S. S. Mandal, Department of Chemistry and Biochemistry, The University of Texas at Arlington, Arlington, TX 76019, USA Fax: +1 817 272 3808 Tel: +1 817 272 3804 E-mail: firstname.lastname@example.org
Mixed lineage leukemias (MLLs) are histone-methylating enzymes with critical roles in gene expression, epigenetics and cancer. Although MLLs are important gene regulators little is known about their own regulation. Herein, to understand the effects of toxic stress on MLL gene regulation, we treated human cells with a common food contaminant mycotoxin, deoxynivalenol (DON). Our results demonstrate that MLLs and Hox genes are overexpressed upon exposure to DON. Studies using specific inhibitors demonstrated that Src kinase families are involved in upstream events in DON-mediated upregulation of MLL1. Sequence analysis demonstrated that the MLL1 promoter contains multiple Sp1-binding sites and importantly, the binding of Sp1 is enriched in the MLL1 promoter upon exposure to DON. Moreover, antisense-mediated knockdown of Sp1 diminished DON-induced MLL1 upregulation. These results demonstrated that MLL1 gene expression is sensitive to toxic stress and Sp1 plays crucial roles in the stress-induced upregulation of MLL1.
Elucidating the regulatory network of proto-oncogenes in normal healthy cells and under toxic stress is important for understanding the mechanism of human diseases [1–5]. Mixed lineage leukemias (MLLs) are a set of evolutionary conserved genes that are often rearranged and misregulated in acute lymphoblastic and myeloid leukemias [1,2,6]. Humans encode several MLL protein families, such as MLL1, MLL2, MLL3, MLL4 and Set1 [1,7–14]. In general, they are master regulators of homeobox (Hox) genes which are critical for cell differentiation and development [1,2,15,16]. Because of their importance in gene regulation and disease, researchers have purified MLL proteins from human cells and have demonstrated that MLLs posses histone H3 lysine-4-specific methyl-transferase activities and play a critical role in gene activation [9,17–20]. MLLs exist as multiprotein complexes inside cells with several common protein subunits such as Ash2, Wdr5, Rbbp5 and CGBP [1,9,10,19,21]. Recently, we affinity purified several MLL complexes from human cells and demonstrated that MLL1 plays critical roles in histone H3 lysine-4 methylation and Hox gene regulation . We also demonstrated that downregulation of MLL1 results in cell-cycle arrest in the G2/M phase indicating its critical role in cell-cycle progression .
Although recent discoveries of MLL-associated histone H3 lysine-4-specific methyl-transferase activities have shed significant light on the complex function of MLLs in gene regulation, little is known about their own regulation in normal cells or in cells under stress . However, it has been reported that certain chemotherapeutic stresses result in MLL rearrangement and misregulation, leading to the development of secondary leukemias in humans [23,24]. These observations indicated that MLL1 is stress-responsive gene. Herein, we studied the effect of a potential carcinogenic mycotoxin, deoxynivalenol (DON) on the regulation of MLL1. Notably, DON is a toxin produced by pathogenic fungi during the infection of cereal crops and is often linked with various acute and chronic human diseases, including cancer [25–27]. Herein, we report that MLL1 and its target Hox genes are upregulated upon exposure to DON and transcription factor Sp1 plays critical roles in the DON-mediated upregulation of MLL1.
DON induces expression of MLL
To understand the effects of mycotoxic stress on MLL expression, we treated cultured human cells (H358 cells) with varying concentrations of DON (up to 33 μm) for 7.5 h. We isolated RNA from the treated and untreated control cells and subjected it to RT-PCR with primers specific to MLL1 and Set1. As seen in Fig. 1A,B, treatment with DON induced two- to five-fold overexpression of MLL1 and Set1 in a concentration-dependent manner. MLL1 overexpression by DON was more dramatic (8.3-fold) at the protein level (lane 4, Fig. 1C,D). The decrease in expression of MLL1 and Set1 at 10 h or longer (Fig. 1C,D) is likely caused by cell death induced by DON. Because MLL1 is upregulated upon exposure to DON, we analyzed the expression of several other proteins (such as Rbbp5, Wdr5 and Ash2) known to interact with MLL1 [9,21]. We also analyzed the effect of DON on expression of some MLL1 target Hox genes (HoxA2, HoxA7, HoxB1, HoxB7, etc.). Importantly, similar to MLL1, Rbbp5 and Wdr5 were overexpressed upon treatment with DON, whereas Ash2 was not affected significantly (Fig. 2). Similarly, HoxA7, HoxA2 and HoxB1 were overexpressed, whereas HoxB7 was downregulated upon exposure to DON (Fig. 2 and data not shown). The upregulation of MLL1, its several interacting proteins or selected target Hox genes upon exposure to DON indicated that expression of these proteins is sensitive to toxic stress.
