|Home | About | Journals | Submit | Contact Us | Français|
Inactivating mutations in the tumor suppressor gene MEN1 cause the inherited cancer syndrome multiple endocrine neoplasia type 1 (MEN1). The ubiquitously expressed MEN1 encoded protein, menin, interacts with MLL (mixed-lineage leukemia protein), and together they are essential components of a multiprotein complex with histone methyl transferase activity. MLL is also essential for hematopoeisis, and plays a critical role in leukemogenesis via epigenetic regulation of Hoxa9 expression that also requires menin. Therefore we chose to explore the role of menin in hematopoiesis. We generated Men1−/− embryonic stem (ES) cell lines, and induced them to differentiate in vitro. While these cells were able to form embryoid bodies (EBs) expressing the early markers Flk-1 and c-Kit, their ability to further differentiate into hematopoietic colonies was compromised. The Men1−/− ES cells show reduced expression of Hoxa9 that can be recovered by reexpression of Menin. We demonstrate that the block in differentiation of Men1−/− ES cell lines can be rescued not only by the expression of menin but also that of Hoxa9. These results suggest that, similar to MLL, menin is required for hematopoiesis, and this requirement may be mediated through regulation of Hoxa9 expression.
Multiple endocrine neoplasia type 1 (MEN1), characterized by the tumors of the parathyroid, enteropancreatic endocrine tissues and anterior pituitary (Marx, 2005), is caused by mutations in the MEN1 gene (Chandrasekharappa et al., 1997). The MEN1 encoded nuclear protein, menin, is ubiquitously expressed, and widely conserved from human to fly, however, the amino acid sequence provides no clues to its function (Chandrasekharappa and Teh, 2003). Men1+/− mice develop endocrine tumors modeling the human disease. Homozygous loss (Men1−/−) results in embryonic lethality (E11.5-E13.5), indicating a requirement for menin in early development (Crabtree et al., 2001).
Menin binds MLL (mixed-lineage leukemia) and as part of a multiprotein complex, methylates lysine 4 of histone H3 (H3K4), a modification associated with activation of gene expression (Hughes et al., 2004). Though the precise biochemical function of menin is unclear, its importance in the complex is apparent from the observation that the loss of menin results in reduced H3K4 methylation, and associated reduction in the expression of known MLL targets Hoxc8 (Hughes et al., 2004) and Hoxa9 (Yokoyama et al., 2004).
MLL is required for normal hematopoiesis. Experiments with chimeric mice demonstrated that Mll −/− ES cells are incapable of contributing to both embryonic and adult hematopoiesis (Ernst et al., 2004a), and Mll −/− ES cells in culture show a greatly reduced ability to form hematopoietic colonies that can be rescued by the reexpression of Hoxa9 (Ernst et al., 2004b). Genomic translocations involving MLL result in hematologic disorders via aberrant Hoxa9 expression and menin is essential for this pathogenesis (Ayton and Cleary, 2003; Caslini et al., 2007; Yokoyama et al., 2005).
To examine the requirement of menin in hematopoiesis we developed Men1−/− ES cell lines, and induced them in vitro to differentiate into hematopoietic lineages. The absence of menin led to a decreased number of hematopoietic colonies, and expression of Hoxa9 rescued the deficiency.
Starting from a Men1+/− cell line (Crabtree et al., 2001) that had a wildtype (Wt) and a deleted (exons 3–8) allele, Men1−/− ES cell lines were generated by targeting and then deleting exons 3 through 8 from the remaining wildtype allele. Figure 1A depicts the schematic of the 2-step targeting/deleting the wildtype allele in the Men1+/− Het cell line. Southern blots of BamH1 digests (Figure 1B) show the deleted allele (2.3kb) and the Wt (17kb) allele in the Men1+/− Het cell line, which upon targeting is replaced by the 7.2kb targeted allele. Subsequent removal of this targeted allele was mediated by transfection with a Cre expression plasmid, and this is indicated by the absence of the 7.2kb fragment in the three null clones (Men1−/−): 3.2N, 173N, and 179N (Figure 1B). Lack of Men1 expression at both the RNA (Figure 1C) and protein level (Figure 1D) was apparent in these three null cell lines; TC1, the parental Men1+/+ cell line, served as a control. Three Men1−/− cell lines, 3.2N, 173N, and 179N that lack expression of the transcript and the encoded protein were established.
