(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .
The transcription factor HBP1 promotes ferroptosis in tumor cells by regulating the UHRF1-CDO1 axis [1]
['Ruixiang Yang', 'Department Of Biochemistry', 'Biophysics', 'School Of Basic Medical Sciences', 'Beijing Key Laboratory Of Protein Posttranslational Modifications', 'Cell Function', 'Peking University Health Science Center', 'Beijing', 'Yue Zhou', 'Tongjia Zhang']
Date: 2023-07
The induction of ferroptosis in tumor cells is one of the most important mechanisms by which tumor progression can be inhibited; however, the specific regulatory mechanism underlying ferroptosis remains unclear. In this study, we found that transcription factor HBP1 has a novel function of reducing the antioxidant capacity of tumor cells. We investigated the important role of HBP1 in ferroptosis. HBP1 down-regulates the protein levels of UHRF1 by inhibiting the expression of the UHRF1 gene at the transcriptional level. Reduced levels of UHRF1 have been shown to regulate the ferroptosis-related gene CDO1 by epigenetic mechanisms, thus up-regulating the level of CDO1 and increasing the sensitivity of hepatocellular carcinoma and cervical cancer cells to ferroptosis. On this basis, we constructed metal-polyphenol-network coated HBP1 nanoparticles by combining biological and nanotechnological. MPN-HBP1 nanoparticles entered tumor cells efficiently and innocuously, induced ferroptosis, and inhibited the malignant proliferation of tumors by regulating the HBP1-UHRF1-CDO1 axis. This study provides a new perspective for further research on the regulatory mechanism underlying ferroptosis and its potential role in tumor therapy.
Funding: This work was supported by grants of the National Natural Science Foundation of China (No. 82073068, 81874141, and 81672717) and Beijing Natural Science Foundation Grant (No. 7212056) to XWZ. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
In this study, we determined that HBP1 up-regulates the expression of CDO1 at the epigenetic level by inhibiting the expression of the UHRF1 gene at the transcription level, thereby promoting CDO1-mediated ferroptosis in tumor cells. Based on our findings, we combined tannic acid, a food additive extracted from green tea and approved by the Food and Drug Administration (FDA), with Fe 3+ to form MPN on the surface of a polyethylenimine-HBP1 plasmid complex (PEI-HBP1). We found that MPN-HBP1 nanoparticles enhanced the induction of ferroptosis. MPN-HBP1 was internalized by tumor cells and introduced a significant amount of Fe 3+ into the cells to induce the Fenton reaction to produce ROS, thus resulting in serious lipid peroxidation in the biomembrane. The exogenous expression of HBP1 enhanced ferroptosis in tumor cells by regulating the UHRF1-CDO1 signaling pathway. Herein, we evaluated the efficacy of MPN-HBP1 as a novel nanodrug with the capacity to induce ferroptosis in cancer therapy. This strategy may provide an option for improving the outcome of traditional cancer therapy.
Over recent years, nanomaterials have been widely used in laboratory research and clinical practice [ 21 , 22 ]. Some researchers have used nanoparticles synthesized by a metal polyphenol network (MPN) approach as carriers for inducers of ferroptosis such as Erastin or iron-based nanomaterials such as Fe 3 O 4 nanoparticles to accurately induce ferroptosis in tumor tissues, so as to achieve therapeutic action [ 23 – 25 ]. The combination of induced ferroptosis and nanotechnology enhances the stability, biosafety, and targeting of drugs in vivo.
HMG box-containing protein 1 (HBP1), as a dual transcription factor, inhibits its target genes by directly binding to specific affinity elements, such as N-MYC, C-MYC, p47phox, DNMT1, EZH2, and AFP [ 6 – 11 ], all of which are oncogenes or genes that promote tumor development. HBP1 also transcriptionally activates several downstream genes, including p16, p21, myeloperoxidase (MPO), and histone H1 [ 12 – 15 ], which are all tumor suppressor genes. HBP1 regulates the cell cycle and inhibits cell proliferation by regulating the expression of its downstream cell cycle regulatory factors, thus inhibiting the occurrence and development of tumors. However, it remains unknown as to whether HBP1, as an important regulator of the cell cycle and metabolism, is involved in ferroptosis and whether it plays a tumor suppressor role by regulating ferroptosis. Ubiquitin-like with PHD and RING finger domains 1(UHRF1) protein is an epigenetic modification factor that plays a significant role in DNA methylation and histone methylation. UHRF1 is expressed at high levels in various malignant tumors, including breast, bladder, and prostate cancer, and is involved in the occurrence and progression of tumors [ 16 – 18 ]. In addition, UHRF1 can inhibit cell apoptosis via ROS-related signaling pathways in gastric cancer [ 19 ] and enhance the invasiveness of tumor cells via the Keap1-Nrf2 pathway in pancreatic cancer [ 20 ]. Therefore, UHRF1 may represent a crucial regulatory factor in tumorigenesis and development; however, the regulatory mechanisms upstream of UHRF1, especially in terms of transcriptional regulation, remain unclear.
Human cysteine dioxygenase 1 (CDO1), an enzyme that adds molecular oxygen to the sulfur of cysteine and converts the thiol to a sulfinic acid known as cysteine sulfinic acid (3-sulfinoalanine). CDO1-induced cysteine deficiency has been shown to reduce glutathione (GSH) synthesis and weaken the antioxidant capacity of cells, eventually leading to an increase in the levels of reactive oxygen species (ROS) and the induction of ferroptosis [ 4 ]. Research has shown that the inhibition of CDO1 expression contributes to the elevated synthesis of GSH as well as a reduction in ROS levels, ultimately resulting in resistance to Erastin-induced ferroptosis [ 5 ]. Thus, high expression levels of CDO1 is an important factor in the development of ferroptosis.
Ferroptosis is an iron-dependent form of regulatory cell death that is caused by the loss of cellular redox homeostasis, thus to uncontrolled lipid peroxidation and eventually cell death [ 1 ]. Ferroptosis is associated with ischemia-induced pathological cell death and different types of cancer [ 2 ]. Many types of tumor cells are susceptible to ferroptosis after drug treatment, including cervical cancer, hepatocellular carcinoma, pancreatic cancer, and renal cell carcinoma [ 3 ]. Therefore, inducing ferroptosis in tumor cells has become an important aspect of tumor therapy. However, the regulatory mechanisms underlying ferroptosis have yet to be fully elucidated.
