Mixed lineage leukaemia histone methylases 1 collaborate with ERα to regulate HOXA10 expression in AML

Homeobox-containing genes are key players in embryogenesis and development. Misregulation of homeobox genes is associated with tumourigenesis. More than 200 homeobox-containing genes have been identified in vertebrates, and they have been classified into two major groups, class I and II which encode transcription factors regulating pattern formation, differentiation and proliferation [1]. There is considerable evidence linking deregulated HOX expression with human acute leukaemia. This includes translocations targeting specific HOX genes, particularly those located at 5′ in their clusters (for example, HOXA9), or myeloid–lymphoid leukaemia, which normally functions to maintain proper HOX levels [2].

HOXA10 is a homeodomain transcription factor that is expressed in the most primitive haematopoietic cell compartment, and overexpression of HOXA10 leads to impaired lymphoid development in vitro. Moreover, overexpression of HOXA10 in murine bone marrow induces a myeloproliferative disorder characterized by expansion of the committed myeloid progenitor population (common granulocyte/monocyte progenitors or GMP) [3]. This myeloproliferative disorder evolving to AML (acute myeloid leukaemia) over time in HOXA10-overexpressing mice, demonstrates that HOXA10 acts as an important regulator of haematopoiesis, governing both proliferation and differentiation of hematopoietic progenitor and stem cells, where distinct fate outcomes depend on the HOXA10 concentration [4–6]. Although HOXA10 is a critical player in differentiation of myeloid cells and leukaemiagenesis, little is known about its own regulation.

HMTs (histone methyltransferases) are key enzymes that post-translationally methylate histones and play critical roles in gene expression, epigenetics and cancer. MLLs (mixed lineage leukaemia histone methylases) are human HMTs, such as MLL1, MLL2, MLL3 and MLL4 that specifically methylate histone H3K4 (H3 at lysine 4) and are linked with genes [7]. MLL gene is a frequent target for recurrent chromosomal translocations found in approximately 10% of human acute leukaemias, which can be defined as AMLs, ALLs (acute lymphoblastic leukaemias) [8]. Studies have demonstrated that several MLLs act as coregulators for E2-mediated activation of E2-sensitive genes. Some MLL-like methyltransferase complexes have been shown to directly interact with site-specific DNA-binding transcription factors [9,10].

Moreover, previous studies have showed there are ERE on promoter of HOXA10 gene [14], which means that ER possibly is regulator of HOXA10. We were interested in determining whether MLL act as coregulator for HOXA10 expression. In the present study, we report that MLL1 act as coregulator of ERα to regulate expression of HOXA10 in an epigenetic way.

HL-60, THP-1 leukaemia cell lines were maintained in RPMI 1640 containing 10% (v/v) FBS. The cells treated with a freshly prepared solution of the cells were grown up to 70% confluency and treated with different concentrations (0–1000 nM) of E2 at different time. The cells were then harvested and subjected to either RNA and protein extraction or ChIP (chromatin immunoprecipitation) assay.

For treatment of cells with antisense oligonucleotides, cells were grown up to 60% confluency in 60 mm culture flasks and transfected with varying amounts (3–9 μg) of different antisense oligonucleotides. Briefly, cocktails of different amounts of antisense oligonucleotide and transfection reagents (ifect, MoleculA) were made in the presence of 300 μl of the culture medium (without supplements) by incubating for 30 min, as instructed by the manufacturer. Cells were washed twice with the supplement-free culture medium, and finally submerged in 1.7 ml of medium (without supplements). The antisense oligonucleotide/transfection reagent cocktail was applied to the cells and incubated for 7 h before the addition of 2 ml of culture medium with all supplements and 20% activated charcoal-stripped FBS. The cells were then incubated for an additional 24 h before being treated with E2.

Cells were washed and lysed with guanidine isothiocyanate solution. Total RNA was extracted using RNA Extraction Kit for Culture Cells (Takara) following the instruction manual. Primers for β-actin were chosen specifically to cross one intron in the β-actin gene. In the presence of contaminating genomic DNA, additional larger bands would be amplified; the lack of amplification of the larger band was used as a control to rule out contamination with any genomic DNA. The sequences and PCR cycling conditions were listed in Table 1.

