studies on the regulation of CXCR4 expression incorporate viewpoints regarding oxidative stress and epigenetics.
Fig. 6 Bisulfite mapping of the CXCR4 promoter region in normal and hepatic failure.A Sections of hepatocytes in both cases and pseudoducts in hepatic failure case were selected by microdissection under a light microscope.The microdissected samples were enclosed in agarose beads, subjected to bisulfite modification, and then used directly as a template for methylation-specific PCR (MSP).CpG dinucleotides adjacent to TA repeat in the CXCR4 gene promoter region were amplified with either unmethylated (U) and methylated (M) specific primers.The MSP demonstrated that methylated cytosines seen in normal hepatocytes were not observed in hepatocytes and pseudoducts in hepatic failure.B Bisulfite mapping confirmed that methylated cytosines located within four bases upstream of TA repeat in normal hepatocytes, forming a typical target sequence of methyl-CpG-binding protein 2 (MeCP2) In general, oxidative stress is associated with the pathogenesis of viral (Dikici et al. 2005), drug-induced (Villanueva-Paz et al. 2021), and ischemic liver injury (Douzinas et al. 2012) that causes acute liver failure.It has also been described as affecting the progression of chronic or persistent liver diseases such as hepatitis C (Lozano-Sepulveda 2015), alcohol-related liver disease (Rouach et al. 1992), and nonalcoholic fatty liver disease (García-Ruiz and Fernández-Checa 2018), all of which are known to be leading causes of irreversible liver fibrosis.Therefore, oxidative stress probably plays a crucial role in exacerbating hepatic failure that requires liver transplantation.The present study also inquired into whether or not oxidative stress affects the CXCR4/ CXCR7 axis and, if so, what its precise molecular mechanism entails.Therefore, the effect of oxidative stress on CXCR4 and CXCR7 expression in hepatocytes and  2A), and increased CXCR4 mRNA expression in a dose-dependent manner (Fig. 2B and Supplementary Fig. 3).As defense reactions against oxidative stress, while CXCR7 was upregulated (Supplementary Fig. 2), the CXCR4/ CXCR7 axis tilted toward CXCR4.
To clarify the signaling pathways by which oxidative stress upregulates CXCR4, the effect of MEK/ERK, p38 MAPK, NF-κB, and PI3K/Akt inhibitors (Tan et al. 2020) on CXCR4 mRNA expression was examined.In Huh-7 cells, CXCR4 mRNA expression was not significantly suppressed at 6 h with any of the inhibitors (Fig. 3A, left panel).H 2 O 2 -induced oxidative stress led to ERK1/2 phosphorylation that was not suppressed by MEK1/2 inhibitor U0126, indicating that ERK1/2 phosphorylation is MEK1/2-independent (Fig. 3B, left panel).Phosphorylation of ERK1/2, however, does not seem to be involved in the elevation of CXCR4 mRNA expression at least after 6 h.In TKKK cells, on the other hand, CXCR4 expression was mostly suppressed by the MEK/ERK system inhibitor U0126 at 6 h (Fig. 3A, right panel).Also, since ERK phosphorylation induced by H 2 O 2 administration was blocked by U0126 (Fig. 3B right panel), CXCR4 elevation in TKKK cells might occur via the MEK/ERK pathway.Indeed, previous studies have described involvement of MEK/ERK and NF-κB signaling pathways in CXCR4 mRNA expression in various cell lines-mouse podocyte cell line MPC5 (Mo et al. 2017), human breast carcinoma cell line MCF-7 (Maroni et al. 2007), and human osteosarcoma cell lines U2OS and MG-63 (Huang et al. 2009).Moreover, the ERK signaling pathway has been identified as a critical modulator of major phenotypic responses exhibited by profibrogenic liver myofibroblasts after the induction of oxidative stress (Foglia et al. 2019).
In the current study, experiments with the use of ERK activators were conducted to further investigate the correlation between CXCR4 expression and the ERK pathway.Western blotting with ERK activators demonstrated rapid and transient ERK phosphorylation in both Huh-7 and TKKK cells (Fig. 4A).In Huh-7 cells, however, no significant elevation of CXCR4 expression was observed after treatment with ERK activators for 6 or 24 h (Fig. 4B, left panel).On the other hand, in TKKK cells, albeit statically not significant, CXCR4 expression tended to be higher after treatment with ERK activator C6-ceramide than with H 2 O 2 at 6 h (Fig. 4B, right panel).Thus, our results suggest that the ERK pathway might account for the upregulation of CXCR4 expression in TKKK cells but not in Huh-7 cells.
Based on the unique reciprocal expression pattern of MeCP2 and CXCR4 (Fig. 1B), the potentiality of an epigenetic mechanism regulating CXCR4 gene expression in the liver was investigated.The results depicted in Fig. 5 (left panel) indicate that CXCR4 mRNA expression was upregulated in Huh-7 cells treated with DNA methyltransferase inhibitor 5-Aza and/or the histone deacetylase inhibitor TSA, indicating that epigenetic pathways, at least in part, mediate CXCR4 expression in Huh-7 cells.In contrast, no significant changes in CXCR4 mRNA expression were observed in TKKK cells, indicating a limited role of epigenetic regulation in CXCR4 expression in TKKK cells (Fig. 5, right panel).The precise epigenetic regulation of Fig. 8 Assumed regulation mechanism of CXCR4 gene expression under oxidative stress.Among several signaling pathways shown to be activated by oxidative stress, our results revealed that MEK/ERK and NF-κB signaling play central roles in CXCR4 expression in intrahepatic cholangiocytes.On the other hand, in hepatocytes, CXCR4 expression is regulated by a unique epigenetic mechanism in that methylation of cytosine in CpG dinucleotides located in promoter region is lost, which contributes to the upregulation of CXCR4 gene expression CXCR4 gene expression in the liver was also investigated, especially in hepatocytes, as well as the distribution pattern of CpG-dinucleotide around the promoter region of the CXCR4 gene and we noted that two continuous CpG sites located adjacent to the TA repeat (overlapping TATA box) form a typical target sequence of MeCP2 (Kitazawa and Kitazawa 2007;Klose et al. 2005;Mori et al. 2014).
The methylation status around the TA repeat in the promoter region of the CXCR4 gene estimated by MSP revealed that the band methylated by MSP seen in the normal liver was lost in the hepatic failure (Fig. 6A).Bisulfite sequencing of two methylated CpG dinucleotides four bases upstream of the TA repeat confirmed the presence of methylation sites in the normal liver and its absence in hepatic failure (Fig. 6B).Taken together with that epigenetic regulation is reversible and essential for spatial and temporal gene expression, reactivation of CXCR4 gene expression may be partly explained by the loss of DNA methylation at the MeCP2 target site in the promoter region of CXCR4 in hepatic failure.Indeed, such dynamics of DNA methylation at the promoter region of the CXCR4 gene, albeit not at the present MeCP2 target site, has been described in instances of estrogen exposure in human endometrial adenocarcinoma cell lines (Kubarek and Jagodzinski 2007) and in hypoxic stimuli in human lung cell lines (Kang et al. 2019).Possibly, the loss of methylation is further compounded by reduced expression of MeCP2 in hepatic failure, resulting in the rapid reactivation of CXCR4 expression.
When liver regeneration is impaired and liver failure ensues, the only currently established life-saving treatment of choice is liver transplantation.Our findings suggest that besides the ordinal MEK/ERK signaling pathway, a unique epigenetic mechanism, attributed to the loss of CpG methylation in the promoter region of CXCR4, contributes to the reactivation of CXCR4 expression in hepatocytes under conditions of severe liver injury.

