This article is mentioned in:

Introduction

Macrophage migration inhibitory factor (MIF) is a regulatory cytokine involved in the immune response, and as such plays an important role in the pathogenesis of autoimmune diseases and cancer (1,2). Findings have shown that CATT repeat number in the MIF promoter is associated with MIF expression level. The CATT5 repeat is the lowest expression allele and CATT8 is the highest (3), with the clinical severity of autoimmune inflammatory diseases and immune susceptibility being linked to higher CATT repeats (4). Thetranscription factor UHRF1 (90 kDa inverted CCAAT box-binding protein) can bind to CATT-repeat polymorphisms to regulate MIF expression, and is essential for the CATT5-8 length-dependent regulation of MIF transcription (5). UHRF1, as an epigenetic regulator, is overexpressed in cancer and coordinates gene silencing of tumor suppressors (6,7), potentially serving as a biomarker to differentiate among different tumor grades (8).

Evidence suggests that HMG-box protein 1 (HBP1) can bind to the MIF promoter and counter-regulate MIF expression (9). HBP1 is a member of the high mobility group (HMG) family of transcription factors and has been demonstrated to act as a transcriptional inhibitor in numerous cell lines, with its activation potently inhibiting the cell cycle and regulating related genes (10); therefore, it has been suggested to function as a tumor suppressor. Moreover, HBP1 maps to chromosome 7q31.1, which has been reported to be frequently deleted in myeloid and other cancers (11,12).

In the present study, we demonstrated that UHRF1 downregulates MIF expression by binding to the CATT repeat of the MIF promoter, and decreases HBP1 expression by promoting the interaction between MIF and HBP1 in T acute lymphoblastic leukemia (T-ALL). In addition, HBP1 negatively regulates MIF expression as a suppressor in T-ALL, and MIF knockdown prolongs the life of T-ALL mice, suggesting that MIF transcriptional regulation plays an important role in the pathogenesis of T-ALL and is a potential treatment target and biomarker for the prediction of prognosis in T-ALL.

Materials and methods

Study approval

The Ethics Committee of Shunde Hospital (Fo Shan) approved the use of discarded peripheral blood from T-ALL patients for T-cell cultivation. Informed consent for the procurement and analysis of these samples was also obtained.

Cells and reagents

The human Jurkat T-cell line was purchased from the American Type Culture Collection and cultured in RPMI-1640 medium supplemented with 10% fetal bovine serum. The apoptotic stains, PI and FITC-Annexin V (BD 556547), were obtained from BD Biosciences. The anti-HBP1 (sc-515281), β-actin (sc-47778), MIF (sc-271631) and anti-UHRF1 (ab57083) antibodies were purchased from Santa Cruz Biotechnology and Abcam, respectively. The Co-IP kit (26149 Pierce) was purchased from Pierce and the EasySep™ Human T Cell Isolation kit was obtained from Stem Cell (cat. no. 17951, Stem Cell).

Western blot analysis

Total protein obtained from Jurkat and PBMC from T-ALL patients was extracted using RIPA lysis and extraction buffer (Thermo Fisher; cat. no. 89900), and equal amounts (Pierce™ BCA Protein Assay, 23227) of 30 µg were resolved in 10% SDS-PAGE at 25 mA for 1 h on ice. The protein bands were subsequently transferred to PVDF (polyvinylidene difluoride) membrane, and non-specific sites were blocked with 5% BSA. The membranes were incubated overnight at 4°C with an anti-UHRF1 (ab57083, 1/1,000) antibody, washed three times with PBST, and incubated with an HRP-conjugated anti-rabbit secondary (ab205719, 1/2,000) antibody for 2 h at room temperature. ECL detection reagent (Pierce) was used to detect the protein complexes. Densitometric analysis was performed using NIH Image (version 1.62f). The MIF promoter CATTX length-dependent retention of protein was detected by western blotting of eluted binding proteins using an anti-UHRF1 or anti-HBP1 antibody (sc-515281, 1/1,000). The following oligos for CATT0,5-8 were used to bind UHRF1: 5′-CTTTCACCCAGCAGTATTAGTCAAT-3′ 5′-CTTTCACCCATTCATTCATTCATTCATTCAGCAGTATTAGTCAAT-3′ 5′-CTTTCACCCATTCATTCATTCATTCATTCATTCAGCAGTATTAGTCAAT-3′ 5′-CTTTCACCCATTCATTCATTCATTCATTCATTCATTCAGCAGTATTAGTCAAT-3′ 5′-CTTTCACCCATTCATTCATTCATTCATTCATTCATTCATTCAGCAGTATTAGTCAAT-3′