Genomic DNA was extracted from leukaemia cell lines and the whole blood of leukaemia patients and healthy volunteers with the QIAamp DNA Blood Mini Kit (Qiagen, Valencia, CA, U.S.A.) following the manufacturer's instructions. Sodium bisulfite modification of genomic DNA was performed with EZ DNA Methylation kit (Zymo Research) following the manufacturer's instructions, in which 2 μg of original genomic DNA was eluted in 20 μl of M-Elution buffer. The bisulfite-converted genomic DNA was amplified by MSP using the primer sets listed in Table 1 to analyse the promoter CpG methylation of HOXA10. MSP was performed in a volume of 25 μl containing 5 pmol of a primer set, 1.0 μl of bisulfite-converted DNA solution in M-Elution Buffer, and contents of ExTaq HS DNA Polymerase kit (TaKaRa) at the suggested concentrations (including 0.625 units of ExTaq HS DNA Polymerase per 25 μl of reaction mixture, 1×ExTaq Buffer [Mg²⁺ plus] and 250 μM of each dNTP).

The whole-cell lysate was prepared in 1 ml of ice-cold buffer [50 mM Tris/HCl (pH 7.4), 150 mM NaCl, 5 mM EDTA, 0.5% (v/v) NP40 (Nonidet P40), 1 mM phenylmethylsulfonylfluoride, 10 μg/ml aprotinin, 10 μg/ml leupeptin]. The lysate was rotated 360° for 1 h at 4°C followed by centrifugation at 12 000× g for 10 min at 4°C to clear the cellular debris. Proteins were quantified using the BCA protein assay kit (Pierce Chemical Co.). Nuclear protein was extracted using Qproteome Nuclear Subfractionation Kit (Qiagen) by following the manufacturer's instructions. Nuclear protein was resolved on SDS/PAGE gels and transferred to nitrocellulose membranes, and Western blot analyses were performed using antibodies of ERα and ERβ (R&D Systems), anti-histone H3 (di-methyl K4) antibody(abcam), MLL1 antibody (Bethyl), anti-histone H3 antibody (Bethyl).

ChIP assays were performed using an EZ Chip chromatin immunoprecipitation kit (Upstate). In brief, cells were fixed in 4% (v/v) formaldehyde, lysed and sonicated to shear the chromatin. The fragmented chromatins were pre-cleaned with protein Gagarose and subjected to overnight immunoprecipitation with specific antibodies for ERa (oestrogen receptor a) and MLL1. Immunoprecipitated chromatins were washed and deproteinized, and DNA fragments were purified by phenol/chloroform extraction followed by precipitation overnight at −80°C. The purified DNA fragments were then used as templates in PCR amplification of four EREs of the HOXA10 promoter, using the primer pairs listed in Table 1.

In order to confirm physical interaction of MLL and ERa, coimmunoprecipitation from HL-60 cells in the absence and presence of E2 was performed. In brief, cells were treated with 100 nM E2 for 6 h, and harvested for preparation of nuclear extract. E2-treated and untreated nuclear extracts were incubated overnight at 4°C with MLL1 antibodies bound to the protein agarose-G beads. The beads were separated, and washed with buffer C (20 mm Tris/HCl, pH 7.9, 5 mm MgCl₂, 420 mm KCl, 0.5 mm dithiothreitol, 0.2 mm phenylmethanesulfonyl) in the presence of 0.1% NP40. The affinity-bound proteins were eluted from the beads using 0.2 m glycine (pH 2.9), and analysed by Western blot, using specific antibodies, for the presence of ERa.

