Abstract

   Chemotherapeutic drugs, such as cisplatin (cis-dichlorodiamineplatinum
   [II], cDDP) and 5-fluorouracil (5Fu), are widely used in transarterial
   chemoembolization (TACE), which is a standard therapy for patients with
   hepatocellular carcinoma (HCC). Chemoresistance is a major cause of
   TACE treatment failure in HCC patients. Our previous studies have
   identified the expression levels of miR-101 responsive genes, such as
   EED, EZH2, STMN1 and JUNB, exhibit significant correlation with the
   occurrence and progression of HCC, while the role of miR-101 responsive
   gene signatures in the chemoresistance of HCC treatment remains
   unclear. In this study, we identified ubiquitin-coupled enzyme E2D1
   (UBE2D1) as a crucial regulatory factor in the chemoresistance of HCC,
   which is a direct target of miR-101 and exhibits significant
   correlation with miR-101-responsive gene signatures. The bioinformatics
   analysis showed the expression of UBE2D1 was significantly increased in
   HCC tissues and was closely correlated with the poor prognosis. In
   addition, we analyzed the role of miR-101/UBE2D1 axis in regulating
   chemo-sensitive of HCC cells. Our results showed that miR-101 increases
   the DNA damage and apoptosis of HCC cells by inhibiting the expression
   of UBE2D1, which in turn increases the sensitivity of HCC cells to cDDP
   and 5Fu both in vitro and in vivo. Therefore, simultaneous assessment
   of miR-101 and UBE2D1 expression levels might provide an effective
   approach in preselecting HCC patients with survival benefit from TACE
   treatment. Moreover, further elucidation of the underlying molecular
   mechanisms of the miR-101/UBE2D1 axis could provide novel insight for
   targeted therapy of HCC.

   Subject terms: Cancer, Biomarkers

Introduction

   Based on the most recent statistical data from 2020, primary liver
   cancer is identified as the third most common cause of cancer-related
   deaths worldwide, with hepatocellular carcinoma (HCC) representing the
   majority of cases at a rate of 75%–85% [[36]1]. The predominant risk
   factors for HCC exhibits regional variation. In China, Hepatitis B
   virus (HBV) is a significant contributor to HCC morbidity, whereas in
   western countries, HCC incidence is primarily linked to hepatitis C
   virus (HCV) infection and alcohol consumption [[37]2]. Early-stage HCC
   can be effectively managed with a variety of treatment modalities;
   however, a significant proportion of patients, ~70%–80%, are diagnosed
   at an unresectable advanced stage, resulting in poor survival rates
   [[38]3, [39]4]. Transarterial chemoembolization (TACE) is considered a
   standard treatment option for patients with HCC, providing a viable
   alternative for individuals who are not candidates for surgical
   intervention or serving as an adjunctive therapy following surgery to
   mitigate the risk of recurrence [[40]5, [41]6].

   TACE initiates cellular apoptosis and ischemic necrosis within liver
   tumors by administering embolic agents and chemotherapeutic agents
   through arterial routes. Anthracyclines, platinum and fluorouracil are
   frequently employed chemotherapeutic agents in TACE procedures [[42]5,
   [43]7]. Cisplatin (cis-dichlorodiamineplatinum [II], cDDP) and other
   platinum compounds demonstrate broad-spectrum efficacy and are widely
   used in the treatment of over 80% of malignancies [[44]8]. The
   structural features of cDDP result in its ability to impede DNA
   replication and disrupt transcription, leading to the induction of DNA
   damage. The subsequent delayed repair of this damage frequently
   initiates the cellular apoptosis pathway [[45]9, [46]10].
   5-Fluorouracil (5Fu) is commonly utilized in the treatment of diverse
   solid tumors through the inhibition of deoxythymidine synthase, thereby
   impacting intracellular DNA synthesis. Nevertheless, the potential for
   diminished efficacy, whether through inherent or acquired resistance,
   presents a significant obstacle to the optimal utilization of this
   chemotherapeutic agent [[47]11]. Various mechanisms of drug resistance
   exist, including alterations in apoptosis-induced signaling proteins
   and the tumor microenvironment (TME), as well as the involvement of
   microRNAs (miRNAs) [[48]12–[49]14].

   MiRNAs are a subset of small non-coding RNAs typically consisting of
   20–25 nucleotides. Within eukaryotic organisms, miRNAs function by
   binding to the 3′ untranslated region (3′UTR) of target messenger RNAs
   (mRNAs) post-transcriptionally, thereby modulating a wide array of
   biological processes including but not limited to cell development,
   infection, immunity, carcinogenesis and chemoresistance [[50]15,
   [51]16]. The impact of miRNAs on chemoresistance is associated with its
   ability to enhance drug efflux, interfere with cell cycle progression,
   and facilitate apoptosis [[52]17]. Previous study has shown that
   miR-101 can be functioning as a tumor suppressor or a facilitator of
   cancer cell apoptosis via targeting crucial oncogenes and
   anti-oncogenes. Its efficacy in impeding the proliferation of various
   cancer types, including HCC, has been empirically validated [[53]18].
   In HCC, studies have shown that miR-101 inhibits tumorigenicity and
   participates in cellular life activities by targeting genes such as
   Mcl-1, EED, EZH2, and VEGF [[54]18–[55]22]. The specific mechanism by
   which miR-101 impacts DNA damage repair and drug resistance in HCC
   remains incompletely understood. As such, we hypothesized that miRNAs
   and their associated signaling pathways may play a role in the
   resistance of HCC to TACE.

   The ubiquitination pathway is also a key factor in regulating cell
   proliferation, apoptosis and DNA damage repair [[56]23]. Ubiquitination
   plays a critical role in mediating the cellular response to DNA
   double-strand breaks (DSBs), with delayed repair potentially resulting
   in genomic instability and tumorigenesis. The ubiquitin-conjugating
   enzyme E2D1 (UBE2D1), a member of the E2 family, is involved in various
   cancer-related signaling pathways through its ubiquitination
   modification mechanism [[57]24]. Previous studies have shown that
   UBE2D1 can mediate the ubiquitination of tumor suppressor protein p53,
   KRAS, TNFα-NFκB [[58]25–[59]29] and other genes to participate in
   important pathways of carcinogenesis. UBE2D1 is believed to play a role
   in DNA damage repair, with increased levels of IL-6 in the serum of HCC
   patients triggering the DNA damage response and genomic instability,
   resulting in the repeated amplification of UBE2D1 gene copies in the
   genome [[60]30]. Additionally, a previous study has shown that UBE2D1
   aggregates at DSB sites and works in conjunction with RNF138 to promote
   CtIP ubiquitination and accumulation, thereby enhancing the efficiency
   of DNA repair [[61]24]. Therefore, UBE2D1 may serve as a pivotal
   regulator of chemoresistance in HCC.