Flow cytometry

PI (propidium iodide) and FITC-Annexin V staining was performed to evaluate apoptosis in Jurkat cells and primary T cells purified from T-ALL patients by Ficoll-Hypaque and a CD3+ T-cell isolation kit following knockdown of UHRF1 or overexpression of HBP1. Staining was analyzed using a FACS Calibur (BD Biosciences).

Immunofluorescence confocal microscopy

PBS-rinsed cultured cells were fixed with methanol on ice for 2 h, blocked with 5% BSA for 1 h, and incubated with the primary antibodies (anti-HBP1, sc-515281; MIF, sc-271631; and anti-UHRF1, ab57083; 1/500) overnight at 4°C. The following day, the cells were rinsed five times with PBS and incubated with a fluorescently labeled secondary antibody in the dark for 2 h at room temperature. After rinsing, ProLong™ Gold Antifade Mountant with DAPI (P36931) was employed for nuclear staining. Imaging was performed on a Leica YSCC SP5 confocal system at a magnification of ×100.

Luciferase reporter assay analysis

MIF-794 CATT5-8-dependent transcription was analyzed using the dual luciferase reporter assay system as previously described (5). Each transfection experiment was performed in triplicate and repeated at least twice.

Co-immunoprecipitation

Cells (1×106) were transfected with empty vector or expression plasmid using Amaxa Nucleofector™. After 24 h, the cells were lysed in IP lysis buffer (Pierce, cat. no. 87787) containing protease inhibitors (Roche), and lysates were centrifuged at approximately 13,000 × g for 10 min to pellet the cell debris at 4°C and incubated with AminoLink™ Plus coupling resin (Pierce 26149) overnight at 4°C. The beads were washed three times with IP wash buffer (0.025 M Tris, 0.15 M NaCl, 0.001 M EDTA, 1% NP-40, 5% glycerol, pH 7.4), and the immunoprecipitates were eluted with elution buffer (DTT-containing SDS sample buffer) and boiled for 5 min in SDS loading buffer. Eluates were analyzed by western blot analysis.

Reverse transcription-quantitative PCR

Total RNA of Jurkat and PBMC of T-ALL patients was isolated using an RNeasy RNA extraction kit (Qiagen), and cDNA was synthesized using a BioRad iScript cDNA synthesis kit. RT-qPCR was carried out using the iQ SYBR-Green system (Bio-Rad). Primer sequences were: UHRF1: (5′-ATGTGGATCCAGGTTCGGA-3′ and 5′-GAACAGCTCCTGGATCTT-3′) and HBP1: (5′-TGAAGGCTGTGATAATGAGGAAGAT-3′ and 5′-CATAGAAAGGGTGGTCCAGCTTA′-3). MIF mRNA was determined using the primers: 5′-CGGACAGGGTCTACATCAA-3′, 5′-CTTAGGCGAAGGTGGAGTT-3′ and 18S 5′-GCAATTATTCCCCATGAACG-3′, 5′-TGTACAAAGGGCAGGGACTT-3′. The emitted fluorescence for each reaction was measured during the annealing/extension phase and relative quantity values were calculated by the standard curve method. The quantity value of 18S in each sample was used as a normalizing control. Differences were evaluated by non-parametric testing using the Mann-Whitney U test.

Jurkat and primary T-cell transfection

Primary T cells were isolated by EasySep™ Human T Cell Isolation Kit (Stem Cell, cat. no. 17951), then transfected using Nucleofector™ solution (Lonza, cat. no. VPA-1002) and the Nucleofector™ II system for transfecting T cells. After 24 h in an incubator at 37°C, the cells were harvested for further experimentation. The shRNA plasmids included UHRF1 (GI333964), HBP1 (TL312507), MIF (TR319111), or control (TR30007). The overexpression plasmid was HBP1 (RG202260) or control (PS100010) (Origene).