Apoptosis was assessed using annexin-V conjugated with FITC. 1−5×10⁶ HL-60, THP-1 cell lines were treated with and without LSD1 (lysine-specific demethylase 1). And Cells were plated at 2.5×10⁵ per 60 mm dish in triplicate. After treatment, cells were washed twice with PBS, and stained with PI and FITC–annexin-V (Apoptosis & Necrosis Quantification Kit, Biotium Hayward) for 15 min in the dark. Cells were immediately analysed on GUAVA flow cytometer for PI and FITC–annexin-V staining.

Results are expressed as the means±standard deviation. The t test and approximate calculation by normal distribution were used for studying the difference between groups. A P value less than 0.05 (double-sided) was considered as significant, and a P value less than 0.01 was considered as very significant.

In order to determine whether HOXA10 is regulated by E2, a steroidogenic human cell line (HL-60 and THP-1 cells cultured in phenol-red free medium containing activated charcoal-treated FBS) was treated with different concentrations (50 and 1000 nM) of E2 for 8 h. The RNA was isolated from the E2-treated cells and analysed by RT–PCR, using HOXA10-specific primers (Table 1). Interestingly, the results demonstrated that HOXA10 was overexpressed upon exposure to E2 in a concentration-dependent manner (Figure 1). In comparison with the control, HOXA10 expression was significantly enhanced in all cell lines. It was found that 100 nM was the optimal concentration for the E2-mediated induction of HOXA10. Time-dependence experiments demonstrated that HOXA10 activation was maximized after 6–8 h of E2 treatment.

In order to study the potential role of ERs in E2-induced activation of HOXA10, ERa and ERβ were knocked down separately, using the specific antisense oligonucleotides, in HL-60 and THP-1 cells and exposed to 100 nM E2 for an additional 8 h. A scramble antisense oligonucleotide (with no homology to ERs) was used as a negative control. The results demonstrated that application of ERa antisense oligonucleotide knocked down the respective genes efficiently, at both the mRNA and the protein level. After confirming effective knockdown, the RNA from these ER knockdown and E2-treated cells for the expression levels of HOXA10 were analysed using RT–PCR. Interestingly, upon knockdown of either ERa, HOXA10 expression was suppressed (Figure 2). These results demonstrated that both ERa are essential for up-regulation of HOXA10.

As MLLs are well known as master regulators of HOX genes, and several MLLs are implicated in E2 signalling, whether any of the MLLs are involved in stimulation of HOXA10 expression was studied. MLL genes (MLL1, MLL2, MLL3 and MLL4) were knocked down separately using specific phosphorothioate antisense oligonucleotides. The efficiencies of different MLL (MLL1–MLL4)-specific antisense oligonucleotides and their most effective doses were studied. The specific MLL knockdowns were confirmed by analysing their respective gene expression at both the mRNA and protein levels. In parallel, a scramble antisense oligonucleotide without homology with any of the MLLs as a negative control was applied. As seen in Figure 3, upon application of MLLs antisense oligonucleotide, only the knocked down of MLL1 could induced that HOXA10 expression decreased significantly (Figure 3). Whereas, MLL2, MLL3 and MLL4 knockdown have no effect on HOXA10 expression. These results demonstrated that MLL1 played critical roles in the regulation of HOXA10.

As the HOXA10 promoter contains several ERE1/2 regions within the first 3000 nucleotides upstream of the transcription start site, the involvement of some of these EREs (ERE1–ERE2) was studied by analysing the in vitro binding of ERs and MLLs. The in vitro binding of the different factors in the absence and presence of E2 were studied by ChIP assays, using antibodies against MLL1, ERα. ChIP experiments were also performed in parallel with the use of antibody against β-actin as a non-specific negative control. In brief, cells were treated with 100 nM E2 for 8 h. Afterwards the control and E2-treated cells were subjected to ChIP analysis. The immunoprecipitated DNA fragments were PCR amplified using primers specifically for ERE1 and ERE2 of the HOXA10 promoter. As seen in Figure 4, no significant binding of actin was observed in any of the EREs, irrespective of the absence and presence of E2. Binding of MLL1, ERα was increased in both ERE1 and ERE2 of the HOXA10 promoter with E2, compared with E2 treatment.