We evaluated the ability of these three Men1−/− ES cell lines to differentiate into embryoid bodies (EBs). The null cell lines formed a similar number of EBs as that from TC1, indicating that they are efficient in the early differentiation process. The proportion of hematopoietic EBs, identified by the presence of blood islands and differentiated cells on the periphery, were also similar to that in TC1 (Figure 2A). Thus the absence of menin does not appear to affect EB formation and early hematopoietic precursors. Consistent with this observation, the percentages of the Flk-1 and c-Kit positive cells in Men1−/− EBs (day 6) were comparable to those from TC1 (Figure 2B). Expression of Flk-1 indicates the presence of cells with both primitive and definitive hematopoietic capabilities (Kabrun et al., 1997). Therefore, the development of these early mesodermal derivatives with potential to develop into hematopoietic colonies is not affected by the loss of Men1. Furthermore, the cells that express both c-Kit and CD41 represent commitment to hematopoiesis. We checked whether loss of menin has any effect on this fraction of cells in the Men1−/− EBs (day 6). The c-Kit+/CD41+ fraction from the three menin-null cell lines was similar to that from the TC1 cell line, and representative data from a menin null cell line 3.2N and the wild type TC1 is shown (Figure 2C). The average fraction of CD41+/c-Kit+ cells for three independent experiments for 3.2N and TC1 were 0.85% and 0.87% respectively. Thus, loss of menin does not appear to compromise the cells that are committed for hematopoietic differentiation.
EBs were dispersed and replated for a secondary round of differentiation with the appropriate growth factors to examine their ability to form hematopoietic colonies. The Men1−/− EBs were deficient in producing hematopoietic colonies (Figure 3A). The total number of hematopoietic colonies formed from EBs of the three null cell lines were reduced nearly 5-fold compared to the wildtype TC1 cell line. In addition, we found a large number of secondary EBs developed from these Men1−/− EBs (Figure 3B). The null EBs generated 6–8 times larger number of secondary EBs compared to that from TC1. The data in 3A and 3B are from the same three independent differentiation experiments. The color of the medium from differentiation experiments of both the TC1 cells and the null cells appeared similar indicating that the increased number of secondary EBs does not appear to result in depletion of nutrients. The decreased number of hematopoietic colonies from null cells thus reflects the consequence of the loss of menin. These Men1−/− ES cells progress through the early differentiation stages normally, but show defects in differentiation during the later stages of hematopoietic colony formation.
These findings from Men1−/− cell are similar to those observed for Mll−/− ES cells (Ernst et al., 2004b): Early differentiation to form EBs was not compromised in Mll−/− ES cells but their subsequent differentiation to hematopoietic colonies was affected. The average number of the distinct types of hematopoietic colonies generated from TC1 and the three null cell lines from five independent differentiation experiments is summarized (Figure 3C).
We found that expression of known menin targets, Hoxa7, Hoxa9, Hoxa10, Hoxc6 and Hoxc8 was reduced in the three Men1−/− EBs (6-day) (Figure 4A). Hoxa9 is the most studied Hox gene, as it regulates hematopoiesis and its aberrant expression leads to leukemia(Popovic et al., 2008). Using conditional knockout mice, a recent study showed that menin does regulate hematopoiesis and myeloid transformation by altering Hox gene expression (Chen et al., 2006). The 3.2N ES cell line was transfected with vector or the Men1 expression construct, and the expression of Hoxa9 was measured 3rd and 4th day post transfection. As anticipated, transfection of menin null cells with Men1 expression construct was accompanied by restored expression of Hoxa9 (Figure 4B) indicating that menin regulates the expression of Hoxa9.
We transfected Men1−/− EBs with MEN1 or control (vector) expression constructs and found that MEN1 expression increased the total number of hematopoietic colonies over that of control (Figure 5A). We also evaluated whether HOXA9 expression can rescue the deficiency in Men1−/− EBs. The reexpression of HOXA9 was shown to result in an increased number of hematopoietic colonies, indicating that menin apparently affects hematopoietic differentiation by affecting the expression of Hoxa9 (Figure 5A). RT-PCR of RNA from transfected cells showed that the constructs did express the respective transcripts (Figure 5B). It has been reported that Hoxa9 mediated growth of leukemic cells transformed by MLL-AF4 and MLL-ENL may involve insulin like growth factor 1- receptor (IGF-1R) and c-Myb, respectively (Hess et al., 2006; Whelan et al., 2008). It is possible that these genes might play a role in normal hematopoiesis requiring MLL, and also menin. It is worth exploring these and many other gene products that are required for hematopoiesis, and this in vitro differentiation system reported here would provide an avenue to explore their role in menin-mediated hematopoiesis.