In vivo leukemia cell transplantation

NOD-SCID-γ (NSG) mice (8–10 weeks old) were obtained from Biocytogen (Beijing) provided with autoclaved food and clear H2O and housed in a specific pathogen-free (SPF) facility. Animal care was carried out in accordance with the local Animal Welfare Act. All food, water, bedding, and cages within the room were autoclaved or sterilized and cages were changed weekly; the room temperature was 26–28°C, and humidity was kept at 40–60%; with a 12-h light/dark cycle. A total of 36 mice were injected via the tail vein with 1×106 cells per mouse (T-ALL cells transfected with UHRF1-shRNA, HBP1-overexpression, or scrambled shRNA control plasmids). For survival experiments, mice (n=12 mice per group) were culled 25 days post-engraftment, or immediately following the appearance of signs of moribund or weight loss exceeding 10–15% of their total weight.

Statistical analysis

Results are expressed as the mean ± standard deviation. To study the difference between two groups, the Student's t-test and approximate calculation of normal distribution were used for all two-tailed comparisons. One-way ANOVA followed by Tukey's post-hoc test was used to compare more than two groups. The similarity of expression levels in the transcriptome was assessed by Pearson's correlation analysis. Expression heat map of genes were selected for 1.5-fold differential expression with an FDR <0.05 to show the different genes associated with MIF between ALL and healthy control. NSG mice survival was assessed via Kaplan-Meier survival curve. Analyses were performed using the GraphPad Prism software. P<0.05 was considered statistically significant, and P<0.01 was considered statistically very significant.

Results

Identification of HBP1 and UHRF1 binding to the MIF promoter and regulation of MIF transcription

To investigate whether HBP1 and UHRF1 can interact with the MIF-794 CATT5-8 microsatellite in T-ALL, 5′ biotin-labeled oligonucleotides, including or excluding CATT repeats of the MIF promoter (without any CATT sequences as a control), were incubated with nuclear lysates of human T cells followed by streptavidin beads. NaCl-eluted bound proteins were evaluated by western blot analysis (Fig. 1A). The effectiveness of this approach was verified by testing the CATT-specific interaction of UHRF1 in T cells. As shown in Fig. 1A, the analysis revealed binding of HBP1 to the MIF CATT8 and control CATT0 oligonucleotides. We previously demonstrated a downregulatory role of UHRF1 in-794 CATT5-8-dependent MIF expression (5); thus, we assessed the functional role of HBP1 in MIF expression by measuring the transcriptional activity of MIF in human T-ALL cells using a luciferase reporter assay. The level of MIF promoter transcript increased progressively with increasing levels of HBP1 shRNA, with CATT5 showing the lowest gene transcription and CATT8 showing the highest one (Fig. 1B).

Figure 1. (A) UHRF1 specifically binds to the CATT polymorphism of MIF but HBP1 does not. Western blot analysis reveals that UHRF1 binds to the CATT motif in the MIF promoter in a CATT length-dependent manner in Jurkat and primary T cells. HBP1, not only binds to promoter sequences including CATT, but also to those excluding CATT, indicating that HBP1 does not specifically bind to the CATT polymorphism. Jurkat or primary T cells were cultured in 6-well plates (1×106 cells/well) and transfected with UHRF1 or control shRNA for 48 h. DNA-bound UHRF1 and HBP1 were detected following incubation of nuclear lysates with 100 nM biotin-labeled 5′ CATT0-8 oligonucleotides spanning the MIF promoter (−865/-833 to −752). The 5′ CATT0-8 oligonucleotide-bound proteins were captured by streptavidin-bead absorption after a 3-h incubation at 4°C, and 1 mg was separated by SDS-PAGE and immunoblotted with anti-UHRF1 and anti-HBP1 antibodies. (B) HBP1 knockdown stimulates MIF expression. T cells were transfected with MIF promoter-luciferase reporter plasmids, treated with HBP1 or control shRNA, and cultured for 24 h. Prior to measurement of luciferase activity, cells were analyzed under basal conditions. Data are expressed as the mean ± SD of three measurements repeated twice (n=three measurements per experiment). **P<0.01 and *P<0.05 for control shRNA vs. HBP1 shRNA (Student's t-test, two-tailed).