Results
HBP1 regulates UHRF1 expression Previously, we reported that HBP1-mediated transcriptional regulation of the methyltransferase DNMT1 induces global DNA hypomethylation during cell senescence [8]. We also identified HBP1 as an intriguing and important factor in regulating the methylation state of DNA. UHRF1 regulates DNA methylation at the epigenetic level by recruiting DNMT1 and links DNA methylation and methylation maintenance following cell division [26–28]. Therefore, in the present study, we investigated whether UHRF1 is another target of HBP1. First, we evaluated the expression levels of HBP1 and UHRF1 in pathological sections taken from various cancer patients (hepatocellular carcinoma, lung adenocarcinoma, and colon adenocarcinoma) by immunohistochemistry (IHC) staining. We observed a statistically significant inverse correlation between the protein levels of HBP1 and UHRF1 in the 3 types of tumors with high expression levels of UHRF1 (Fig 1A and 1B). Accordingly, the expression levels of HBP1 mRNA were negatively correlated with that of UHRF1 in TCGA public databases for liver, breast, and lung cancers (S1A Fig). PPT PowerPoint slide
PNG larger image
TIFF original image Download: Fig 1. HBP1 regulates UHRF1 expression. (A) IHC staining of HBP1 and UHRF1 in pathological sections of cancer patients (n = 12/group). Scale bar = 100 μm. (B) Two-tailed Pearson correlation analysis method was used to calculate the correlation between the expression of HBP1 and UHRF1. (C) HBP1 overexpression decreases UHRF1 protein and mRNA expression. The protein levels of HBP1 and UHRF1 in cell lysate was measured by western blotting in HeLa, HepG2, and Huh7 cells transfected with pCDNA3-HBP1 or pCDNA3 (as a control). β-actin was used as a control, respectively (left panel). The mRNA level of UHRF1 was measured by real-time PCR in HeLa, HepG2, and Huh7 cells transfected with pCDNA3-HBP1 or pCDNA3 (right panel). (D) HBP1 knockdown by shRNA increases UHRF1 protein and mRNA expression. The protein levels of HBP1 and UHRF1 in cell lysate was measured by western blotting in HeLa, HepG2, and Huh7 cells stably transfected with pLL3.7-shHBP1-1, pLL3.7-shHBP1-2, or pLL3.7 (as a control) through lentiviral infection. β-actin was used as a control, respectively (left panel). The mRNA level of UHRF1 was measured by real-time PCR in HeLa, HepG2, and Huh7 cells stably transfected with pLL3.7-shHBP1-1, pLL3.7-shHBP1-2, or pLL3.7 (right panel) through lentiviral infection. The underlying data for Fig 1B–1D can be found in S1 Data. Differences between 2 groups were calculated using a two-tailed Student t test. Error bars represent SD., p < 0.05, **, p < 0.01, ***, p < 0.001. COAD, colon adenocarcinoma; HBP1, HMG box-containing protein 1; HCC, hepatocellular carcinoma; IHC, immunohistochemical; LUAD, lung adenocarcinoma; shRNA, short hairpin RNA; UHRF1, ubiquitin-like with PHD and RING finger domains 1.
https://doi.org/10.1371/journal.pbio.3001862.g001 To determine whether HBP1 is a repressor of the UHRF1 gene, we overexpressed HBP1 in HeLa, HepG2, and Huh7 cells (Fig 1C). The levels of UHRF1 protein (left panel) and mRNA (right panel) were reduced by HBP1 overexpression in HeLa, HepG2, and Huh7 cells. To confirm the endogenous regulatory activity of HBP1 on UHRF1 expression, we knocked down the expression of HBP1 using short hairpin RNAs (shRNAs). As shown in Fig 1D, the knockdown of HBP1 resulted in increased levels of UHRF1 protein (left panel) and mRNA (right panel) in the 3 cell lines. These results suggest that HBP1 inhibits UHRF1 expression at the transcriptional level.
HBP1 represses the UHRF1 gene by binding to an affinity site in the UHRF1 promoter Next, we investigated whether HBP1 inhibits the transcriptional activity of the UHRF1 promoter through sequence-specific DNA binding. We cotransfected HEK293T cells with distinct fragments of the UHRF1 promoter (−1,783 to +74, −1,536 to +74, −1,285 to +74, −1,121 to +74 from the transcriptional start site) and HBP1. As shown in Fig 2A, HBP1 inhibited some UHRF1 promoter fragments (−1,783 to +74, −1,536 to +74, and −1,285 to +74) but had no effect on one UHRF1 fragment (−1,121 to +74), thus indicating that site of affinity for HBP1 was between −1,285 and −1,121 bp in the UHRF1 promoter. To further verify the DNA binding requirement for the functional activity of HBP1, we constructed a deletion reporter for the UHRF1 promoter, which abolished the HBP1 affinity site between −1,173 and −1,155 bp. HBP1 inhibited the activity of the wild-type UHRF1 promoter but had no effect on the mutant promoter (Fig 2B). In order to investigate whether the transcriptional repression of HBP1 depends on DNA binding, we used 2 mutants of HBP1: pmHMG (which had 3 amino acid mutations in the HMG domain and lacked DNA binding ability) and DelEx7 (which was isolated from breast cancer tissue and lacked the DNA binding domain and part of the repression domain) [8,12]. As shown in Fig 2C and 2D, wild-type HBP1 reduced the activity of the UHRF1 promoter and the level of protein while the overexpression of pmHMG and DelEx7 had no effect, indicating that HBP1 transcriptional inhibition of UHRF1 gene expression depends on DNA binding. Since HBP1 inhibited the activity of the UHRF1 promoter, we tested whether HBP1 binds directly to the UHRF1 promoter. Chromatin immunoprecipitation (ChIP) assay (Fig 2E) demonstrated that HBP1 bound directly to the specific affinity site in the UHRF1 promoter in vivo. In contrast, pmHMG and DelEx7 did not bind to the UHRF1 promoter. Therefore, we concluded that HBP1 inhibits the UHRF1 gene by binding to a site of affinity within the UHRF1 promoter. PPT PowerPoint slide