   This study has established a significant correlation between UBE2D1 and
   miR-101-responsive gene signatures, demonstrating the role of the
   miR-101/UBE2D1 axis in mediating resistance to cDDP and 5Fu by
   enhancing DNA damage repair mechanisms. Our initial exploration of the
   oncogenic potential of UBE2D1 in HCC revealed that UBE2D1 is directly
   regulated by miR-101, influencing DNA damage repair and apoptosis
   processes, ultimately leading to resistance to both cDDP and 5Fu in HCC
   cells. In conclusion, targeting UBE2D1 combined with miR-101 may prove
   to be an effective strategy for overcoming multidrug resistance during
   chemotherapy for HCC.

Results

Screening and identification of UBE2D1 as a potential candidate is closely
associated with miR-101-responsive gene signatures

   MiR-101 is thought to exhibit tumor suppressive properties in multiple
   types of malignancies. To verify its role in HCC, we analyzed the
   expression of miR-101 using clinical data obtained from the TCGA-LIHC
   cohort and the [62]GSE6857 dataset. Both datasets consistently showed
   that the expression of miR-101 in LIHC tissues was dramatically reduced
   when compared with normal tissues (Fig. [63]1A). The TCGA dataset
   contained 49 pairs of LIHC tissues and matched normal samples, and
   [64]GSE6857 contained 48 paired samples, and the expression of miR-101
   in these tumor tissues showed an overall downward trend relative to
   corresponding normal tissues (Fig. [65]1B). Moreover, there was a
   significant correlation observed between miR-101 expression levels, WHO
   grade, and survival outcomes in HCC patients. A higher grade of
   malignancy was found to be associated with a greater prevalence of low
   miR-101 expression, while the converse was true for lower degrees of
   malignancy (Fig. [66]1C). Patients exhibiting low levels of miR-101
   expression experienced a significantly reduced survival duration in
   comparison to those with elevated levels (Fig. [67]1D). The above
   findings confirm the independent prognostic significance of miR-101 in
   HCC. The expression of EED, EZH2, STMN1 and JUNB [[68]31] was closely
   associated with miR-101 and GEPIA analysis revealed that the elevated
   expression of these miR-101-responsive gene signatures was correlated
   with poor prognosis in HCC patients (Fig. [69]1E). Further research is
   needed to explore the potential clinical significance of molecules with
   characteristics similar to miR-101-responsive gene signatures in the
   diagnosis and prognosis of HCC patients.

Fig. 1. Expression of miR-101 in HCC patients and screening and
identification of its downstream target gene UBE2D1.

   [70]Fig. 1
   [71]Open in a new tab

   A Expression analysis of miR-101 in normal hepatic tissues (N) and LIHC
   tumor tissues (T) was conducted using data from the TCGA database and
   the [72]GSE6857 dataset. B The differential expression of miR-101
   between normal and paired LIHC tissues was analyzed using data from the
   TCGA database and the [73]GSE6857 dataset. C Distribution of high,
   medium and low expression of miR-101 in HCC patients with different
   malignant degrees based on TCGA data. D The correlation between miR-101
   expression levels and the overall survival of HCC patients was assessed
   using the Log-rank test. E The correlation between miR-101-responsive
   gene signatures (EED, EZH2, STMN1, JUNB) expression levels and the
   overall survival of HCC patients was derived from GEPIA. F Overlap of
   the predicted downstream targets of miR-101 from TargetScan and miRDB
   prediction databases. G The overlapping 595 predicted genes were
   subjected to KEGG analysis and enriched to the ubiquitin-mediated
   proteolysis pathway. The pathways were categorized based on the −log10
   P value interval. H Fold-change differences with
   miR-101^low/miR-101^high based on TCGA data, UBE2D1 showed the most
   pronounced alteration in expression. I Correlation analysis of UBE2D1
   and miR-101-responsive gene signatures in different tumors according to
   GEPIA. J A positive correlation was observed between the expression
   levels of UBE2D1 and EED, EZH2, STMN1, JUNB in LIHC tissues based on
   TCGA data. K A negative correlation was observed between the expression
   levels of UBE2D1 and miR-101 in LIHC tissues based on TCGA data.
   Statistical significance was determined by Student’s t-test.
   ****p < 0.0001.

   Next, we utilized TargetScan and miRDB to identify potential downstream
   target genes regulated by miR-101 (Fig. [74]1F). A total of 595 novel
   underlying molecular targets of miR-101 in the intersection were used
   for KEGG analysis. The ubiquitin-mediated proteolysis pathway exhibited
   enrichment in eight ubiquitination-related molecules, specifically
   UBE2D1, UBE2A, CUL4B, UBE2F, UBE2D2, FBXW11, FBXW7 and CUL3 (Fig.
   [75]1G, H). Based on the mean cutoff value of miR-101 expression, we
   compared the expression changes of these ubiquitinating factors between
   HCC patients with high miR-101 levels (named as miR-101^high) and low
   miR-101 levels (named as miR-101^low). The log2 ratio of miR-101^low to
   miR-101^high was utilized to determine the fold change in the
   expression levels of miR-101 between groups exhibiting high and low
   miR-101 expression. UBE2D1 demonstrated the most pronounced alteration
   in expression (Fig. [76]1H). Following correlation analysis, the
   relationship between UBE2D1 and miR-101-responsive gene signatures in
   LIHC was found to be the second highest among various types of tumors
   (Fig. [77]1I). The evidence presented also indicated a positive
   correlation between the expression levels of UBE2D1 and
   miR-101-responsive gene signatures (Fig. [78]1J). However, the
   expression levels of miR-101 was negatively correlated with UBE2D1 in
   LIHC tissues (Fig. [79]1K). Therefore, it is worth exploring whether
   UBE2D1 is regulated by miR-101 and involved in HCC progression.

