The transcription factor MZF1 differentially regulates murine Mtor promoter variants linked to tumor susceptibility

Mechanistic target of rapamycin (MTOR) is a highly conserved serine/threonine kinase that critically regulates cell growth, proliferation, differentiation, and survival. Previously, we have implicated Mtor as a plasmacytoma-resistance locus, Pctr2, in mice. Here, we report that administration of the tumor-inducing agent pristane decreases Mtor gene expression to a greater extent in mesenteric lymph nodes of BALB/cAnPt mice than of DBA/2N mice. We identified six allelic variants in the Mtor promoter region in BALB/cAnPt and DBA/2N mice. To determine the effects of these variants on Mtor transcription, we constructed a series of luciferase reporters containing these promoter variants and transfected them into mouse plasmacytoma cells. We could attribute the differences in Mtor promoter activity between the two mouse strains to a C → T change at the −6 position relative to the transcriptional start site Tssr 40273; a T at this position in the BALB promoter creates a consensus binding site for the transcription factor MZF1 (myeloid zinc finger 1). Results from electrophoretic mobility shift assays and DNA pulldown assays with ChIP-PCR confirmed that MZF1 binds to the cis-element TGGGGA located in the −6/−1 Mtor promoter region. Of note, MZF1 significantly and differentially down-regulated Mtor promoter activity, with MZF1 overexpression reducing Mtor expression more strongly in BALB mice than in DBA mice. Moreover, MZF1 overexpression reduced Mtor expression in both fibroblasts and mouse plasmacytoma cells, and Mzf1 knockdown increased Mtor expression in BALB3T3 and NIH3T3 fibroblast cells. Our results provide evidence that MZF1 down-regulates Mtor expression in pristane-induced plasmacytomas in mice.

Significant progress has been made in understanding how MTOR serves as a sensor that integrates a variety of exogenous cues to regulate cellular growth and metabolism, in both physiological and pathological conditions, including cancer. Deregulation of multiple elements in MTOR pathways has been reported in numerous cancers. Inhibition of MTOR kinase activity has been the aim of intense anti-cancer clinical research, because MTOR is often overexpressed or mutated in cancer cells (1).
Cancer development in human populations is a complex genetic trait. Pristane-induced plasmacytomas (PCTs) in mice provide an animal model system relevant to several human B cell malignancies, predominantly non-Hodgkin's lymphomas, including human plasma cell tumors, Burkitt's lymphoma, and multiple myeloma. In our previous studies, Mtor was identified as a tumor susceptibility/resistance gene in the pristane-induced PCT mouse model (5). BALB/cAnPt (BALB) mice are susceptible to PCT induction, whereas DBA/2N (DBA) and most other strains of mice are resistant (6). BALB mice carry a rare hypomorphic allele-encoding cysteine instead of a conserved arginine, R628C, in the HEAT domain of Mtor-leading to less kinase activity (5) and less phosphorylation of EIF4EBP1 (7) compared with the DBA allele. Furthermore, constitutive reductions in Mtor alter cell size, immune cell development, and antibody production (7), as well as limiting humoral responses in vivo (8,9). This work was supported by the Intramural Research Program of the National Institutes of Health, National Cancer Institute. The authors declare that they have no conflicts of interest with the contents of this article. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. This article contains Tables S1 and S2 and Figs. S1-S4. The nucleotide sequence(s) reported in this paper has been submitted to the DDBJ/GenBank TM /EBI Data Bank with accession number(s) MN076323. 1  cro ARTICLE In the current study, decreased levels of Mtor mRNA and protein expression were found in the mesenteric lymph nodes (MLNs) of mice after intraperitoneal injection of pristane. Mtor expression was more reduced in BALB than in DBA mice or BALB congenic mice carrying a DBA/2 allele of mTOR. Sequencing of the Mtor promoter region revealed six allelic variants between BALB and DBA mice in the region from Ϫ674 to ϩ28 relative to Tssr 40273. We therefore sought to identify the relevant cis-regulatory element (CRE) and the transcription factor that could repress Mtor and contribute to its differential expression in MLNs from BALB and DBA mice after pristane induction.

Transcription of the Mtor gene is differentially down-regulated by pristane treatment in plasmacytoma-susceptible versus plasmacytoma-resistant mice
Pristane treatment led to a significant reduction of Mtor mRNA levels in the MLNs of mice 18 days postinjection with pristane versus PBS (RT-qPCR; Fig. 1A). Mtor expression was reduced ϳ3-fold in BALB, which was significantly greater than the ϳ2-fold reduction in DBA. Similarly, in a separate experiment, pristane treatment of a pair of BALB congenic strains led to lower Mtor levels in splenic B cells isolated from tumor-susceptible C.D2Pnd7B mice harboring the BALB allele of Mtor, compared with levels in B cells from tumor-resistant, C.D2-Pnd7A mice harboring the DBA allele (Fig. 1B). Both of these strains of mice carry BALB/c alleles of genes throughout the genome (99% BALB), and 1% of their genome are from DBA Chr 4 near the Mtor locus. The segments introgressively backcrossed onto BALB contain the DBA allele of Mtor in the C.D2-Pnd7A mouse and a small region of chromosome 4 adjacent to, but not including DBA Mtor, in the C.D2-Pnd7B mouse (5).