PNG larger image
TIFF original image Download: Fig 2. HBP1 represses UHRF1 gene by binding an affinity site in the UHRF1 promoter. (A) Relative activity of HBP1 on the UHRF1 promoters with various lengths. (B) The integrity of affinity site is indispensable for HBP1 suppressing UHRF1 promoter in vivo. Shown is schematic diagram of the wild-type UHRF1 promoter and its mutant promoter (left panel) and the relative activities of HBP1 on the wild-type UHRF1 promoter and mutant UHRF1 promoters (right panel). (C) Relative activities of HBP1 and associated mutants on the UHRF1 promoter in cotransfected HEK293T cells. Shown is schematic diagram of wild-type HBP1 and associated mutants (left panel). Luciferase activity was determined after transfection (right panel). (D) Expression of exogenous HBP1 decreases UHRF1 protein level. HEK293T cells were transfected with HBP1 and associated mutants. The protein level was measured by western blotting. (E) HBP1 binding to the endogenous UHRF1 promoter requires the HMG domain. ChIP assays were used to test the binding of exogenous HBP1 to endogenous UHRF1 gene. HEK293T cells were transfected with HA-HBP1, HA-pmHMG, or HA-DelEx7. The region from position −1,289 to position −1,067 contains the HBP1 affinity site and was analyzed by specific PCR. Anti-HA antibody was used in the indicated lanes. The underlying data for Fig 2A, 2B, 2C and 2E can be found in S1 Data. Differences between 2 groups were calculated using a two-tailed Student t test. One-way ANOVA was performed to assess differences among multiple groups. Error bars represent S.D. *, p < 0.05, **, p < 0.01, ***, p < 0.001. ChIP, chromatin immunoprecipitation; HBP1, HMG box-containing protein 1; UHRF1, ubiquitin-like with PHD and RING finger domains 1.
https://doi.org/10.1371/journal.pbio.3001862.g002
HBP1 reduces cellular antioxidant capacity and therefore sensitizes tumor cells to ferroptosis To investigate potential physiological or pathological biological processes involving HBP1, we first downloaded and analyzed relevant mRNA sequence data from the LIHC data set (Cbioportal) of TCGA database. Gene Set Enrichment Analysis (GSEA) identified HBP1 as generally associating with oxidative stress, cell death in response to oxidative stress, oxidoreductase activity, and lipid oxidation (S1B Fig). We specifically focused on oxidation stress process and then asked if HBP1 expression could be linked to the regulation of redox homoeostasis in tumor cells. The effect of HBP1 levels on redox balance in HepG2 cells was then explored using a HBP1 overexpression system and lentivirus-mediated knockdown. We first examined the subsequent effect on the nicotinamide adenine dinucleotide phosphate (NADPH) and GSH systems that represent important guardians for maintaining cell redox homoeostasis. HBP1 overexpression in HepG2 cells resulted in significantly decreased ratios of NADPH/NADP+ and GSH/GSSG (S2A and S2B Fig, left panels). In contrast, HBP1 knockdown in HepG2 cells led to an increase in the ratios (S2A and S2B Fig, right panels), suggesting that HBP1 can alter the redox balance in HepG2 cells by transitioning to oxidation. To determine whether redox changes caused by HBP1 is accompanied by ROS accumulation, we tested intracellular ROS levels in the HepG2 cells with HBP1 overexpression or knockdown. Since endogenic ROS are mainly produced in mitochondria, we also tested mitochondrial ROS levels in these cells. We found that both intracellular ROS levels and mitochondrial ROS levels in HBP1 overexpressed cells were significantly elevated compared to the vector, and this phenotype could be reversed by adding of the antioxidant N-acetyl-l-cysteine (NAC). In contrast, the intracellular total and mitochondrial ROS levels in HBP1 knockdown cells were significantly reduced compared with the vector (S2C and S2D Fig). We further examined potential changes in mitochondrial membrane potential (ΔΨm) in these cells to determine if mitochondrial function was impaired by the elevated ROS. Compared with the vector, HBP1 overexpressed cells had a higher proportion of depolarized mitochondria, showing a decrease in ΔΨm, while HBP1 knockdown cells had an increase in ΔΨm (S2E Fig). NAC treatment restored the decreased ΔΨm associated with HBP1 overexpression in HepG2 cells. These results suggest that HBP1 raises ROS levels and thus damages ΔΨm in HepG2 cells. We further hypothesized that HBP1 might restrict the antioxidant capacity of HepG2 cells, thereby sensitizing these cells to ROS damage. In support of this hypothesis, HBP1 overexpressed cells showed impaired cellular antioxidant capacity and HBP1 knockdown cells showed enhanced antioxidant capacity compared with the vectors (S2F Fig). We then treated HepG2 cells with hydrogen dioxide (H 2 O 2 ) and found that HepG2 cells with HBP1 overexpression were more sensitive to H 2 O 2 treatment than the vector, while HepG2 cells with HBP1 knockdown were more resistant to H 2 O 2 treatment (S2G Fig). Taken together, these results strongly suggest that HBP1 overexpression impairs antioxidant capacity of HepG2 cells, disrupts redox balance, and sensitizes HepG2 cells to oxidative stress-induced damage. Then, we used the Cancer Therapeutics