Notably, we analyzed the effects of DON on cell growth and determined the cytotoxicity (IC50) towards H358 cells using a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromideassay, as described previously . Upon treatment with 3.3 μm DON, up to 5, 47 and 68% of H358 cells were killed at 7.5, 24 and 72 h post treatment, respectively. The IC50 value is determined to be 1 μm. These results demonstrated that DON is significantly cytotoxic towards human cells.
Src kinase inhibitor suppressed the DON-induced upregulation of MLL1
To understand potential mechanism of DON-mediated upregulation of MLL1 and Hox genes, we examined the involvement of different DON-responsive signaling pathways. Because DON is known to induce ribotoxic stress that instigates various signaling cascades, including MAP/Src kinases [29–33], we initially examined whether inhibition of MAP/Src kinase activation had any effect on DON-induced upregulation of MLL1. We treated cells with a Src kianse inhibitor (PP2) or a MAP kinase inhibitor (PD98059) and then exposed the cells to DON. As expected, MLL1 was upregulated upon treatment with DON (lanes 1 and 4–7, Fig. 3). However, upon treatment with PP2, DON-induced expression of MLL1 and HoxA7 was suppressed in a concentration-dependent manner at both the mRNA and protein levels (compare lanes 4–7 with lane 1, Fig. 3A,B). These results indicated that Src kinases play a critical role in regulating upstream events that lead to MLL1 and HoxA7 upregulation by DON. Notably, PP2 has no significant effect on DON-induced expression of Set1, Rbbp5, Ash2 and Wdr5 (data not shown) suggesting the involvement of alternate pathways. Because MLL1 induction was suppressed by Src kinase inhibitor (PP2), we examined whether MAP kinases are also involved in DON-mediated MLL1 upregulation. However, application of PD98059 did not have any significant effect on DON-induced upregulation of MLL1, indicating no involvement of MAP kianses in this process (Fig. 3C).
Sp1 plays a critical role in DON-induced MLL1 upregulation
To understand the mechanism of MLL1 upregulation by DON, we analyzed the MLL1 promoter for the presence of various cis-elements recognized by specific transcription factors (such as Sp1, AP2), particularly those known to be activated by mycotoxins [27,29–36]. Interestingly, we found the presence of multiple Sp1-binding sites in the MLL1 promoter (−3000 to +500 nucleotide region; Fig. 4). To investigate possible role of Sp1 in MLL1 gene regulation, we knocked down Sp1 in H358 cells by using Sp1-specific antisense and then analyzed the expression of MLL1 in the absence and presence of DON (3.3 μm). As seen in Fig. 5A,B, treatment with Sp1 antisense effectively knocked down Sp1 expression at both the mRNA and the protein level (compare lanes 1 with 3). Upon knockdown of Sp1, the basal level of MLL1 expression was not significantly affected at the mRNA or the protein level (Fig. 5A,B, lanes 1 and 3). Interestingly, however, DON-induced upregulation of MLL1 (mRNA and protein level) was suppressed to almost normal levels under an Sp1 knocked down environment (Fig. 5A,B, lanes 2 and 4). These results indicated that Sp1 is critical for MLL1 regulation, especially in presence of DON.