Determination of allelic variants in the Mtor promoter between BALB and DBA mice and identification of MZF1 CREs in the Mtor promoter
To analyze the differential expression of Mtor, BALB and DBA Mtor promoters were cloned and sequenced (GenBank TM accession numbers: MN076323.1 and MN076324.1). Six allelic variants were identified in the promoter between positions Ϫ674 and ϩ28 (the transcriptional start site Tssr 40273 is designated ϩ1; see sequence alignment schema in Fig. 1C and Figs. S1A and S2). The allelic variants between DBA and BALB are at the following positions: Ϫ625 (A/T), Ϫ407 (A/C), Ϫ301 (G/A), Ϫ299 (C/A), Ϫ293 (G/A), and Ϫ6 (C/T).
Versions of BALB and DBA mTOR promoters from position Ϫ674 to ϩ28 were cloned into the pGL3 vector (Table S2 and to ϩ28, relative to Tssr 40273) between BALB and DBA mice. Dual-Luciferase reporter assays to identify allelic variants in the Mtor promoter responsible for its down-regulation. D, Luciferase activity in serially truncated versions of the BALB and DBA Mtor promoter. The t tests indicated significance in the difference in luminescence between BALB and DBA. E and F, chimeric versions of the Mtor promoter (Ϫ427 to Ϫ1) differing at one or more of five nucleotide variants were tested for luciferase activity. The t tests indicated a significant difference in luminescence (all constructs were compared with D5 (DBA/2N)). Numbers indicate nucleotide positions with respect to ϩ1 as the transcriptional start site. In all luciferase assays: 9 g of the mTOR promoter-pGL3-luc reporter construct and 1 g of pRL-SV40-Rluc plasmid DNA (transfection control) were used to co-transfect TEPC1165 (PCT) cells using electroporation (2 million cells per well in 6-well plates). The cells were harvested 24 h post-transfection for the Dual-Luciferase assay. Each data point is the mean Ϯ S.D. (error bar) of three independent assays (n ϭ 3). *, p Ͻ 0.05; **, p Ͻ 0.01. Red lines indicate the presence of the BALB variant at that nucleotide.

BALB/cAnPt Mtor promoter is sensitive to repression by MZF1
Fig. S1), and luciferase assays were used to measure promoter activity. These constructs were used to create a series of truncated and chimeric plasmids to evaluate the six divergent nucleic acids in the BALB versus DBA Mtor promoter. The results indicated that the DBA Mtor promoter was 3-fold more active than the BALB promoter when transfected into mouse plasmacytoma cell lines, TEPC1165 (10) (Fig. 1D), MOPC460 (Fig. S1E), and XRPC24 (X24; Fig. S1F), and the difference in promoter activity was attributable to the C-to-T divergence at the Ϫ6 nucleotide (nt) position (Fig. 1, E and F).
The GTGGGGAA sequence surrounding the Ϫ6 nt site in the BALB/c Mtor promoter is an established consensus MZF1binding site (MZF1 CRE; CRE1B (nt Ϫ22 to ϩ8, named 1B for the BALB Mtor promoter)) with 100% sequence identity based on MatInspector analysis and the recognition model of No Read Left Behind (NRLB) (11)(12)(13)(14) (Fig. 2A). The sequence (GCGGGGAA) at the same site in the DBA/2N Mtor promoter has 90% identity with the MZF1 binding site (MZF1 CRE1D for the DBA allele in the promoter). Co-transfection of Mzf1 along with the Mtor promoter variants (deletions and chimeras) in NIH3T3 cells led to greater down-regulation of promoter activity in constructs carrying the BALB allele (T) versus the DBA allele (C) at the Ϫ6 nt position, as measured by luciferase activity (Fig. 2, B and C). There were three other potential MZF1 CREs found (nt Ϫ14 to Ϫ44, Ϫ164 to Ϫ134, and Ϫ294 to Ϫ264) with at least 85% sequence identity ( Fig. 2A and Fig. S2), but they did not affect Mtor promoter activity (Fig. 2, B and C). Luciferase activity of the B7 promoter (T at Ϫ6) was repressed by co-transfection with Mzf1; in contrast, D5 promoter (C at Ϫ6) activity was not affected (Fig. 2, C and D). This was seen in both fibroblasts and plasmacytoma cells.
MZF1 reduced the luciferase activity of the Mtor promoter in a dose-dependent manner (Fig. 2, E-G). As the amount of Mzf1 plasmid was increased in co-transfections with the BALB Mtor promoter (B1 depicted in Figs. 1D and 2B) in TEPC1165 plasmacytoma cells, promoter activity was decreased (Fig. 2E). Furthermore, low doses of the Mzf1 plasmid did not significantly down-regulate the DBA D1 promoter (which has a single-nucleotide polymorphism (SNP) relative to BALB) (Fig. 2F); only the high dose of Mzf1 was able to down-regulate the DBA (D1 depicted in Figs. 1D and 2B) Mtor promoter (Fig. 2G).