Response Portal (CTRP), which allowed us to analyze correlations between gene expression and the response to 481 compounds across different cancer cell lines. Liver cancer cell line data from the CTRP revealed a significant correlation between HBP1 expression and sensitivity to RSL-3 (S2H Fig), which targets glutathione peroxidase 4 (GPX4) to induce ferroptosis [29]. Our previous work has shown that HBP1 is involved in the biological process of H 2 O 2 -induced apoptosis [30]. Therefore, we planned to investigate whether HBP1 participates in ferroptosis, another form of cell death, by regulating UHRF1. Next, we investigated the induction of ferroptosis by RSL-3 or Erastin, both of which are well-known inducers of ferroptosis. As shown in S2I Fig, HeLa and HepG2 cells were treated with Erastin or RSL-3. Cell viability decreased with the increase of Erastin or RSL-3 concentrations. Ferrostatin-1 (Fer-1), an inhibitor of ferroptosis, was able to rescue RSL-3- or Erastin-induced cell death (S2J Fig), thus confirming the induction of ferroptosis by RSL-3 or Erastin. To determine whether the HBP1-UHRF1 axis plays a pivotal role in regulating ferroptosis, we analyzed the protein levels of HBP1 and UHRF1 during ferroptosis in HeLa, HepG2, and Huh7 cells. Increasing concentrations of Erastin or RSL-3 led to an increase in HBP1 protein level and a reduction in UHRF1 protein level (Fig 3A and 3B, left panels), whereas Fer-1 rescued the Erastin/RSL-3-induced changes in protein levels (Fig 3C). We also introduced an iron scavenger, Deferoxamine (DFO), another ferroptosis inhibitor, which could rescue the Erastin-induced changes in HBP1 and UHRF1 protein levels as well (S2K Fig). However, the mRNA expression of HBP1 remained unchanged following Erastin or RSL-3 treatment (Fig 3A and 3B, right panels). These results suggest that the HBP1-UHRF1 axis may contribute to ferroptosis and that the stability of HBP1 protein may change during ferroptosis. PPT PowerPoint slide
PNG larger image
TIFF original image Download: Fig 3. HBP1 reduces cellular antioxidant capacity and therefore sensitizes tumor cells to ferroptosis. (A) HBP1 protein increased and UHRF1 protein decreased during ferroptosis. After HeLa, HepG2, and Huh7 cells were treated with different concentrations of Erastin, the protein expression of HBP1 and UHRF1 was detected by western blotting (left panel) and the mRNA expression level of HBP1 was detected by qPCR after HeLa cells were treated with 10 μM Erastin (right panel). (B) After HeLa, HepG2, and Huh7 cells were treated with different concentrations of RSL-3, the protein expression of HBP1 and UHRF1 was detected by western (left panel) and the mRNA expression level of HBP1 was detected by qPCR after HeLa cells were treated with 5 μM RSL-3 (right panel). (C) Protein levels of HBP1 and UHRF1 in HeLa, HepG2, and Huh7 cells were treated with Erastin (10 μM, left panel)/RSL-3 (5 μM, right panel) alone or in combination with Fer-1 (10 μM). (D) Erastin inhibits HBP1 ubiquitination-mediated proteasome degradation. HeLa cells were treated with 10 μM Erastin with or without MG132; the protein level of HBP1 was measured by western blotting. (E) Erastin extends the half-life of HBP1 protein. HeLa cells were treated with Erastin for 24 h, and cells were incubated with the protein synthesis inhibitor CHX for 0, 30, 60, 90, or 120 min before collect. HBP1 protein levels were detected by western blotting. (F) HEK293T cells were transfected HA-HBP1 with or without 10 μM Erastin treatment for 24 h and then exposed to MG132 for another 6 h prior to lysis. HBP1 protein was then isolated by immunoprecipitation and analyzed by anti-Ub antibody. (G) Erastin promotes the suppression of HBP1 on UHRF1 protein. HeLa cells were transfected HBP1 with or without 10 μM Erastin treatment. The protein levels of HBP1 and UHRF1 were measured by western blotting. (H) Erastin promotes the suppression of HBP1 on UHRF1 mRNA. The mRNA level of UHRF1 was measured by qPCR in HeLa cells transfected HBP1 with or without 10 μM Erastin treatment. (I) Erastin promotes the suppression of HBP1 on UHRF1 promoter. HEK293T cells were cotransfected with UHRF1 promoter and HBP1 with or without 10 μM Erastin treatment. Luciferase activity was determined after transfection. (J) Erastin promotes the interaction between HBP1 and UHRF1 promoter. HEK293T cells were transfected HA-HBP1 with or without 10 μM Erastin treatment. The region from position −1,289 to position −1,067 contains the HBP1 affinity site and was analyzed by specific PCR. (K) Representative TEM images of the mitochondrial morphology in HBP1-OE and HBP1-KD HeLa cells treated with 10 μM Erastin for 24 h. Red arrows indicate mitochondria. Scale bar = 1 μm (top)/500 nm (bottom). (L) Representative images (left) and quantification (right) of ROS level in HBP1-OE and HBP1-KD HeLa cells treated with 10 μM Erastin for 24 h. Scale bar = 10 μm. The underlying data for Fig 3A, 3B, 3H, 3I and 3L can be found in S1 Data. Differences between 2 groups were calculated using a two-tailed Student t test. One-way ANOVA was performed to assess differences among multiple groups. Error bars represent S.D. *, p < 0.05, **, p < 0.01, ***, p < 0.001. CHX, cycloheximide; Fer-1, ferrostatin-1; HBP1, HMG box-containing protein 1; KD, knockout; OE, overexpression; ROS, reactive oxygen species; TEM, transmission electron microscopy; UHRF1, ubiquitin-like with PHD and RING finger domains 1.