Abnormally high expression of UBE2D1 is closely associated with poor
prognosis of HCC patients in multiple clinical cohorts

   To investigate the potential oncogenic role of elevated UBE2D1
   expression in HCC, we utilized the TCGA-LIHC cohort and the
   [80]GSE14520 dataset to merge clinical data with genomic information on
   UBE2D1 for subsequent analysis. Our findings revealed a statistically
   significant increase in UBE2D1 expression levels in LIHC tissues
   compared to adjacent tissues, indicating a consistent upward trend
   relative to paired normal tissues (Fig. [81]2A, B). Furthermore, there
   was a significant correlation between UBE2D1 expression and the WHO
   grade of patients with HCC, with a greater prevalence of elevated
   UBE2D1 expression in cases of higher malignancy, and conversely, a
   lower incidence in cases of lower malignancy (Fig. [82]2C). In order to
   investigate the potential correlation between the expression levels of
   UBE2D1 and the clinical prognosis of patients with HCC, survival
   analysis was conducted. The results indicated that elevated levels of
   UBE2D1 expression in tumor tissues were significantly linked to
   decreased overall survival rates in HCC patients (Fig. [83]2D). In
   conclusion, abnormally high expression of UBE2D1 can be used as a poor
   prognostic factor for the overall survival in HCC patients.

Fig. 2. Abnormally high expression of UBE2D1 is closely associated with poor
prognosis of HCC patients.

   [84]Fig. 2
   [85]Open in a new tab

   A Analysis of UBE2D1 expression in normal hepatic tissues (N) and LIHC
   tumor tissues (T) was conducted using data from the TCGA database and
   the [86]GSE14520 dataset. B The differential expression of UBE2D1
   between normal and paired LIHC tissues was analyzed by utilizing data
   from both the TCGA database and the [87]GSE14520 dataset. C
   Distribution of high, medium and low expression of UBE2D1 in HCC
   patients with varying degrees of malignancy based on TCGA data. D
   Overall survival rate of HCC patients from the TCGA database was
   analyzed based on UBE2D1 mRNA levels using the Log-rank test. E
   Survival curve of 367 LIHC patients enrolled in TCGA database. Patients
   were classified into four groups based on the expression levels of
   miR-101 and UBE2D1 mRNA in their tumors. F The expression patterns of
   miR-101 and UBE2D1 mRNA in 367 LIHC samples and the statistical
   analysis of patient survival outcomes. Statistical significance was
   determined by Student’s t-test. ****p < 0.0001.

   To facilitate further data analysis, we employed appropriate critical
   values to divide LIHC tumor samples into high and low miR-101
   expression groups, as well as high and low UBE2D1 expression groups,
   and then combined into four groups: miR-101^highUBE2D1^high,
   miR-101^highUBE2D1^low, miR-101^lowUBE2D1^high, miR-101^lowUBE2D1^low
   for survival comparison purposes. As hypothesized, patients classified
   within the miR-101^highUBE2D1^low group demonstrated the most favorable
   prognosis, whereas those categorized in the miR-101^lowUBE2D1^high
   group exhibited the poorest prognosis (Fig. [88]2E). Divided by four
   quadrants, the highest survival rate was observed in the
   miR-101^highUBE2D1^low group whereas the lowest survival rate was found
   in the miR-101^lowUBE2D1^high group (Fig. [89]2F). The observed
   expression patterns of miR-101 and UBE2D1 in HCC patients exhibit a
   strong correlation with overall survival rates, thereby offering
   valuable insights for prognostic prediction within this specific
   patient cohort.

Knockdown of UBE2D1 enhances sensitization of HCC cells to chemotherapeutic
drug-induced cell death

   In order to investigate the function of UBE2D1 in various HCC cell
   lines, we established stable UBE2D1-knockdown cell lines in SNU-739 and
   HCC-LM3 cells through lentiviral transduction with two UBE2D1-specific
   shRNAs, while control cells were transduced with an empty pLKO.1 vector
   lentivirus. The results indicated that specific shRNAs were able to
   effectively knockdown endogenous UBE2D1 protein in SNU-739 and HCC-LM3
   cell lines (Fig. [90]3A). The cell proliferation was analyzed by CCK8
   assay, revealing a proliferation inhibitory effect on SNU-739 and
   HCC-LM3 cells via reduction of UBE2D1 expression (Fig. [91]S1A). To
   evaluate the impact of UBE2D1 on chemosensitivity in HCC, we employed
   cDDP and 5Fu at the minimal concentrations that effectively inhibited
   cell proliferation to treat stable transfected cell lines (Fig.
   [92]S1B). In both SNU-739 and HCC-LM3 cells, the knockdown of UBE2D1
   resulted in a notable enhancement in the susceptibility of the cells to
   chemotherapeutic agents, as evidenced by reduced rates of cell
   proliferation and colony formation (Fig. [93]3B–E). In addition,
   knockdown of UBE2D1 resulted in a notable increase of cleaved caspase 3
   protein, a reliable marker of apoptosis (Fig. [94]3F). Similarly, flow
   cytometry analysis showed that UBE2D1 knockdown significantly increased
   the apoptosis rate after treatment with cDDP or 5Fu (Fig. [95]3G).

Fig. 3. Knockdown of UBE2D1 enhances sensitization of HCC cells to
chemotherapeutic drug-induced cell death.

   [96]Fig. 3
   [97]Open in a new tab

   A Protein levels identification of shRNA-mediated UBE2D1 knockdown
   effect in SNU-739 and HCC-LM3 cells. B, C Cell viability was analyzed
   in stable UBE2D1-knockdown cell lines following treatment of 0.5 μM
   cDDP for 0, 24, 48 and 72 h respectively and treatment of 0.5 μM 5Fu
   for 0, 24, 48 and 72 h respectively by CCK8 assay. D, E Cell
   proliferation ability detection of stable UBE2D1-knockdown cell lines
   after treatment with 0.5 μM cDDP or 0.5 μM 5Fu by plate clone formation
   assay. F A western blotting analysis of cleaved caspase 3 protein
   levels in UBE2D1-knockdown cells with or without the cDDP (50 μM, 24 h)
   or 5Fu (50 μM, 48 h) treatment. G Apoptosis rate detection of
   UBE2D1-knockdown cells after treatment with cDDP or 5Fu (5 μM, 24 h) by
   flow cytometry. All experiments were performed three times. Statistical
   significance was determined by Student’s t-test. ****p < 0.0001,
   ***p < 0.001, **p < 0.01, *p < 0.05.