Fig. 1
Fig. 1 Representative histochemical demonstration of liver tissues from normal and hepatic failure.A Representative histopathological findings of cases used in this study (a-f).In normal liver tissue, hepatocytes (HC) form polygonal-shaped lobule containing a central vein (CV) at the center and portal tracts at the corner of the hexagon (a, hematoxylin and eosin (HE) staining).By immunohistochemical analyses, Hep par 1, a marker of hepatocytes, is uniformly positive in the cytoplasm of normal hepatocytes (b).Cytokeratin7 (CK7), a marker of bile duct epithelial cells, shows strong staining in the cytoplasm of normal bile ducts in the portal regions (c).In the hepatic failure, HE staining shows a typical regenerative nodule (asterisk) with eosinophilic cytoplasm in the lower-right corner.Pseudoducts are also observed around the regenerative nodule (indicated by arrows in d).Irreversible cell injury with nuclear pyknosis is observed in some hepatocytes, where Hep par 1 immunostaining shows uneven

Fig. 4
Fig.4Roles of ERK pathway in Huh-7 and TKKK cells.A Western blotting confirmed that ERK activators induced rapid and transient ERK1/2 phosphorylation.B In Huh-7 cells, after 24 h of serum starvation (0.1% FBS) followed by 6 or 24 h of 1 mM of H 2 O 2 treatment, CXCR4 mRNA expression increased 3.1-and 7.3-fold, respectively.However, treatment with two ERK activators, C6-ceramide and tertbutylhydroquinone (tBHQ), at any dose, did not affect the CXCR4 expression, indicating that ERK pathway do not play a major role in

Fig. 7
Fig. 7 Effects of CpG methylation of CXCR4 promoter region on binding with methyl-CpG-binding protein 2 (MeCP2) and TATA binding protein (TBP).A Electrophoretic mobility shift assay (EMSA): the unmethylated oligonucleotides (U) spanning the TA repeat in the CXCR4 gene promoter region show protein-DNA bindings (black arrow) while the methylated oligonucleotides (M) show another protein-DNA binding (black arrowhead) at higher position (left panel).The protein-DNA bindings in unmethylated oligonucleotides (black arrow) are washed out by excess amounts of cold TA repeat competitor (middle panel).In methylated oligonucleotides, the protein-DNA binding (black arrowhead) is block-shifted with an anti-MeCP2 antibody (right panel).Right panel shows an effect of anti-MeCP2 antibody on the protein-DNA binding.The addi-