https://doi.org/10.1371/journal.pbio.3001862.g003 Subsequently, we used a selective 26S proteasomal inhibitor (MG-132) to address the effect of ubiquitination-mediated proteasomal degradation on HBP1 protein during ferroptosis. The levels of HBP1 protein were increased by treatment with MG-132. Erastin treatment did not induce a further increase in HBP1 (Fig 3D). As shown in Figs 3E and S3, Erastin clearly increased the half-life of HBP1 protein from 20.7 min to 53.5 min, suggesting that Erastin-induced increase in HBP1 protein level is mediated in a proteasome-dependent manner. To determine whether Erastin inhibits the ubiquitination and degradation of HBP1, we transfected HEK293T cells with HA-HBP1 with or without Erastin treatment and then exposed the cells to MG132 for 6 h. Subsequently, the cells were subjected to anti-HA immunoprecipitation followed by western blotting with an anti-ubiquitin antibody. As shown in Fig 3F, Erastin significantly inhibited HBP1 ubiquitination. Thus, we concluded that Erastin increases the expression of HBP1 protein by inhibiting HBP1 ubiquitination-mediated proteasomal degradation. Analyses showed that HBP1 inhibited the protein level, mRNA level, and promoter activity of UHRF1 in the presence of Erastin in a more significant manner than with HBP1 alone (Fig 3G–3I). ChIP assays further showed that Erastin enhanced the binding of HBP1 to the UHRF1 promoter (Fig 3J). These data indicate that the up-regulation of HBP1 protein by Erastin further enhances the transcriptional inhibition of HBP1 on UHRF1 by enhancing the binding of HBP1 to the UHRF1 promoter. To further determine the role of HBP1 in ferroptosis, we used transmission electron microscopy (TEM) to detect morphological changes in HeLa cells with HBP1 overexpression or HBP1 knockdown treated with Erastin. The cells overexpressing HBP1 and treated with Erastin possessed smaller mitochondria, diminished or vanished mitochondria cristae, and condensed mitochondrial membrane densities; these were typical mitochondrial phenotypes of ferroptosis when compared to vector cells. In contrast, HBP1 knockdown alleviated the abnormalities of mitochondrial morphology and cell death induced by Erastin (Fig 3K). We also measured ROS levels after interference with HBP1 expression; ROS levels are known to be the primary cause of ferroptosis. Results showed that the overexpression of HBP1 led to a significant increase in ROS levels in Erastin-induced HeLa cells, whereas HBP1 knockdown reduced the levels of ROS (Fig 3L). In conclusion, these results suggest that HBP1 promotes ferroptosis in tumor cells and sensitizes tumor cells to ferroptosis.
HBP1 sensitizes tumor cells to ferroptosis by inhibiting UHRF1 expression in tumor cells Since the HBP1-UHRF1 axis may contribute to ferroptosis in tumor cells, we speculate that HBP1 may increase the sensitivity of tumor cells to ferroptosis by inhibiting UHRF1. We next investigated the effect of HBP1 on UHRF1 expression in the context of Erastin treatment. HBP1 knockdown recovered the reduced UHRF1 levels induced by Erastin treatment in HeLa, HepG2, and Huh7 cells (Fig 4A), thus indicating that the inhibitory effect of HBP1 on UHRF1 is involved in the ferroptosis process in tumor cells. PPT PowerPoint slide
PNG larger image
TIFF original image Download: Fig 4. HBP1 sensitizes tumor cells to ferroptosis by inhibiting UHRF1 expression in tumor cells. (A) Protein levels of HBP1 and UHRF1 in HBP1 knockdown cells treated with 10 μM Erastin. (B) Indicated cells were treated with or without 10 μM Erastin for 24 h in the presence of different cell death inhibitors (Fer-1, 10 μM; DFO, 50 μM; Ac-DEVD-CHO, 59 μM; Necrosulfonamide, 20 μM; Chloroquine, 10 μM). Cell viability was measured using MTT. (C) Cell viability was conducted with HeLa cells stably transfected with vector, HBP1, HBP1+UHRF1 or vector, shHBP1, shHBP1+shUHRF1. (D-G) HBP1-UHRF1 axis sensitizes HeLa cells to ferroptosis. Indicated cells were treated with or without 10 μM Erastin for 24 h. (D) Cells were collected, and confocal was used to detect Fe2+ levels. (E) Flow cytometry was used to detect lipid peroxides. (F, G) Indicated cells were lysed and MDA, GSH, and iron contents were measured. Scale bar = 10 μm. The underlying data for Fig 4B–4G can be found in S1 Data. Differences between 2 groups were calculated using a two-tailed Student t test. One-way ANOVA was performed to assess differences among multiple groups. Error bars represent SD. *, p < 0.05, **, p < 0.01, ***, p < 0.001. DFO, Deferoxamine; Fer-1, ferrostatin-1; GSH, glutathione; HBP1, HMG box-containing protein 1; MDA, malondialdehyde; UHRF1, ubiquitin-like with PHD and RING finger domains 1.
https://doi.org/10.1371/journal.pbio.3001862.g004 To determine the type of cell death associated with increased HBP1 expression and Erastin induction, we also evaluated the effect of ferroptosis inhibitor (Fer-1, DFO), necroptosis inhibitor Necrosulfonamide (Nec-1), apoptosis inhibitor Ac-DEVD-CHO (Apo), and autophagy inhibitor Chloroquine (Chq) on the viability of cells. As shown in Fig 4B, for all tested cells (HeLa/HBP1, HeLa/shHBP1, and their vector counterparts), only Fer-1 and DFO were able to protect cells from Erastin-induced cell death. Nec-1, Apo, and Chq had no significant effect on Erastin-induced cell death. The overexpression of HBP1 enhanced Erastin-induced cell death while HBP1 knockdown had the opposite effect. These data indicate that Erastin causes tumor cell death by inducing ferroptosis and that HBP1 may enhance the antitumor effect of Erastin by regulating ferroptosis. Reduced cell activity, increased intracellular Fe2+, lipid peroxidation, and the depletion of GSH are recognized hallmarks of ferroptosis [31].To confirm whether HBP1 can promote Erastin-induced ferroptosis via UHRF1 down-regulation, we measured the levels of these markers in tumor cells transfected with HBP1, HBP1+UHRF1, HBP1shRNA, or HBP1shRNA+UHRF1shRNA following the administration of Erastin. We observed a reduction of cell viability following Erastin treatment in HeLa vector cells. The overexpression of HBP1, however, resulted in a further reduction in cell viability that could be restored by the overexpression of UHRF1. Conversely, the knockdown of HBP1 partially blocked the reduction of cell viability after Erastin treatment; this effect was reversed by knocking down UHRF1 (Fig 4C). Labile iron (Fe2+) is an important source of hydroxyl radical formation and can initiate lipid peroxidation via the Fenton reaction [32]. We found that both the overexpression and knockdown of HBP1 had a slight effect on the intracellular Fe2+ content of HeLa cells (Fig 4D). However, the combination of HBP1 overexpression and Erastin treatment led to an obvious increase in Fe2+ production than that seen with Erastin treatment alone in HeLa cells. This effect of HBP1 was restored by the