High expression of UBE2D1 attenuates sensitization of HCC cells to
chemotherapeutic drug-induced cell death

   To investigate the impact of UBE2D1 overexpression on chemotherapy
   resistance, SNU-739 and HCC-LM3 cells were infected with lentivirus
   carrying either control or UBE2D1-overexpressing constructs. Following
   western blotting analysis, our findings indicated that HCC cells
   exhibited successful upregulation of UBE2D1 (Fig. [98]4A). The efficacy
   of chemotherapy agents was markedly reduced following the upregulation
   of UBE2D1 in SNU-739 and HCC-LM3 cells, as indicated by enhanced
   cellular proliferation and colony formation (Fig. [99]4B–E). In
   addition, the overexpression of UBE2D1 resulted in a significant
   reduction in caspase 3 cleavage and apoptosis rate after treatment with
   cDDP or 5Fu (Fig. [100]4F, G).

Fig. 4. High expression of UBE2D1 attenuates sensitization of HCC cells to
chemotherapeutic drug-induced cell death.

   [101]Fig. 4
   [102]Open in a new tab

   A Protein levels identification of UBE2D1 overexpression effect in
   SNU-739 and HCC-LM3 cells. B, C Cell viability was assessed in stable
   UBE2D1-overexpression cells following treatment of 0.5 μM cDDP for 0,
   24, 48, and 72 h respectively and treatment of 0.5 μM 5Fu for 0, 24,
   48, 72 h respectively by CCK8 assay. D, E The cell proliferation
   ability of stable UBE2D1-overexpression cells was assessed after
   treatment with 0.5 μM cDDP or 0.5 μM 5Fu using plate clone formation
   assay. F A western blotting analysis of cleaved caspase 3 protein
   levels in stable UBE2D1-overexpression cells with or without the cDDP
   (50 μM, 24 h) or 5Fu (50 μM, 48 h) treatment. G Apoptosis rate of
   stable UBE2D1-overexpression cells was assessed by flow cytometry after
   treatment with cDDP or 5Fu (5 μM, 24 h). All experiments were performed
   three times. Statistical significance was determined by Student’s
   t-test. ****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05.

UBE2D1 attenuates DNA damage induced by chemotherapy drugs in HCC cells

   To investigate whether UBE2D1 induces chemoresistance by attenuating
   DNA damage and inhibiting cell apoptosis, we performed DNA damage
   assays. The formation of nuclear γ-H2AX foci at DNA damage sites is the
   earliest marker of DNA damage occurs [[103]32]. In order to evaluate
   this sensing phenomenon, western blotting and immunofluorescence
   staining were performed, and the results revealed that reduced UBE2D1
   significantly increased the occurrence of γ-H2AX after treatment with
   cDDP and 5Fu (Fig. [104]5A, B). The comet assay, employed for
   evaluating the levels of DNA damage, revealed a correlation between
   decreased levels of UBE2D1 and an increase in DNA damage, with a
   notable accumulation of damage observed when treated with cDDP and 5Fu
   (Fig. [105]5C). In contrast, the overexpression of UBE2D1 led to a
   decrease in DNA damage, consequently diminishing the susceptibility of
   cells to chemotherapeutic agents. This was supported by the decreased
   presence of γ-H2AX and shortened tailing induced by DNA damage (Fig.
   [106]5D–F).

Fig. 5. UBE2D1 attenuates DNA damage induced by chemotherapy drugs in HCC
cells.

   [107]Fig. 5
   [108]Open in a new tab

   A Western blotting analysis of γ-H2AX protein levels in HCC cells
   following stable knockdown of UBE2D1, with or without treatment of cDDP
   or 5Fu (10 μM, 2 h). B Representative images and quantification of
   γ-H2AX foci in stable UBE2D1-knockdown cells following treatment with
   or without cDDP or 5Fu (10 μM, 2 h). γ-H2AX accumulates at DNA
   double-strand break sites (scale bar = 10 μm). The statistical scatter
   points represent the number of γ-H2AX foci in each cell, which were
   counted by Image J software. C The comet assay was employed to evaluate
   the extent of DNA damage in stable UBE2D1-knockdown cells treated with
   or without cDDP or 5Fu (10 μM, 12 h). The head of the comet as a
   spherical mass of undamaged DNA, while the damaged DNA (DNA loops
   around strand breaks) emanates from the head in the form of a tail
   (scale bar = 150 μm). D Western blotting analysis of γ-H2AX protein
   levels in HCC cells following stable overexpression of UBE2D1, with or
   without the cDDP or 5Fu treatment (10 μM, 2 h). E Representative images
   and quantification of γ-H2AX foci in stable UBE2D1-overexpression cells
   following treatment with or without cDDP or 5Fu (10 μM, 2 h) (scale
   bar = 10 μm). F Comet assay was performed on stable
   UBE2D1-overexpression cells, with or without treatment of cDDP or 5Fu
   (10 μM, 12 h) (scale bar = 150 μm). All experiments were performed
   three times. Statistical significance was determined by Student’s
   t-test. ****p < 0.0001, ***p < 0.001, **p < 0.01.

UBE2D1 is identified as a de novo target of miR-101 in HCC cells

   As shown the predicted binding site between the 3ʹUTR of UBE2D1 and
   miR-101 according to the TargetScan online tool (Fig. [109]6A). In
   order to investigate the potential targeting of UBE2D1 by miR-101,
   plasmids containing either matching or mutant 3ʹUTR sequences of UBE2D1
   were generated and transfected into HEK293 cells along with miR-101
   mimics. Subsequent luciferase reporter assays revealed a significant
   reduction in relative luciferase activity of the reporter gene
   containing the wild-type UBE2D1-3ʹUTR in SNU-739 and HCC-LM3 cells upon
   transfection with miR-101 mimics. Conversely, no significant change in
   relative luciferase activity was observed in cells transfected with the
   mutant UBE2D1-3ʹUTR reporter gene (Fig. [110]6B). Moreover, the results
   of western blotting and qRT-PCR analysis revealed that miR-101 mimics
   could significantly suppress UBE2D1 protein and mRNA expression in
   SNU-739 and HCC-LM3 cells (Fig. [111]6C, E). In cells transfected with
   miR-101 inhibitors, the opposite result was observed (Fig. [112]6D, F).