Because the MLL1 promoter contains multiple Sp1-binding sites and our results demonstrated that Sp1 is critical in regulating MLL1 on exposure to DON, we hypothesized that DON modulates binding of Sp1 to the MLL1 promoter. To confirm our hypothesis, we treated H358 cells with 3.3 μm DON and subjected them to chromatin immunoprecipitation (ChIP) using anti-Sp1 Ig (Fig. 5C,D). In parallel, we also performed ChIP with an unrelated antibody (actin antiserum). The immunoprecipitated DNA fragments were PCR amplified using primers specific for MLL1 promoter regions R1 (−497 to −593), R2 (−5 to −105), MLL1-ORF (as control) and β-actin-ORF (a second unrelated control). Our results demonstrated that no Sp1 was bound to the ORF region of actin in either the absence or presence of DON (Fig. 5C, upper, lanes 5 and 6). Similarly, Sp1 binding was not enriched in the MLL1-ORF region in the presence of DON (MLL1-ORF; Fig. 5C, lanes 5 and 6). Interestingly, however, the binding of Sp1 was significantly enriched in the Sp1-binding sites of the MLL1 promoter regions R1 and R2 in the presence of DON, although more enrichment was observed in the R2 region (closer to the transcription start site) (Fig. 5C,D, lanes 5 and 6). ChIP analysis showed no binding of β-actin to the MLL1 promoters and ORF region, irrespective of the presence or absence of DON (Fig. 5C, lanes 3 and 4). These results demonstrated that binding and enrichment of Sp1 to the MLL1 promoter regions (R1 and R2) in the presence of DON is specific and this demonstrates that Sp1 is crucial for transcriptional activation of MLL1 under DON treatment.
Furthermore, because phosphorylation of Sp1 is well known to be associated with toxic stress, we analyzed the state of Sp1 phosphorylation upon DON exposure [33,37]. We performed immunoprecipitation of Sp1 using anti-Sp1 Ig (nonphosphorylated) from DON-treated and untreated cells. We analyzed the immunoprecipitates by western blot using both anti-Sp1 Ig (nonphosphorylated) and anti-phosphotyrosine Ig that recognize tyrosine-phosphorylated proteins. Interestingly, upon DON treatment, the protein level of Sp1 was not significantly affected, although, the level of tyrosine-phosphorylated Sp1 was increased (Fig. 5E). These results indicated that DON induces phosphorylation of Sp1 and this might be linked with MLL1 upregulation.
Because MLLs are proto-oncogenes and are known to be rearranged or misregulated under chemotherapeutic stress, leading to secondary leukemias [23,24], elucidating the stress responsive regulatory mechanism of MLL is important. Herein, our studies showed that exposure to mycotoxin DON-induced expression of MLL1, several MLL interacting proteins and MLL target Hox genes. Notably, MLL1 exists as a multiprotein complex inside the cell with subunits like Ash2, Wdr5 and Rbbp5, and MLL1 executes its histone methyl-transferase activity and regulates target Hox genes in the context of the multiprotein complex [1,9,10]. Therefore, because MLL1 and several Hox genes (HoxA7, HoxA2, etc.) were overexpressed upon exposure to DON, we anticipated that MLL-interacting proteins might be upregulated in a similar fashion. However, our results demonstrated that although several MLL1-interacting proteins such as Wdr5 and Rbb5 were upregulated upon exposure to DON, Ash2 expression was not significantly affected. These observations suggest that Ash2 is a unique component of MLLs and may have other distinct functions that are yet to be revealed. It is also possible that Ash2 is normally distributed in different protein complexes which may be redistributed (without being induced) under stress to compensate for the higher expression of MLL1. This aspect needs further investigation for complete understanding. Similarly, although the MLL1 target genes HoxA7, HoxA2 and HoxB1, along with MLL1, were upregulated upon exposure to DON, we observed that HoxB7 expression decreased. These results suggested that the mechanism of regulation, especially in presence of DON, is different for HoxB7 (as well as HoxA2 and HoxB1) and HoxA7, although they are all targets of MLL1 in normal circumstances (without DON). Nevertheless, our results showing the DON-induced upregulation of MLL1 and related proteins indicated that MLL1 and its associated genes are sensitive to toxic stress.