overexpression of UHRF1. In contrast, the combination of HBP1 knockdown and Erastin treatment resulted in reduced Fe2+ production when compared with Erastin treatment alone; this effect was reversed by the knockdown of UHRF1. These data suggest that HBP1 may account for Fe2+ production or iron metabolism in ferroptosis by down-regulating UHRF1. Furthermore, the overexpression of HBP1 led to a further increase in Erastin-induced lipid peroxidation, while HBP1 knockdown restrained the increase of lipid peroxidation induced by Erastin in HeLa cells. These actions of HBP1 overexpression and knockdown could be reversed by the overexpression and knockdown of UHRF1, respectively (Fig 4E). Malondialdehyde (MDA), a product of lipid peroxidation, another sensitive indicator of ferroptosis, is consistent with the level of lipid peroxidation (Fig 4F and 4G, left panels). In addition, GSH was significantly reduced after Erastin treatment. The overexpression of HBP1, however, resulted in a further reduction of GSH levels, and overexpression of UHRF1 restored this effect (Fig 4F, middle panel). HBP1 knockdown partially blocked the depletion of GSH arising from Erastin treatment. This effect could be reversed by the knockdown of UHRF1 (Fig 4G, middle panel). We also used a colorimetric method to detect the intracellular Fe2+ content (Fig 4F and 4G, right panels), and the results were the same as those of fluorescent probe staining shown in Fig 4D. Collectively, these results suggest that HBP1 sensitizes tumor cells to ferroptosis by inhibiting UHRF1 expression. Sorafenib is a first-line molecular-target drug for advanced hepatocellular carcinoma (HCC). The main function of sorafenib is to inhibit the proliferation and promote the apoptosis of HCC cells. However, the resistance of HCC cells to apoptosis makes hepatoma patients prone to drug resistance to sorafenib. Recently, sorafenib was identified as an inducer of ferroptosis [33]. The classic phenotype of ferroptosis, such as GSH depletion, elevated Fe2+and increased ROS production, is often observed in the cells treated with sorafenib. Therefore, promoting ferroptosis induced by sorafenib may be a new way to better treat HCC. According to the regulation of ferroptosis by HBP1-UHRF1 axis, we reasonably speculated that HBP1 might also increase the sensitivity of HCC cells to sorafenib by inhibiting the expression of UHRF1. To confirm this conjecture, we used Erastin or sorafenib to treat HepG2 cells transfected with HBP1, HBP1+UHRF1, shHBP1, shHBP1+shUHRF1. As shown in S4 Fig, similar to Erastin, solafenib treated HepG2 cells with HBP1 overexpression, resulting in a further decline in cell viability (S4A Fig), a more significant increase in intracellular Fe2+, ROS, and MDA levels (S4B–S4D Fig), and a further decrease in GSH levels compared with solafenib treated alone (S4D Fig). Overexpression of UHRF1 reversed these effects of HBP1. However, solafenib treated HepG2 cells with HBP1 knockdown caused the opposite of these changes, and UHRF1 knockdown rescued these effects caused by HBP1 knockdown (S4A–S4C, S4E Fig). These data suggest that HBP1 also promotes the sensitivity of HCC cells to sorafenib-induced ferroptosis by inhibiting UHRF1 expression, thereby reducing the resistance of HCC cells to sorafenib.
UHRF1 epigenetically regulates the methylation status of the CDO1 promoter during ferroptosis UHRF1 is the “core protein” of epigenetic regulation and has been shown to bind DNMT1 to hemi-methylated DNA to maintain the stability of methylated DNA. UHRF1 often represses the transcription of target genes by regulating DNA methylation of the promoter of target genes [34]. Therefore, we next investigated whether UHRF1 epigenetically regulates the expression of a specific marker gene for ferroptosis, thereby regulating the process of ferroptosis. We overexpressed UHRF1 in HeLa cells to detect the mRNA expression of ferroptosis-related genes. As shown in Fig 5A, the mRNA expression of CDO1 decreased dramatically, while there were no significant changes in the expression of other mRNAs. In addition, the mRNA expression of CDO1 increased when UHRF1 was knocked down (Fig 5B), thus indicating that UHRF1 may epigenetically regulate the expression of the CDO1 gene. PPT PowerPoint slide
PNG larger image
TIFF original image Download: Fig 5. UHRF1 epigenetically regulates methylation status of CDO1 promoter during ferroptosis. (A) Quantitative RT-PCR showing expression of ferroptosis-related genes in UHRF1 overexpressed cells. (B) The mRNA expression of CDO1 in UHRF1 knockdown cells. (C) The protein expression of CDO1 increased in Erastin-induced cells. Hela, HepG2, and Huh7 cells were treated with 10 μM Erastin for 24 h; cells were collected to detected using western blotting. (D) UHRF1 overexpression decreases CDO1 protein expression. The protein levels of UHRF1 and CDO1 in cell lysate were measured by western blotting in Hela, HepG2, and Huh7 cells transfected with pCDNA3.1UHRF1 or pCDNA3.1 (as a control). (E) UHRF1 knockdown by shRNA increases CDO1 protein expression. The protein levels of UHRF1 and CDO1 in cell lysate were measured by western blotting in Hela, HepG2, and Huh7 cells stably transfected with pLL3.7-shUHRF1-1, pLL3.7-shUHRF1-2, or pLL3.7 (as a control) through lentiviral infection. (F) UHRF1 suppresses CDO1 promoter activity in a dose-dependent manner. Luciferase activity was detected in HEK293T cells cotransfected with the indicated reporter genes and UHRF1 plasmids. (G) Erastin promotes the suppression of UHRF1 on CDO1 promoter. HEK293T cells were cotransfected with CDO1 promoter and UHRF1 with or without 10 μM Erastin treatment. Luciferase activity was determined after transfection. (H) Erastin rescues UHRF1-induced specific CDO1 promoter hypermethylation. Methylation levels of CDO1 promoter were measured by MSP (right) in HeLa cells infected with pCDNA3.1 (as control) and pCDNA3.1-UHRF1. MSP using primer sets were listed in supplements. (I) CDO1’s promoter methylation by UHRF1 depends on the simultaneous presence of its hemi-mCpG and H3K9me2/3 binding activities. Several mutants were constructed as shown. (J, K) The protein and mRNA levels of CDO1 were detected in HeLa cells transfected with vector, UHRF1, UHRF1YP191/192AA, UHRF1YT478/479AA, and UHRF1YP191/192AA+UHRF1 YT478/479AA. (L) The occupation of UHRF1 on CDO1 promoter in HEK293T cells was measured by luciferase assays. The underlying data for Fig 5A, 5B, 5F, 5G, 5J and 5L can be found in S1 Data. Differences between 2 groups were calculated using a two-tailed Student t test. One-way ANOVA was performed to assess differences among multiple groups. Error bars represent SD. *, p < 0.05, **, p < 0.01, ***, p < 0.001. CDO1, cysteine dioxygenase 1; MSP, methylation-specific PCR; RT-PCR, real-time PCR; shRNA, short hairpin RNA; UHRF1, ubiquitin-like with PHD and RING finger domains 1.