Fig. 6. UBE2D1 is identified as a de novo target of miR-101 in HCC cells.

   [113]Fig. 6
   [114]Open in a new tab

   A The predicted binding sites between the wild-type (WT) and mutant
   (MUT) 3ʹUTR of UBE2D1 and miR-101. B WT UBE2D1 3ʹUTR and MUT UBE2D1
   3ʹUTR firefly luciferase activities were measured in HEK293T cells
   transfected with NC mimics or miR-101 mimics. C, D Western blotting
   analysis of UBE2D1 expression in SNU-739 and HCC-LM3 cells transfected
   with miR-101 mimics or miR-101 inhibitors. E, F The miR-101 and UBE2D1
   transcript levels were quantified by qRT-PCR in cells transfected with
   miR-101 mimics or miR-101 inhibitors. Graph represents miR-101 and
   UBE2D1 mRNA levels, expressed as fold changes relative to the control.
   All experiments were performed three times. Statistical significance
   was determined by Student’s t-test. ****p < 0.0001, ***p < 0.001,
   *p < 0.05, ns nonsignificant.

The miR-101–UBE2D1 axis enhances HCC chemosensitivity by increasing DNA
damage

   To explore the functional significance of miR-101 in cDDP and 5Fu
   resistance in HCC, we constructed SNU-739 and HCC-LM3 cells that stably
   overexpressed miR-101 and identified the UBE2D1 expression changes by
   western blotting and qRT-PCR analysis (Fig. [115]7A, B). the results of
   CCK8 assays and plate clone formation assays indicated that
   overexpression of miR-101 could inhibit the proliferation of SNU-739
   and HCC-LM3 cells, while the rescue expression of UBE2D1 reversed the
   inhibition effect induced by miR-101 after treatment with cDDP and 5Fu
   (Fig. [116]7C–F).

Fig. 7. The MiR-101–UBE2D1 axis enhances HCC chemosensitivity by increasing
DNA damage.

   [117]Fig. 7
   [118]Open in a new tab

   A Western blotting analysis of UBE2D1 expression changes in SNU-739 and
   HCC-LM3 cells following stable transfection of miR-101 and rescue
   expression of UBE2D1 on this basis. B The miR-101 and UBE2D1 transcript
   levels were quantified by qRT-PCR in miR-101 overexpression cells and
   rescued cells. C, D Cell proliferation of HCC cells was analyzed in the
   rescue experiments following treatment of 0.5 μM cDDP or 0.5 μM 5Fu for
   0, 24, 48 and 72 h respectively by CCK8 assay. E, F The cell
   proliferation of HCC cells was analyzed in the rescue experiments after
   treatment with 0.5 μM cDDP or 0.5 μM 5Fu using plate clone formation
   assay. G Cleaved caspase 3 expression levels were analyzed using
   western blotting in the rescue experiments performed on SNU-739 and
   HCC-LM3 cells, with or without administration of cDDP (50 μM, 24 h) or
   5Fu (50 μM, 48 h). H Cell apoptosis rate was detected by flow cytometry
   in the rescue experiments with the cDDP or 5Fu treatment (5 μM, 24 h).
   I, J Detection of γ-H2AX DNA damage foci in the rescue experiments of
   SNU-739 and HCC-LM3 cells by western blotting and immunofluorescent
   staining (scale bar = 10 μm). The statistical scatter points represent
   the quantification of γ-H2AX foci per cell. K Comet assay was performed
   on the rescue experiments with or without administration of cDDP or 5Fu
   (10 μM, 12 h) (scale bar = 150 μm). All experiments were performed
   three times. Statistical significance was determined by Student’s
   t-test. ****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05.

   Moreover, miR-101 overexpression combined with cDDP or 5Fu treatment
   considerably potentiate the expression of cleaved caspase 3 protein in
   SNU-739 and HCC-LM3 cells. In the rescue experiments, the rescue
   expression of UBE2D1 in miR-101 overexpression cells attenuated the
   expression of cleaved caspase 3 protein (Fig. [119]7G). Similarly,
   apoptosis rates were assessed by flow cytometry, which showed the same
   trends (Fig. [120]7H). The formation of γ-H2AX foci was detected by
   western blotting and immunofluorescence staining (Fig. [121]7I, J), as
   well as the comet assay (Fig. [122]7K) all showed that miR-101 enhanced
   chemotherapy-induced DNA damage via regulation of UBE2D1. These results
   suggest that the downregulation of UBE2D1 by miR-101 is able to
   sensitize HCC cells to chemotherapeutic drug-induced apoptosis.

The miR-101–UBE2D1 axis regulates tumor growth and chemoresistance in vivo

   The xenograft tumors of HCC were established to further investigate the
   function of UBE2D1 in vivo and its impact on the development of
   chemoresistance. UBE2D1 stable-knockdown cells of SNU-739 mixed with
   matrigel were injected subcutaneously into nude mice (male mice aged 5
   weeks old; Beijing Vital River Laboratory Animal Technology Co., Ltd.,
   China). When the tumor volume reached ~150 mm^3, cDDP treatment
   (2 mg/kg/3 days) was initiated. As shown, tumor volume and tumor weight
   were significantly reduced in both the sh-UBE2D1 group and
   control + cDDP group compared to the control group. Notably, the tumor
   volume and weight of the shUBE2D1 + cDDP group were the lowest among
   all groups (Fig. [123]8A–C), indicating that UBE2D1 knockdown enhanced
   the inhibitory effect of cDDP on tumor growth. We further detected the
   expression levels of γ-H2AX and Ki67 in the tumors using IHC staining.
   The results showed that knockdown of UBE2D1 significantly enhanced DNA
   damage and inhibited cell proliferation activity, as indicated by the
   percentage of γ-H2AX and Ki67-stained positive cells compared with the
   control group (Fig. [124]8D, E). Tunel staining showed that UBE2D1
   inhibition significantly increased chemotherapy-induced apoptosis (Fig.
   [125]8F).

Fig. 8. The MiR-101–UBE2D1 axis regulates tumor growth and chemoresistance in
vivo.