The effect of DON is very well studied in plants [38,39]. In mammalian cells, DON induces oxidative stress, activates MAP/Src kinases and induces inflammation and oxidative stress-responsive genes such as interleukins and cyclooxygenase [32,36,40–42]. Using RT-PCR analysis, we also observed that interleukin-8 and cyclooxygenase are overexpressed in H358 cells upon exposure to DON, indicating the induction of oxidative stress in human cells, as reported earlier (data not shown) [36,41]. Furthermore, using Src kinase inhibitor (PP2), we demonstrated that DON-induced MLL1 and HoxA7 gene upregulation were alleviated in the presence of PP2. These observations demonstrated that Src kinases are involved in upstream events in DON-mediated upregulation of MLL1 and HoxA7. Notably, our results demonstrated that application of PP2 has no significant effect on the DON-induced upregulation of other proteins such as Set1, Wdr5 and Rbbp5 (data not shown), suggesting the involvement of alternate pathways in the regulation of these genes.
Our sequence analysis demonstrated that the MLL1 promoter contains multiple binding sites for Sp1, a transcription factor that is well known to be activated and phosphorylated under stress [29,33,34,36,43]. The literature relating to mycotoxin-mediated activation of Sp1 and our results showing the presence of multiple Sp1-binding sites in the MLL1 promoter, prompted us to hypothesize that Sp1 plays a critical role in the regulation of MLL1, especially under mycotoxic stress [33,37,43]. Our studies demonstrated that antisense-mediated knockdown of Sp1 suppressed the effects of DON on upregulation of MLL1. In addition, the level of Sp1 is enriched in the Sp1-binding regions of the MLL1 promoter upon exposure to DON. These results demonstrated that Sp1 acts a mediator in translating the effects of DON on MLL1 gene upregulation. Notably, cells respond to stress by activating signaling pathways that regulate defense responsive genes [36,38,39]. An early step in the stress response includes phosphorylation of the MAP/Src kinases leading to their activation . Sp1 and other Sp1 family members are differentially acetylated, phosphorylated and/or glycosylated, and bind variants of a GC-rich box in promoter of target genes. Because the MLL1 promoter contains multiple Sp1-binding sites and is regulated by Sp1, as well as the Src family of kinases on DON treatment, we hypothesized that Sp1 is likely phosphorylated and recruited to the MLL1 promoter, resulting in its upregulation. Our studies demonstrated that Sp1 is phosphorylated upon exposure to DON. Although, at this point we could not directly analyze recruitment of the phosphorylated Sp1 into the MLL1 promoter because of the unavailability of the phospho-Sp1-specific antibody, the increased recruitment of the Sp1 in the MLL1 promoter may be linked with phosphorylation of Sp1.
In conclusion, we demonstrated that MLL1, several MLL-associated proteins and Hox genes are upregulated upon exposure to mycotoxin DON via involvement of Src kinase activation. The transcription factor Sp1 plays critical role in upregulating MLL1 gene expression under mycotoxic stress. Although further analysis is needed to understand the detailed mechanism of MLL gene (and other DON-responsive genes) regulation in normal cell or under stress, our studies established a novel link between MLL gene regulation, the stress response and DON, and revealed critical stress-responsive MLL1 gene regulatory pathways. Although, the mechanism is not clear, MLL is well known to be rearranged and misregulated in various cancers and it is likely that different types of stresses cause MLL misregulation and rearrangement. As exposure to DON induces upregulation of MLL1, we hypothesize that this may be one of the possible mechanism by which DON exerts is carcinogenic action in human cells.
Cell culture and treatments with DON
Human cells (H358, a lung cancer, ATCC) were grown on RPMI media supplemented with 10% fetal bovine serum, l-glutamine (1%) and penicillin/streptomycin (0.1%) (Sigma, St Louis, MO, USA). For the toxin treatment, cells were grown to 80% confluence and treated with varying concentrations of DON (Sigma) for different times, as needed. Total RNA and proteins were isolated from the treated and untreated cells and subjected to RT-PCR and western blot analysis. For the RT-PCR analysis, each experiment was performed in two to four replicates in parallel. For the western blot analysis, proteins from replicate experiments were pulled together prior to SDS/PAGE.