ERs are well known to bind directly to the EREs of the E2-responsive genes via their DNA-binding domains. MLLs (MLL1–MLL4) also have DNA-binding domains that might be involved in direct binding of promoters. This binding may be critical for regulation of basal transcription of the target genes. On the other hand, MLLs might be recruited to the HOXA10 promoter via protein–protein interactions directly or indirectly with ERs. Amino-acid sequence analysis demonstrated that MLL1–MLL4 had LXXLL domains [also called NR (nuclear receptor) boxes], which are known to be involved in E2-dependent interactions with ERs. The physical interactions of MLLs with ERα were further confirmed by using coimmunoprecipitation experiments. The interaction of MLL1 with ERa was studied separately. In brief, HL-60 cells were treated with 100 nM E2 for 8 h. Nuclear extracts were prepared from these E2-treated and untreated cells, and were incubated with MLL1 antibody (bound to protein-G agarose beads) overnight at 4°C. Proteins bound to the MLL1-attached and control beads were analysed by Western blotting using antibodies specific for ERa and MLL1. The results demonstrated that the interactions of both ERa with MLL1 were increased in the presence of E2 (Figure 5).

As shown that MLL1 could bind to HOXA10 EREs to regulate HOXA10 expression through protein complex formation with ERα, the H3K4 methylation status with and without E2 treatment in HL-60 and THP-1 cells were studied since MLL contains specifically methylate histone H3K4. The H3K4 methylation significantly increased with E2 in both leukaemia cell lines (Figure 6A). On the contrary, the H3K4 methylation decreased with ERα knockdown (Figure 6B).

MSP, RT–PCR and Western blot were performed to respectively analyse the methylation status of HOXA10 promoter CpG island, mRNA expression of MLL1 and HOXA10, protein expression and H3K4 methylation status in the leukaemia cell line (HL-60) with and without MLL1 knockdown (Figure 7).

With MLL1 knockdown, promoters of HOXA10 was fully methylated in HL-60, compared with semi-methylated without MLL1 knockdown (Figure 7C). However, the methylation status of H3K4 decreased with MLL1 knockdown (Figure 7D).

Pi (propidium iodide) simple staining was performed to analyse the cell apoptosis on the HL-60 and THP-1 lines with or without LSD1 treatment. With the H3K4 demethylation treatment of LSD1, the cell apoptosis had significant increase in HL-60 and THP-1 cells (AML), compared with leukaemia cell line without LSD1 demethylation treatment (Figure 8).

HOX genes play a major role in embryonic development. HOX gene products act as transcription factors that regulate critical genes that are necessary for cell differentiation and development. Hox proteins are transcription factors, which are proteins that are capable of binding to specific nucleotide sequences on the DNA called enhancers where they either activate or repress genes. The same Hox protein can act as a repressor at one gene and an activator at another gene [5,11].

Studies have been shown that HOXA10 is a homeodomain transcription factor that is frequently overexpressed in human AML. In murine bone marrow transplantation studies, HOXA10 overexpression induces a myeloproliferative disorder with accumulation of mature phagocytes in the peripheral blood and tissues. Over time, differentiation block develops in these animals, resulting in AML [12]. In general, ERs, along with various coregulators, are recruited to EREs present in the promoters of E2-responsive genes [13]. Studies have demonstrated that the HOXA10 promoter contains two EREs within 3000 bp upstream of the transcription start site that ERs could bind, which means ERs may play an important role in HOXA10 expression [14]. Indeed, the ER gene has been shown to be a growth and/or metastasis suppressor in multiple cell types, and the negative effect of oestrogen on haematopoiesis suggests that this is also the case in leukaemias [15]. And studies have suggested a potential role of ER in leukomogenesis because ER CpG island methylation at very high frequency caused inactivation of ER expression in human leukaemias. Also it has been shown that the inactivation of ER in mice leads to a leukaemia-like syndrome [16–18].