   [126]Fig. 8
   [127]Open in a new tab

   A Representative tumor images in nude mice after the injection of
   SNU-739 cells that stably expressing UBE2D1-shRNA and control scramble
   shRNA, with or without the cDDP treatment (2 mg/kg/3 days). B, C Growth
   curves of xenograft tumors in nude mice and quantification of tumor
   weight. The initiation time of drug administration is recorded as 0 d.
   D–F Representative images of γ-H2AX, Ki67 and tunel staining in tumor
   samples. On the right is the result of statistical analysis on positive
   signals using Image J Profiler and GraphPad Prism (scale bar = 20 μm).
   All experiments were performed three times. Statistical significance
   was determined by Student’s t-test. ***p < 0.001, *p < 0.05, ns
   nonsignificant.

Discussion

   TACE chemotherapy has been suggested as an initial treatment option for
   both primary and recurrent cases of liver cancer, with the potential to
   enhance patient survival outcomes to differing extents [[128]33]. The
   methodology employed not only integrates the use of chemotherapy agents
   like cDDP and 5Fu for the eradication of tumors, but also initiates
   tumor vascular embolization and subsequent necrosis in order to impede
   or postpone tumor progression. Nevertheless, the persistence of drug
   resistance-induced tumor relapse and metastasis poses significant
   challenges to the prognosis of patients with liver cancer, with
   therapeutic shortcomings frequently stemming from a myriad of genetic
   and epigenetic modifications, as well as subsequent disruption of
   crucial genes and protein networks governing the fate of tumor cells
   [[129]34, [130]35]. A thorough understanding of the tumor biological
   pathways underlying liver cancer, combined with the identification of
   genes responsive to chemotherapy agents to enhance drug sensitivity,
   will facilitate the implementation of precision medicine for patients
   and improve therapeutic outcomes in liver cancer. However, a limitation
   of our research is the exclusion of the TACE mouse model, owing to the
   challenges associated with its establishment.

   MiRNA comprises only about 1% of the total genes in the human genome,
   yet it plays a regulatory role in ~60% of protein-coding genes,
   demonstrating significant functional potency [[131]36]. To date, over
   50% of human miRNAs have been found associated with carcinogenesis,
   suggesting that abnormal miRNA expression profiles could serve as a
   promising indicator of malignancy [[132]11]. Among these, miR-101 is
   acknowledged as a tumor suppressor miRNA in a range of malignant
   tumors, as it demonstrates the capacity to modulate multiple oncogenes
   and play a role in the modulation of drug resistance across various
   cancer types [[133]37–[134]41]. In the context of liver cancer
   progression, a significant downregulation of miR-101 expression was
   observed in LIHC tissues relative to adjacent tissues, resulting in
   dysregulation of tumor suppressor pathways and correlation with disease
   stage. Deletion of miR-101 has been implicated in the acceleration of
   liver cancer development. Various external and internal factors,
   including atmospheric particles, viruses and pro-inflammatory
   cytokines, may contribute to alterations in miR-101 levels [[135]42].
   Therefore, our study sought to investigate the potential biological
   relevance of miR-101 in tumorigenesis, building upon prior research
   conducted by our group that established miR-101’s ability to suppress
   malignant characteristics in HCC cells by concurrently targeting
   oncogenes such as STMN1 (proliferation), JUNB (angiogenesis) and CXCR7
   (migration/invasion) [[136]31]. However, the precise mechanisms by
   which miR-101 modulates chemoresistance in HCC remain to be elucidated.

   First, bioinformatics predictions were generated using TargetScan and
   miRDB online tools, followed by validation through molecular biology
   experiments. Our findings indicated that the ubiquitination factor
   UBE2D1 is a direct target gene negatively modulated by miR-101 at both
   the transcriptional and post-transcriptional levels in SNU-739 and
   HCC-LM3 cells. Importantly, a significant association was observed
   between UBE2D1 and miR-101-responsive gene signatures. Overexpression
   of UBE2D1 has been shown to be associated with poor prognosis in
   several human cancers, such as breast cancer, liver cancer, bladder
   cancer and lung adenocarcinoma [[137]43–[138]46]. These findings are
   consistent with the LIHC sample data information from the TCGA
   database. In addition, UBE2D1 had also been found to be overexpressed
   in precancerous lesions, and it was only upregulated in cancer cells
   rather than hepatitis cells, suggesting its potential as an early-stage
   HCC marker [[139]30]. Subsequently, we investigated the impact of
   miR-101 and UBE2D1 on the proliferation of HCC cells following
   chemotherapy through a series of gain and loss of function experiments
   conducted both in vitro and in vivo. Silencing UBE2D1 using shRNA was
   found to enhance the sensitivity of HCC cells to cDDP and 5Fu, as
   evidenced by a significant reduction in cell viability and an increase
   in apoptosis. To further elucidate the role of UBE2D1 in
   miR-101-mediated chemosensitivity, we reintroduced UBE2D1 into
   miR-101-overexpressing cells and treated them with chemotherapeutic
   agents. The upregulation of miR-101 resulted in a significant
   inhibition of tumor cell proliferation and colony formation in vitro.
   Importantly, the inhibitory effects of miR-101 were found to be
   reversed by UBE2D1.

   DNA damage is a crucial anti-cancer mechanism of action for most
   chemotherapy drugs. For instance, cDDP and 5Fu are commonly employed in
   TACE therapy, which can directly or indirectly induce DNA damage,
   induce apoptosis, and inhibit cell proliferation [[140]5, [141]47]. DNA
   damage was identified through the detection of γ-H2AX using western
   blotting and immunofluorescence staining, which revealed the products
   formed following the occurrence of DNA damage. Additionally, based on
   the alteration of the physicochemical properties of damaged DNA, we
   assessed the extent of DNA damage with a comet experiment [[142]48].
   The results showed that downregulation of UBE2D1 and overexpression of
   miR-101 resulted in an increase in the population of HCC cells
   displaying DNA damage, as evidenced by the presence of distinct nuclear
   foci of γ-H2AX and migration of damaged DNA. This effect was
   consistently observed in two HCC cell lines, SNU-739 and HCC-LM3. In
   conclusion, miR-101 has been identified as a promoter of DNA damage by
   targeting and binding ubiquitinated proteins, which subsequently
   activates the ubiquitination pathway and modulates the sensitivity of
   HCC cells to cDDP and 5-fluorouracil treatments. However, the
   involvement of UBE2D1 in DNA damage repair has been previously
   documented. This suggests the potential existence of additional, yet
   undiscovered mechanisms that merit further investigation, particularly
   those elucidating the intricate processes of DNA damage repair.