https://doi.org/10.1371/journal.pbio.3001862.g005 CDO1, a non–heme iron metalloenzyme, transforms cysteine to taurine by catalyzing the oxidation of cysteine to its sulfinic acid [35,36]. A deficiency of cellular cysteine reduces GSH synthesis and impairs cellular antioxidant capacity, ultimately resulting in enhanced ROS levels and the induction of ferroptosis. Therefore, studies have confirmed that the overexpression of CDO1 promotes the process of ferroptosis, thereby inhibiting tumorigenesis [5]. We confirmed that CDO1 protein levels were increased during Erastin-induced ferroptosis in HeLa, HepG2, and Huh7 cells (Fig 5C). Western blotting further showed that the overexpression of UHRF1 in HeLa, HepG2, and Huh7 cells led to significant reductions in CDO1 expression, as compared with vector cells (Fig 5D). The down-regulation of UHRF1 expression led to an evident increase in CDO1 expression (Fig 5E). Next, we investigated whether UHRF1 regulates the activity of the CDO1 promoter in a manner that depended on DNA binding. Fig 5F showed that UHRF1 inhibited the activity of the CDO1 promoter in a dose-dependent manner. The combination of UHRF1 overexpression and Erastin treatment partially reversed the inhibitory effect of UHRF1 alone on the CDO1 promoter (Fig 5G). MSP assays were used to detect the effect of UHRF1 on the methylation of the CDO1 promoter during ferroptosis. As shown in Fig 5H, UHRF1 overexpression increased the methylation of the CDO1 promoter; the combination of UHRF1 overexpression and Erastin treatment reversed the increased methylation of the CDO1 promoter induced by UHRF1. These data suggest that UHRF1 may inhibit ferroptosis by increasing the methylation level of the CDO1 promoter region and by decreasing the expression level of CDO1. Next, we investigated the role of UHRF1 hemi-mCpG and H3K9me2/3 binding activities in regulating methylation of the CDO1 promoter. We constructed 2 mutants based on previous studies [37]. In one mutant, Tyr191 and Pro192 were changed to Ala (YP191/192AA); this mutant exhibited defective binding in H3K9me2/3. In the other mutant, Tyr478 and Thr479 were changed to Ala (YT478/479AA), thus abolishing hemi-mCpG-binding activity (Fig 5I). Analysis showed that the loss of H3K9me2/3 or hemi-mCpG binding activity in UHRF1 only partially affected the expression levels of mRNA and protein, or the activity of the CDO1 promoter. However, the SRA/TDD double mutant of UHRF1 had no effect on CDO1 mRNA and protein expression or promoter activity (Fig 5J–5L). These results suggest that UHRF1 methylation of the CDO1 promoter depends on the simultaneous presence of its hemi-mCpG and H3K9me2/3 binding activities, thus indicating that UHRF1 plays a role in ferroptosis by regulating the methylation of the CDO1 promoter. As a transcription factor, does HBP1 directly and specifically enhance CDO1 gene transcription, thereby inducing ferroptosis? S5 Fig showed that although HBP1 increased CDO1 mRNA level and had no effect on the mRNA levels of other ferroptosis-related genes (S5A Fig), HBP1 did not directly bind to the CDO1 promoter (S5B Fig), indicating that HBP1 transcriptionally enhance CDO1 gene expression through UHRF1 in ferroptosis.