Oestrogens are important signalling molecules that regulate various physiological processes such as cell growth, reproduction, development and differentiation. They also have a role in several pathological processes of hormone-dependent diseases, including cancers. Cellular signalling by oestrogens is mediated via the ERa and ERβ (oestrogen receptor β), and the two ER's exert different biological responses. ERa serve as an oncogene, however, ERβ serve as tumour suppress gene. The metabolism of endogenous oestrogens can be associated with the development of hormone-dependent human cancers. Moreover, it has recently been demonstrated that some haematological malignancies, including leukaemia and lymphoma cells, are positive for oestrogen receptors (ERs). Furthermore, some lymphomas, including various subtypes of NHL (non-Hodgkin's lymphoma), develop with different frequencies between men and women, and suggested that E2 significantly increased the proliferation of Hut-78 T cell and Jurkat cell [19].

In the present study, we treated leukaemia cell lines with 100 nM E2, that concentration has been used by many studies [20,21] and proved by experiment. And there was a study found that the leukaemia cell line RMPI-8226 contained the highest levels of oestrogen (182.06 pg/106 cells) and 17b-estradiol (753.45 pg/106 cells), compared with breast and ovarian cancer. The present study shows the HOXA10 expression could be activated with E2 treatment for 8 h and antisense oligonucleotide-mediated knockdown of either ERa down-regulates the E2-mediated activation of HOXA10, indicating their critical roles in the process. ER-mediated regulation of E2-sensitive genes is a complicated process. In the presence of E2, ERs are activated and bind to the EREs of E2-responsive genes, eventually resulting in transcription activation. In addition to ERs, E2-mediated gene activation requires various other coregulators and coactivators that result in chromatin modification and remodelling. It demonstrates that MLL1 plays a crucial role in the regulation of HOXA10. Knockdown of MLL1 suppresses the activation of HOXA10.

There is sequence analysis demonstrated that the HOXA10 promoter contains two EREs that close to transcription start site. In vitro binding analysis (ChIP), it is demonstrated that, in the presence of E2, more MLL1 could bind to both ERE1 and ERE2, compared with absence of E2. Furthermore, the coimmunoprecipitation experiments demonstrate that MLL1 interacts with ERa in an E2-dependent manner. These results suggest that MLL1 could bind to HOXA10 promoter through conforming a protein complex with ERa.

To further understand the MLL1 mediated regulation of HOXA10 gene in AML, we study the methylation status of H3K4 with and without E2, and the methylation status of H3K4 and the methylation status of HOXA10 promoter with and without MLL1 knockdown. The level of H3K4 methylation with E2 is higher than that without E2. And the HOXA10 expression is also increased with E2 treatment. With MLL1 knock down, the level of H3K4 methylation significantly decreases and the methylation status of HOXA10 is fully methylated, which induces the HOXA10 expression to elevate significantly both in mRNA and protein.

Moreover, HL-60 and THP-1 cell with LSD1 demethylation treatment, the apoptosis is significantly increased. These above results suggest MLL1 is the key factor in regulation of HOXA10 and MLL1 regulates the HOXA10 expression in epigenetic way. MLL1 binds to promoter of HOXA10 through ERs to regulate HOXA10 expression, because MLLs possess H3K4-specific HMT activity to methylate H3K4, which could elevate HOXA10 expression.

ALL

acute lymphoblastic leukaemia

AML

acute myeloid leukaemia

ChIP

chromatin immunoprecipitation

ER

oestrogen receptor

ERE

oestrogen response element

H3K4

H3 at lysine 4

HMT

histone methyltransferase

LSD1

lysine-specific demethylase 1

MSP

methylation-specific PCR

MLL

mixed lineage leukaemia histone methylase

NP40

Nonidet P40

Jie Yao and Junsong Zheng designed research; Jie Yao and Lichao Fang performed research; Zai-Lin Yang, Hui Huang, Yan Li and Jun Deng contributed new reagents/analytic tools; and Jie Yao wrote the paper.

This work was supported by the National Nature Scientific Foundation of China [grant number 81000225], and the Natural Science Foundation Project of CQ CSTC [grant number CSTC2013jcyja0102].