   MiRNA has long been advocated as a biomarker for early diagnosis and
   therapeutic efficacy assessment according to Chinese guidelines
   [[143]49–[144]51]. The miRNA panel consisting of miR-122, miR-192,
   miR-21, miR-223, miR-26a, miR-27a and miR-801 has high accuracy in
   diagnosing early HBV-associated HCC, with a 30% increase in sensitivity
   compared to AFP levels alone [[145]51]. The miR-101/UBE2D1 signaling
   axis plays a crucial role in promoting genomic alterations in oncogenes
   during HCC carcinogenesis. Simultaneous evaluation of miR-101 and
   UBE2D1 could potentially serve as valuable biomarkers or therapeutic
   targets for predicting prognosis and disease severity in HCC. A
   comprehensive understanding of the molecular mechanisms involved in the
   regulation of the miR-101/UBE2D1 axis provides new perspectives for the
   diagnosis and targeted treatment of HCC, suggesting the potential
   benefit of combining UBE2D1 inhibition with conventional chemotherapy.

Materials and methods

Clinical data analysis and bioinformatics prediction

   Curated datasets obtained from The Cancer Genome Atlas (TCGA) and Gene
   Expression Omnibus (GEO) were used for gene expression analysis,
   correlation analysis and overall survival (OS) analysis. For the
   prediction of canonical target genes of miR-101, TargetScan and miRDB
   prediction databases were used. In order to conduct further screening,
   KEGG pathway enrichment analysis was performed using online analysis
   tools DAVID and KOBAS, with both the screened P value and corrected P
   value being <0.01.

Cell culture and drug treatment

   For human HCC cell lines, SNU-739 and HCC-LM3 were cultured in
   RPMI-1640 medium (GIBCO BRL, Grand Island, NY, USA) supplemented with
   10% FBS (GIBCO BRL) and 100 mg/ml penicillin–streptomycin at 37 °C in a
   5% CO[2] incubator. Similarly, the basal medium of HEK293T cells was
   changed to DMEM (GIBCO BRL) under the same culture conditions. All
   three cell lines were purchased from ATCC (Manassas, VA, USA). All cell
   lines were tested without mycoplasma contamination. For
   chemotherapeutic drugs, cDDP and 5Fu were purchased from Selleckchem
   catalog #S1166 and #S1209, respectively.

Plasmid construction and lentiviral transduction

   The short-hairpin sequences against human UBE2D1 (Table [146]S1) were
   designed by Sigma-Aldrich. The synthesized primers were annealed and
   cloned into the pLKO.1-TRC vector (Addgene, Cambridge, USA; 10878) to
   obtain plasmids for stable knockdown of UBE2D1 (shUBE2D1). The UBE2D1
   overexpression plasmid was constructed by cloning the open reading
   frame of the cDNA into the polyclonal sites of pCDH vector, and the
   primers used were shown in Table [147]S2. A pLenti6.3/V5-DEST vector
   was used to construct a plasmid for rescuing UBE2D1. The pCDH-miR-101
   plasmid was retained in our laboratory.

   When HEK293T cells reached 60%–80% confluence, the target plasmid,
   packaging plasmid psPAX2 (Addgene, 12260) and envelope plasmid pMD2.G
   (Addgene, 12259) were mixed at a mass ratio of 4:3:1 and transfected
   using PEI reagent (Invitrogen) to form a virus packaging system.
   Viruses were collected twice after 48 and 72 h, centrifuged at 1000 rpm
   for 5 min, and purified with 0.45 μm syringe filter. Subsequently,
   stably transfected cells were established, SNU-739 and HCC-LM3 were
   incubated with lentivirus and medium supplemented with 10 μg/ml
   polybrene for 24 h in 6-well plates when 40%–50% fusion was reached.
   The transduced cells were screened with 2 μg/ml puromycin for 3 days,
   and when there were no surviving cells in the blank treatment group.
   The selected cells maintained a low concentration of puromycin during
   culture and identified by western blotting.

Quantitative real-time PCR (qRT-PCR)

   Total RNA was extracted from cells using Trizol Reagent (Invitrogen),
   followed by reverse transcription into cDNA using PrimeScript RT
   reagent (Takara). Detection of gene expression was performed using
   qRT-PCR method with SYBR Premix Ex Taq (Takara). The UBE2D1 primer
   sequences utilized were as follows: forward primer:
   5′-TAGCGCATATCAAGGTGGAGT-3′, reverse primer: 5′-TGGTGACCATTGTG
   ACCTCAG-3′. β-actin and U6 were used for the normalization of UBE2D1
   and miR-101, respectively. Fold change of expression was calculated as
   2^−^ΔΔCt.

Western blotting

   Cells were lysed in RIPA lysis buffer (added protease and phosphatase
   inhibitors), and the protein concentration in the lysate was determined
   using a BCA assay kit. Protein lysates were resolved by SDS-PAGE with
   10%–15% gel concentration based on the quality of the protein and
   transferred to nitrocellulose membranes. The membranes were blocked in
   5% skim milk prepared in TBST buffer for 1 h and then incubated with
   primary antibodies against UBE2D1 (1:1000, Proteintech, 11677-1-AP),
   caspase 3 (1:1000, Cell Signaling Technology, #9662), γ-H2AX (1:2000,
   Abmart, [148]M63324) and β-actin (1:2500, Cell Signaling Technology,
   #3700) overnight at 4 °C. Membranes were washed and incubated with
   HRP-conjugated secondary antibodies for 1 h at room temperature. The
   labeled bands were exposed to the ECL chemiluminescent reagent and
   imaged using a chemiluminescent imaging system.

CCK8 assay

   The proliferation of stable transfected SNU-739 and HCC-LM3 cell lines
   were detected via CCK8 assay. Each cell type was cultured in 96-well
   plates with 1 × 10^3 cells/well in triplicate with continuous 0.5 μM
   drug treatment or no drug treatment, while setting up blank wells.
   Proliferative activity was measured every 24 h and maintained for 4
   days. At the designated time, each well was added 10 μl CCK8 reagent
   and placed in 37 °C incubator for 1–4 h. The absorbance was then
   determined with a microplate reader at 450 nm. In the drug sensitivity
   test, SNU-739 and HCC-LM3 cells were treated with drugs at gradient
   concentrations, and the absorbance was detected after 48 h. Cell
   proliferation curves were created and analyzed using Excel and GraphPad
   Prism software.