Using conditional knockout mice, by excising the Men1 alleles from bone marrow, and bone-marrow transplantation experiments, the role of menin in hematopoiesis has been explored in two reports (Chen et al., 2006; Maillard et al., 2008). It was discovered that menin does regulate hematopoiesis and myeloid transformation by altering Hox gene expression (Chen et al., 2006). This observation has now extended in a more recent study (Maillard et al., 2008) and the authors conclude that menin has an essential and novel role in hematopoietic stem cell (HSC) homeostasis, and menin plays a distinct role particularly in hematopoietic recovery, or when the marrow is exposed to stress. Our results complement the studies in these two reports, and allows for further exploring the molecular dissection of the menin-mediated events in hematopoietic differentiation, as well as other differentiation processes.
In summary, we found that menin loss did not compromise the formation of EBs, nor their commitment to hematopoietic differentiation, but affected the subsequent steps in the hematopoietic differentiation. This deficiency in hematopoiesis is associated with reduced expression of Hoxa9, and we have shown that this block in differentiation can be corrected by reexpression of Hoxa9. Loss of Mll results in a similar scenario, and thus we demonstrate that menin and MLL together are integral components of the epigenetic regulation of hematopoiesis. The in vitro differentiation system described here provides a novel avenue to follow the molecular events in hematopoiesis and leukemogenesis that are mediated by menin.
The wild type allele in the Men1+/− cell line (Crabtree et al., 2001) was targeted and deleted by sequential transfections (Figure 1A). The same floxed targeting vector was transfected into the Men1+/− cell line. The correctly targeted clone from the first step, identified by Southern blot analysis, was transfected with a plasmid expressing Cre recombinase to generate the null allele. The Men1−/− clones were identified by Southern blot analysis, and characterized for Men1 expression by RT-PCR and western analysis, as described earlier (Crabtree et al., 2001; Guru et al., 1998).
The ES cell lines were differentiated using the protocol from Stem Cell Technologies. Briefly, ES cells (2.5 × 103 cells/ml) were plated in methylcellulose to form embryoid bodies (EBs) for 10 days. On day 7, a solution containing mSCF, IL3, IL6 (Stem Cell Technologies) and erythropoietin (Pharmingen) was added. The EBs were harvested on day 11, single cell suspensions were prepared and 2.5 × 104 cells/ml replated onto methylcellulose containing the same combination of cytokines. Equal number of cells were plated for both the Wt and the null cell lines. After 10–12 days, the colonies were identified and counted. A knock-in mutant ES cell line, KIKL55, is used as a control as it is described to be competent to form EBs but not the differentiated hematopoietic colonies (Castilla et al., 1996). Individual colonies were either identified by morphological characteristics of the different kinds of hematopoietic cell types or harvested from the methylcellulose medium using a pipette tip and spun in a cytospin onto glass slides. These cells were then stained with Wright-Giemsa stain and examined under a microscope (Nikon E800) to identify the different cell types present.
For flow cytometry, 6-day EBs were harvested, dissociated to form single cell suspensions, and 1 × 106 cells were labeled with antibodies for Flk-1, CD41 or c-Kit (Pharmingen) and analyzed on a FACS Calibur.
RNA from 6 day EBs was prepared by directly lysing them in Trizol reagent (Invitrogen) after washing them free of methylcellulose medium. Total RNA was used for cDNA synthesis (Invitrogen first strand synthesis kit using oligo dT primers) and quantitative PCR was performed on a Bio-Rad iCycler using Fam labeled TaqMan probes from Applied Biosystems.
The human HOXA9 insert was replaced with the human MEN1 coding sequence in the MSCV-based HOXA9 expression construct (Ernst et al., 2004b) to generate the MEN1 expression construct. Single cell suspensions were prepared from 6-day EBs, and aliquots of 2 × 106 cells were transfected with 10µg plasmid using the mouse ES cell nucleofector kit (Amaxa Biosystems). Cells from multiple transfections were combined and then plated in methylcellulose medium containing mSCF, IL3, IL6, and erythropoietin. Hematopoietic colonies were identified and counted after 10 to 12 days in culture.
To confirm that the transfection did result in the expression of the intended transcript, a fraction of transfected cells were collected two days post-transfection, and total RNA was extracted. PCR primers for MEN1 (F: ggtgaagaaggtctccgatg; R: cgtgagctggtgaagaaggt) or HOXA9 (F: ggtgactgtcccacgcttgac; R: gagtggagcgcgcatgaag).
We thank Patricia Ernst for the HOXA9 and Jyotshna Kanungo for the MEN1 expression constructs, David Bodine for helpful suggestions, Stacie Anderson for flow cytometry. This research was supported by the Intramural Research Program of the National Human Genome Research Institute, National Institutes of Health.