Protein crosstalk between the transcription factor HBP1 and MIF

Co-immunoprecipitation (Co-IP) in T cells verified the interaction between HBP1 and MIF, but not between UHRF1 and HBP1 or between UHRF1 and MIF (Fig. 2A). Subsequently, the location of the interaction between HBP1 and MIF was confirmed in HeLa cells by confocal microscopy, showing co-localization in the cytosol but not in the nucleus (Fig. 2B). Moreover, there was no co-localization of UHRF1 and HBP1 (Fig. 2C).

Figure 2. Protein interaction between MIF and HBP1 in the cytosol. (A) Co-immunoprecipitation and immunoblotting showing protein interaction. Jurkat cells were cultured in 6-well plates (1×106 cells/well) and lysed in IP lysis buffer containing protease inhibitors. Lysates were centrifuged and incubated with AminoLink coupling resin (Pierce 26149) and anti-UHRF1, anti-HBP1, or anti-MIF antibodies for at least 4 h. The beads were subsequently washed four times with cold wash buffer and eluted with DTT-containing SDS sample buffer by boiling for 10 min. Protein samples were prepared in loading buffer containing 0.125 mM Tris-HCl (pH 6.8), 30% (v/v) glycerol, and 2% (w/v) deoxycholate and subjected to SDS-PAGE at 25 mA for 1 h on ice. (B) Confocal microscopy of UHRF1 and HBP1 in HeLa cells. (C) Magnification, ×100. UHRF1 localizes mainly to the nucleus but HBP1 localizes to both the nucleus and the cytosol in HeLa and Jurkat cells. HBP1 (green) and UHRF1 (red).

UHRF1 downregulates both MIF and HBP1 expression

Given that UHRF1 regulates MIF expression by binding to MIF CATT motifs, and HBP1 acts as a suppressor, we focused further attention on defining the relationship among the three genes. Following knockdown of UHRF1, HBP1, or both genes simultaneously, RT-qPCR results showed that UHRF1, not only downregulated MIF expression, but also HBP1 expression, causing loss of HBP1 repressive function in both the T-ALL cell line and T cells from ALL patients (Fig. 3A and B). It was further confirmed by western blotting that UHRF1 can downregulate HBP1 at the protein level (Fig. 3C and D). Subsequently, the effect of the overexpression of UHRF1 and HBP1 on MIF regulation was assessed. The MIF expression level was not significantly different following upregulation of UHRF1 or HBP1 (Fig. 4A and B), indicating that UHRF1 cannot upregulate MIF and HBP1 cannot downregulate MIF in T-ALL. The influence of the interaction between MIF and HBP1 on the downregulation of HBP1 mediated by UHRF1 was further evaluated. Knockdown of MIF by MIF shRNA increased HBP1 expression in T-ALL cells, which was also observed following UHRF1 knockdown (Fig. 5A and B). Moreover, UHRF1 knockdown could not downregulate HBP1 following inhibition of the MIF protein using an inhibitor (Fig. 5C).

Figure 3. UHRF1 downregulates both MIF and HBP1 in T-ALL cells. (A) RT-qPCR showing that HBP1 knockdown alone can increase MIF expression but UHRF1 knockdown can decrease both MIF and HBP1 expression in Jurkat cells. (B) RT-qPCR showing that HBP1 knockdown alone can increase MIF expression but UHRF1 knockdown can decrease both MIF and HBP1 expression in primary T cells from ALL patients (n=3). (C) Intracellular UHRF1 protein level following HBP1 knockdown as analyzed by western blotting of cell lysates. β-actin served as a loading control. UHRF1 protein expression levels were no different following HBP1 knockdown. (D) Lower HBP1 protein levels were observed following treatment with UHRF1 shRNA. Data are expressed as the mean ± SD of three measurements repeated twice (n=three measurements per experiment). **P<0.01, by Student's t-test (two-tailed). Displayed blots are representative of three independent experiments.

Figure 4. UHRF1 cannot upregulate either MIF or HBP1, and HBP1 overexpression does not downregulate MIF in T-ALL cells. (A) RT-qPCR showing that increased UHRF1 cannot increase MIF or HBP1 expression in Jurkat cells. (B) RT-qPCR showing that HBP1 upregulation cannot decrease MIF expression in primary cells from T-ALL patients. Data are expressed as the mean ± SD of three measurements repeated twice (n=three measurements per experiment). **P<0.01, by Student's t-test (two-tailed). Displayed blots are representative of three independent experiments.