Colony formation assay

   Each group was seeded with 1000 cells into 6-well plates by adjusting
   the density of the single-cell suspension. After the cells were grown
   overnight, the medium containing 0.5 μM drug was replaced and continued
   for 1–2 weeks. Until a single cell was cloned into a colony of more
   than 50 cells, the cell clones were washed with PBS and fixed with 4%
   paraformaldehyde, followed by staining with 0.5% (w/v) crystal violet
   dye (Sigma-Aldrich). Images of cell clones were captured by the Odyssey
   scanner (LI-COR, Lincoln, NE, USA) and then counted using the Image J
   software.

Measurement of apoptosis

   CDDP and 5Fu were added to the cell medium at a final concentration of
   5 µM to treat the cells. After 24 h of incubation, cells were collected
   and stained with Annexin V/PI (Sigma-Aldrich) for 15 min at room
   temperature. Flow cytometry was used to carry out early and late cell
   apoptosis analysis.

Immunofluorescence staining

   Slides were sterilized prior to use, cells were grown on slides and
   treated with or without 10 μM cDDP or 5Fu for 2 h. Then the slides were
   washed and fixed with 4% paraformaldehyde. After permeabilized with
   0.5% Triton X-100 for 15 min, slides were blocked for 1 h with 1% (w/v)
   BSA in PBS. This was followed by overnight incubation with primary
   γ-H2AX antibody (1:400, Abmart, [149]M63324), and incubated with goat
   anti-mouse secondary antibody conjugated with FITC for 1 h. The nuclei
   were stained with DAPI (Sigma-Aldrich) for a duration of 5 min at room
   temperature. Expression levels of γ-H2AX were evaluated based on the
   number of positively stained foci per cell.

Comet assay

   Cells were treated with or without 5 μM cDDP or 5Fu for 12 h, and
   samples were collected through digestion and centrifugation. According
   to the instructions, 1% normal melting point agarose gel and 0.7% low
   melting point agarose gel were first prepared. Mixing 10,000 cells into
   a low melting point agarose gel and spread evenly on a slide, which is
   pre-coated with the first layer of normal melting point gel, and the
   third layer of low melting point agarose gel as protection. After cell
   lysis, the DNA was uncoiled under alkaline buffer, and the single-cell
   gel electrophoresis (~0.75–1 V/cm) should be performed in an ice bath
   and protected from light. Then the DNA was neutralized with neutral
   buffer, stained with PI, and DNA tailing was visualized under red
   fluorescence by fluorescence microscopy.

Dual-luciferase reporter assay

   A fragment of the UBE2D1 3′UTR containing the putative binding site of
   miR-101 was subcloned into pGL3-Enhancer Vector containing the firefly
   luciferase gene to construct the UBE2D1 wild-type (WT) or mutant (MUT)
   reporter plasmid. The phRL-TK vector containing Renilla luciferase gene
   was used as an internal control. HEK293 cells were seeded in 48-well
   plates at 70% cell confluence, followed by co-transfected with the
   above luciferase plasmids (5 ng of plasmids encoding Renilla luciferase
   and 100 ng of plasmids containing firefly luciferase reporters) and
   20 ng miR-101 mimics using Lipofectamine 2000. 48 h after transfection,
   the cells were lysed and luciferase activity was detected using the
   Dual-Luciferase Reporter assay kit (Promega). Reporter luciferase
   activity was normalized to its respective Renilla luciferase activity
   for each transfected well. All values were expressed as fold change
   over control.

MiRNA and transfection

   MiR-101 mimics were chemically synthesized and purified by GenePharma
   (Shanghai, China), which mimic naturally occurring miR-101 after
   transfection of cells. The synthesized miR-101 inhibitors specifically
   inhibited the function of miR-101 when transfected into cells. The
   sense strand sequences of miR-101 mimics, miR-101 inhibitors and their
   negative controls are listed in Table [150]S3. According to the
   manufacturer’s instructions, 20 μM miR-101 mimics or miR-101 inhibitors
   were transfected into cells with liposome reagent 2000 in a 1:1 ratio.

Treatment of xenograft with cDDP

   The animal experiments conducted in this study were approved by the
   Committee for Experimental Animal Use and Care of the Fourth Military
   Medical University. All experimental procedures adhered to the National
   Guidelines for Animal Experimentation. The 5-week-old male nude mice
   used in this experiment were obtained from the Experimental Animal
   Center of the Fourth Military Medical University. The xenograft models
   were generated by injecting SNU-739-shUBE2D1 or control cells
   subcutaneously in the flank of nude mice. A suspension of 5 × 10^6
   tumor cells in 100 μl PBS was mixed with 100 μl matrigel. When the
   tumor volume reached 150 mm^3 and grow normally, mice were randomly
   divided into four groups of three mice. For drug intervention, mice
   were administered an intraperitoneal injection of 2 μM of the
   chemotherapeutic drug cDDP for every 3 days. Mouse body weight and
   tumor size were recorded every 3 days during the period. The tumor
   volume was calculated according to the formula:
   V = (length × width^2)/2. Following which animals were euthanized, the
   tumors were harvested and weighed, then fixed with 4% paraformaldehyde
   for histological analyses.

Immunohistochemistry

   The fresh tumor samples were fixed in neutral buffered formalin for
   24 h at room temperature and then embedded and processed according to
   standard protocols. The sections underwent deparaffinized through
   graded ethanol solutions and rehydration. After an antigen retrieval
   procedure of 30 min using antigen retrieval solution, the sections were
   stained with Ki67, γ-H2AX and TUNEL. The stained slices are scanned by
   the Servicebio company.

Statistics

   Experiments were repeated at least three times. Statistical analyses
   were performed using GraphPad Prism 8 software. The data in statistical
   graphs were presented as mean ± standard deviation (SD). Unless
   otherwise specified, group comparisons were assessed using unpaired
   Student’s t-test or one-way ANOVA. Survival curves were analyzed using
   the Log-rank test. The statistical significance of differences was
   determined at a threshold of P < 0.05.

Supplementary information

   [151]Full and uncropped western blots^ (10.5MB, docx)
   [152]Supplemental Material^ (209.6KB, docx)

Acknowledgements