The HBP1-UHRF1-CDO1 axis inhibits tumor cell proliferation and tumorigenesis It was previously reported that CDO1 silencing promotes the proliferation of non-small cell lung cancer (NSCLC) by limiting the metabolism of cysteine to the wasteful and toxic by-products CSA and sulfite (SO 3 2−) and by depleting cellular NADPH [38]. To verify whether CDO1 inhibits the proliferation of HeLa, HepG2, and Huh7 cells, we constructed CDO1 knockout cells using the CRISPR/Cas9 system. We designed an sgRNA to target the third exon of the CDO1 gene to prevent its transcription. CDO1 knockout inhibited PTEN protein levels and increased the protein levels of p-AKT without affecting the total AKT protein level (S6A Fig). MTT and EdU assays showed that the cells in which CDO1 had been knocked out proliferated at a higher rate than vector cells (Fig 6A and 6B); the overexpression of CDO1 yielded opposite results (Figs 6C, 6D, and S6B). These results suggest that CDO1 inhibits the proliferation of tumor cells by regulating the PTEN-PI3K-AKT signaling pathway. Recently, some research showed that PI3K-AKT-mTOR signaling suppresses ferroptosis via SREBP-mediated lipogenesis [39]. Therefore, we speculate that CDO1 may promote ferroptosis by regulating PTEN-PI3K-AKT signaling pathway in addition to directly inhibiting glutathione synthesis. We added Erastin to induce ferroptosis in HeLa, HepG2, and Huh7 cells overexpressing CDO1 and also added PTEN inhibitor SF1670. The results showed that SF1670 rescued the decreased cell viability caused by overexpressing CDO1 (S6C Fig). We proved that CDO1 can partially promote ferroptosis through PTEN-AKT signal pathway. PPT PowerPoint slide
PNG larger image
TIFF original image Download: Fig 6. The HBP1-UHRF1-CDO1 axis inhibits tumor cell proliferation and tumorigenesis. (A, B) CDO1 knockout promotes cell proliferation. Cells with CDO1 knockout were analyzed by MTT assay (A) and EdU assay (B). (C, D) CDO1 overexpression inhibits cell proliferation. Cells with CDO1 overexpression were analyzed by MTT assay (C) and EdU assay (D). (E, F) Stable transfected HeLa, HepG2, and Huh7 cells were analyzed by western blotting. β-actin was detected as a control. (G) HeLa cells were stably transfected with vector, HBP1, HBP1+UHRF1 or vector, shHBP1, shHBP1+shUHRF1, and growth rates of cells were measured by MTT assay. (H) Cell proliferation was examined by EdU incorporation assay. (I, J) HepG2 cells stably transfected with vector, HBP1, HBP1+UHRF1 or vector, shHBP1, shHBP1+shUHRF1 were subcutaneously injected into nude mice. The tumors were weighed, 4 weeks after injection. (K) GSH synthesis is the key to mediate HBP1-regulated ferroptosis. HeLa, HepG2, and Huh7 cells with HBP1 overexpression treated with Erastin (10 μM) and GSH (10 μM) for 24 h. HeLa, HepG2, and Huh7 cells with HBP1 knockdown treated with Erastin (10 μM) and BSO (10 μM) for 24 h. Cell viability was measured by MTT. (L) HBP1 inhibits tumor proliferation by inducing ferroptosis. HepG2 cells stably transfected with vector and HBP1 were subcutaneously injected into nude mice. Mice were injected daily with Liproxstatin-1 (10 mg/kg, IP) during the course of the experiment. The tumors were weighed, 4 weeks after injection. The underlying data for Fig 6A–6D, 6G, and 6H–6L can be found in S1 Data. Data were analyzed using a two-tailed, unpaired Student t test. One-way ANOVA was performed to assess differences among multiple groups. *, p < 0.05, **, p < 0.01, ***, p < 0.001. BSO, Butylamine-Sulfoximine-L; CDO1, cysteine dioxygenase 1; GSH, glutathione; HBP1, HMG box-containing protein 1; IP, intraperitoneal; UHRF1, ubiquitin-like with PHD and RING finger domains 1.
https://doi.org/10.1371/journal.pbio.3001862.g006 Next, we investigated the effect of HBP1-UHRF1 axis on CDO1 growth inhibition. We detected expression of related proteins in PTEN/PI3K/AKT pathway in HeLa, HepG2, and Huh7. In addition, HBP1 overexpression increased the levels of CDO1 protein, thereby increasing the expression of PTEN protein and decreasing the levels of p-AKT level; in contrast, the coexpression of UHRF1 rescued the HBP1-induced changes in PTEN and p-AKT expression (Figs 6E and S6D). Furthermore, the knockdown of HBP1 reduced the expression of CDO1, thereby reducing PTEN levels and increasing the levels of p-AKT; however, there was no effect when UHRF1 was also knocked down (Figs 6F and S6E). Consistent with protein expression data, MTT and EdU assays showed that UHRF1 rescued the HBP1-induced reduction in cell proliferation. HBP1 knockdown cells grew at a faster rate, while UHRF1 knockdown rescued the high growth rate induced by HBP1 knockdown (Fig 6G and 6H). These data suggest that the HBP1-UHRF1-CDO1 axis inhibits tumor cell proliferation via the PTEN/PI3K/AKT signaling pathway. To investigate the role of the HBP1-UHRF1-CDO1 axis during tumorigenesis in vivo, we subcutaneously injected 4-week-old BALB/c nude mice with HepG2 cells that stably expressed vector, HBP1 or HBP1+UHRF1, along with vector, shHBP1 or shHBP1+shUHRF1 individually. We found that UHRF1 reversed HBP1-induced tumor volume reduction (Fig 6I). HBP1 knockdown promoted the tumorigenic behavior of HepG2 cells; this was abolished by UHRF1 knockdown in HBP1 knockdown cells (Fig 6J). Next, we explored whether HBP1-UHRF1-CDO1 axis can inhibit tumor proliferation by inducing ferroptosis. Because CDO1 mainly promotes ferroptosis by inhibiting the synthesis of glutathione, we added Erastin to HeLa, HepG2, and Huh7 cells overexpressing HBP1 and then added exogenous GSH. After 24 h, the cell viability was tested. The results showed that the exogenous GSH supplementation reversed the decline of cell viability caused by HBP1 (Fig 6K, left panel). Then, we added Erastin to HeLa, HepG2, and Huh7 cells with HBP1 knockdown and then added Butylamine-Sulfoximine-L (BSO) to inhibit glutathione synthesis. Finally, we detected the cell viability. The results showed that BSO reversed the increase in cell viability caused by HBP1 knockdown (Fig 6K, right panel). In addition, we constructed HepG2 cell lines that simultaneously knocks out CDO1 and overexpresses HBP1 to detect its cell viability. The results showed that HBP1 enhanced Erastin-induced ferroptosis, and the induced ferroptosis could be rescued by Fer-1. However, when CDO1 was knocked out, the cell viability was mostly restored (S6F Fig) but not completely recovered, indicating that HBP1 mainly induces ferroptosis through CDO1 but may also induce a small part of ferroptosis through other ways. Taken together, these results suggest that HBP1 inhibits cell viability primarily by inhibiting glutathione synthesis. We further verified the above results in mice. We injected HepG2 cells stably overexpressing HBP1 subcutaneously in nude mice. After tumorigenesis, we injected ferroptosis inhibitor Liproxstatin-1 intraperitoneally into nude mice for treatment. The results showed that Liproxstatin-1 reversed the effect of HBP1 on tumor proliferation (Fig 6L). These results suggest that HBP1-UHRF1-CDO1 inhibits tumor growth by activating ferroptosis. Collectively, our results suggest that the HBP1-UHRF1-CDO1 axis inhibits tumor cell proliferation and tumorigenesis.
[END]
---
[1] Url:
https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3001862
Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.
via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/