Figure 5. UHRF1 regulates HBP1 expression by promoting the interaction between HBP1 and MIF in T-ALL cells. (A) RT-qPCR showing that MIF negatively regulates HBP1 expression, but not UHRF1 expression, following MIF knockdown in Jurkat cells. (B) RT-qPCR showing that MIF negatively regulates HBP1 expression, but not UHRF1 expression, following MIF knockdown in cells from T-ALL patients (n=3). (C) UHRF1 knockdown cannot decrease HBP1 expression in the absence of MIF in cells from T-ALL patients. Data are expressed as the mean ± SD of three measurements repeated twice (n=three measurements per experiment). **P<0.01, by Student's t-test (two-tailed). Displayed blots are representative of three independent experiments.

MIF silencing induces cell apoptosis and slows leukemia progression in vivo

We also examined cell apoptosis and animal survival following the regulation of MIF expression by UHRF1 and HBP1, which induced apoptosis, retarded the progression of ALL, and extended survival time. As expected, knockdown of UHRF1 or overexpression of HBP1 in T-ALL cells reduced basal MIF expression (Fig. 6A). MIF overexpression is known to inhibit apoptosis in many cell types, and the two mediators act in concert to regulate apoptotic sensitivity in the context of inflammatory activation (10–13). Experimental reduction of UHRF1 or upregulation of HBP1 enhanced T-ALL cell sensitivity to apoptosis, which is consistent with the interpretation that functional UHRF1 and HBP1 regulate MIF expression and protect cells from apoptosis (Fig. 6B). To further assess the functions of UHRF1 and HBP1 in the progression of T-ALL in vivo, mice were injected with transduced Jurkat cells to observe survival time. The results show that mice transplanted with UHRF1-knockdown cells lived longer than those in both the control and HBP1-overexpression groups, which is consistent with the cell apoptosis data (Fig. 6C).

Figure 6. MIF knockdown by its regulators causes apoptosis and significantly increases survival in vivo. (A) Primary human T-ALL cells extracted using a lymphocyte separation kit were transfected with control, UHRF1 shRNA, or HBP1 overexpression plasmids for 24 h. Cells were subsequently washed once with PBS, stained with apoptotic markers, and subjected to FACS analysis. Right lower quadrant: early apoptotic cells; left upper quadrant: dead cells; left lower quadrant: live cells; right upper quadrant: late apoptotic cells. All apoptosis experiments were carried out at least three times. One representative experiment is shown. (B) Peripheral blood leukocytes were isolated from T-ALL patients (n=3) and transfected with control, UHRF1 shRNA, or HBP1 overexpression plasmids for 24 h. Cells were subsequently harvested, and the intracellular MIF was measured by flow cytometry. Data represent three individuals per group. **P<0.01 unpaired t-test. (C) Kaplan-Meier survival curve of NSG mice transplanted with HBP1, UHRF1, or empty vector-transfected Jurkat cells. **P<0.01 (log-rank test as compared with the control group).

Pathogenic role of MIF

Human genetic studies indicate a high expression of MIF and UHRF1 and low expression of HBP1 in T-ALL, and experimental data suggest a pathogenic role of MIF in promoting proliferation and downstream expression of chemokines in T-ALL (Fig. 7A). Notably, a significant correlation was observed between the mRNA expression levels of UHRF1 and MIF (R=0.9192, P<0.0001) and HBP1 and MIF (R=0.6977, P<0.0001) in the T-ALL group as compared with those in the healthy control group (Fig. 7B). This correlation between the expression levels of UHRF1 and MIF and between HBP1 and MIF supports a functional role of UHRF1 downregulation and HBP1 upregulation with respect to MIF in T-ALL.

Figure 7. (A) The difference in expression of UHRF1, HBP1, MIF, and MIF-related cytokines between T-ALL patients and healthy controls. Expression heat map of genes selected for 1.5-fold differential expression with an FDR <0.05. Data represent three healthy individuals in comparison with three patients with T-ALL cells. (B) Correlation between UHRF1 and MIF and between HBP1 and MIF, but not between UHRF1 and HBP1 in ALL. Correlation plots for UHRF1 vs. MIF, UHRF1 vs. HBP1, and HBP1 vs. MIF mRNA expression in RNA samples obtained from patients with ALL (n=66) or healthy controls (n=20). The expression scores were calculated from the mean of the normalized values. P<0.0001 for the mean MIF expression in ALL patients (12.81±0.29) vs. healthy controls (9.37±0.15) by a two-tailed Student's t-test. **P<0.01 unpaired t-test.