Preparation of RNA, nuclear proteins and whole-cell extract
DON-treated and untreated cells were harvested by centrifugation at 500 g, resuspended in diethyl pyrocarbonate-treated buffer A (20 mm Tris/HCl, pH 7.9, 1.5 mm MgCl2, 10 mm KCl, 0.5 mm dithiothreitol and 0.2 mm phenylmethanesulfonyl fluoride), incubated on ice for 10 min and then centrifuged at 3500 g for 5 min. The supernatant containing the cytoplasmic extracts was subjected to phenol–chloroform extraction followed by LiCl precipitation of cytoplasmic mRNA by incubating overnight at −80 °C. The mRNA was washed with diethyl pyrocarbonate treated 70% EtOH, air dried and resuspended in diethyl pyrocarbonate-treated water. Nuclear proteins extracts were prepared from the nuclear pellets, as descried previously [21,22]. For preparation of whole-cell protein extracts cells were incubated in whole cells extract buffer (50 mm Tris/HCl pH 8.0, 150 mm NaCl, 5 mm EDTA, NP-40, 0.2 mm phenylmethanesulfonyl fluoride, 1 × protease inhibitors) for 20 min on ice. The whole cell extract was separated from histone protein by centrifugation at 12 000 g for 10 min.
RT-PCR and western blots
Reverse transcription reactions were performed in a total volume of 25 μL containing 1 μg of total RNA, 2.4 μm of oligo-dT, 100 U of MMLV reverse transcriptase (Promega, Madison, WI, USA), 1 × first strand buffer (Promega), 100 μm dNTPs, 1 mm dithiothreitol and 20 U of RNaseOut (Invitrogen, Carlsbad, CA, USA). This cDNA (1 μL) was used for PCR with primer pairs listed in Table 1. Each of the experiments was performed in two replicates for three times. The normality of the data was analyzed by using t-test and analyses of the variants (ANOVA) were performed at 5% level of significance.
Table 1. Primers used for RT-PCR, chromatin immunoprecipitation and antisense experiments.
Forward primer (5′- to 3′)
Reverse primer (5′- to 3′)
a Phosphorothioate antisense oligonucleotide.
Equivalent amount of proteins were analyzed in SDS/PAGE and subjected to western blot analysis with specific antibodies. MLL1, MLL2, Set1, Ash2 and Rbbp5, antibodies were purchased from Bethyl laboratory (Montgomery, TX, USA).
Immunoprecipitation and western blotting of Sp1 and phosphorylated Sp-1
For western blot analysis of the Sp1 expression, equivalent amounts of whole-cell extract (DON-treated and untreated) were separated in 8% SDS/PAGE and subjected to western blot analysis using anti-Sp1 Ig (Upstate, Waltham, MA, USA). For the analysis of DON-induced phosphorylation of Sp1, we performed immunoprecipitation of Sp1 from the whole-cell protein extract using anti-Sp1 Ig, as described earlier . The Sp1 immunoprecipitates were electrophoresed in 8% SDS/PAGE and subjected to western blot using both anti-Sp1 (nonphosphorylated) and anti-phosphotyrosine Ig (Upstate) that recognize tyrosine phosphorylated proteins.
Antisense-mediated knockdown of Sp1
The Sp1 antisense (5′-CTGAATATTAGGCATCACTCCAGG-3′) was transfected into H358 cells using Maxfect transfection reagent (MoleculA). In brief, H358 cells were grown to 60% confluence, washed twice with fetal bovine serum-free RPMI media and then incubated with transfection reagent–antisense complex for 5 h in serum-free RPMI prior to the addition of complete growth medium (with 10% serum). Cells were then incubated for 48 h followed by treatment with 3.3 μm DON for 7.5 h. Cells were then harvested for RNA, nuclear protein extraction or ChIP analysis. A scramble antisense without any sequence homology with Sp1 (5′-CGTTTGTCCCTCCAGCATCT-3′) was used as control.
The ChIP assay was performed using H358 cells and anti-Sp1 mAb (Bethyl lab) using EZ Chip™ chromatin immunoprecipitation kit (Upstate) as described previously [21,22]. Immunoprecipitated DNA obtained from the ChIP was PCR amplified with different primers (specific to Sp1 rich sites in MLL1 promoter, Table 1).
This work was supported by grants from Texas Advanced Research Program (00365-0009-2006) and American Heart Association (SM 0765160Y). We also thank Saoni Mandal and other Mandal lab members for critical discussions.