Discussion

Macrophage migration inhibitory factor (MIF) has been suggested to be a pro-tumorigenic factor that promotes the proliferation, migration, and invasion of tumor cells (13). Previous findings have shown that ALL cells constitutively express high levels of MIF (14). Leukemic cells from most patients express the chemokine IL-8 and the receptor CXCR1, but at lower levels. Moreover, one report used a mouse model in which subcutaneous ALL tumors were partially suppressed by locally injected endothelial IL-8 (15–17). In the present study, T cells were isolated from healthy controls and T-ALL patients and subjected to microarray. These data are consistent with reports that MIF expression is high and causes increased expression of downstream chemokines such as IL-8, which are associated with cell proliferation, suggesting that MIF plays a pathogenic role in T-ALL. Our mechanistic understanding of the MIF-mediated regulation of tumor cell proliferation has expanded since the identification of MIF transcription. Compelling evidence suggests that MIF overexpression and regulation is associated with, and contributes to, the pathogenesis of inflammatory autoimmune and malignant diseases (18,19); however, the mechanism underlying MIF regulation in T-ALL has yet to be clarified.

Evidence suggests that UHRF1 positively regulates MIF transcription (5) and HBP1 has a negative regulatory function (9). To determine the key regulatory mechanism of MIF in T-ALL, we identified that UHRF1 and HBP1 co-regulate MIF expression. Of note, UHRF1 can also regulate HBP1 transcription by promoting the interaction between MIF and HBP1 proteins. We verified a specific association between MIF and HBP1 by co-immunoprecipitation of MIF-HBP1 complexes in vitro. To confirm that MIF can interact with intracellular HBP1, we showed the co-localization of endogenously expressed MIF and HBP1 in the cytosol. We also observed that UHRF1 can only regulate MIF, but not HBP1 without the presence of MIF. Moreover, HBP1 was able to only upregulate MIF expression but could not downregulate MIF expression when HBP1 expression was elevated, indicating that UHRF1 is the key regulator in the knockdown of MIF in T-ALL.

Furthermore, we found that the expression levels of UHRF1 and MIF were elevated but that of HBP1 was decreased in T-ALL patients as compared with those in healthy controls. In addition, analysis of the correlation among UHRF1, MIF, and HBP1 expression in a gene expression dataset of T cells from T-ALL patients shows a high correlation between UHRF1 and MIF expression and between HBP1 and MIF expression, supporting a positive regulatory role of UHRF1 and a negative regulatory role of HBP1 in MIF transcription in vivo. Moreover, UHRF1 knockdown and HBP1 overexpression induced a greater level of apoptosis in T cells from T-ALL patients and significantly prolonged the survival time of transplanted mice.

Taken together, our results indicate an important role of the UHRF1 protein in the survival and homing of malignant T cells, which is mediated through a functional interaction between MIF and HBP1. In conclusion, a high level of UHRF1 and a low level of HBP1 cause MIF overexpression, resulting in tumor cell proliferation and inhibition of cell death. The MIF/UHRF1/HBP1 axis may represent a novel target for the therapeutic intervention of ALL.

Acknowledgements

Not applicable.

Funding

We would like to thank the National Natural Scientific Foundation of China (81770148, to J. Yao) for funding.

Authors' contributions

JY contributed to the conception of the study. JY and CZ analyzed data for the study. JY, CZ, HH, YL and XZ performed the experiments for the study. JY and YL wrote the study. All authors approved the study, and JY and YSL confirm the authenticity of the data.

Availability of data and materials

All data generated or analyzed during this study are included in this article.

Ethics approval and consent to participate

The Ethics Committee of Shunde Hospital (Fo Shan) approved the use of discarded peripheral blood from T-ALL patients for T-cell cultivation. Informed consent for the procurement and analysis of these samples was also obtained.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Glossary

Abbreviations

Abbreviations:

References