MZF1 binds to the Mtor promoter CRE encompassing the variant nucleotide (؊6) closest to the transcription start site, designated Tssr 40273
A series of electrophoretic mobility shift assays (EMSAs) were carried out with nuclear extracts (NEs) from either Figure 2. MZF1 cis-regulatory elements and differential promoter activity associated with the ؊6 nucleotide variant in the Mtor promoter. A, schema of four potential MZF1 CREs in the Mtor promoter (Ϫ674 to ϩ28). The conserved nucleotides were identified using the recognition model/algorithm denoted as NRLB. B and C, Luciferase activity assays in NIH3T3 cells transfected with serially truncated (B) or chimeric (C) versions of the Mtor promoter in pGL3-luc (24-well plates) along with a pTCP plasmid containing full-length cDNA of Mzf1 (1:1 ratio, 2 g/well each), and pRL-SV40-Rluc (0.5 g/well) to provide an internal control for normalization of luciferase activity. D, Luciferase activity of the D5 and B7 Mtor promoters co-transfected in mouse plasmacytoma cell lines (TEPC1165, MOPC460, and XRPC24) with Mzf1 (2 g). E, dose responsiveness of luciferase activity (Luc/Rluc) of the Mtor promoter (pGL3-B1 plasmid, 2 g/well) after co-transfection by electroporation with pRL-SV40-Rluc plasmid (0.5 g/well) and a Mzf1 expression plasmid (pTCP-Mzf1, concentration ranging from 0 to 2.5 g/well) into TEPC1165 plasmacytoma cells (2 ϫ 10 6 cells/well in 6-well plates). F and G, Luciferase activity of the B1 and D1 Mtor promoters (pGL3-B1 plasmid (2 g/well)/Renilla luciferase values obtained from co-transfection by electroporation with pRL-SV40 plasmid DNA (0.5 g/well) and either pTCP-Mzf1 in low concentrations (F, 2 g/well) or in high concentrations (G, 5 g/well) in TEPC1165 plasmacytoma cells (2 ϫ 10 6 cells/well in 6-well plates). Cells were harvested at 24 h post-transfection in the Dual-Luciferase assays. Each data point is the mean Ϯ S.D. of at least three independent assays (n ϭ 6 or 3). The t tests indicate significant differences in luminescence between control (pTCP empty vector, solid bars or circles) versus Mzf1 overexpression (pTCP-Mzf1, open bars or circles). **, p Ͻ 0.01; *, p Ͻ 0.05. Red, BALB; blue, DBA.

BALB/cAnPt Mtor promoter is sensitive to repression by MZF1
BALB and DBA splenic B cells (Fig. 3B) or TEPC1165 plasmacytoma cells (Fig. 3, C-F) to examine the binding of MZF1 to its putative CREs (Fig. 3A) in the Mtor promoter. NEs isolated from splenic B cells, TEPC1165 cells, or TEPC1165 cells with Mzf1 transfection were used as the source of MZF1 protein. Formation of a DNA-protein complex could only be detected with CRE1B (T allele at Ϫ6 nt in BALB) incubated with NE from BALB splenic B cells (Fig. 3B). The addition of a 200-fold molar excess of unlabeled CRE1B "cold" probe abolished formation of the complex (Fig. 3C, lane 3). In contrast, excess unlabeled CRE1D (contains C allele at Ϫ6 SNP in DBA) or CRE2-4 did not completely abolish complex formation (Fig. 3C, lanes 4 -7). Biotin-labeled probes for CRE2-4 did not detect any specific protein complex (data not shown) when performing EMSA with NE of T1165 or splenic B cells. Complex formation was enhanced in a dose-dependent manner in NE of Mzf1 transfected TEPC1165 cells (Fig.  3D). The BALB probe (CRE1B) bound more efficiently than the DBA probe in NE from TEPC1165 (Fig. 3E), and unlabeled mutated probe (unlabeled 5Ј-CCGGCCTAGAAGAC-AGTTTTTAAGGCGGGT-3Ј) could not compete with the BALB/c probe (CRE1B) for the formation of a protein complex with T1165 NE (Fig. 3F, lane 3).

MZF1 interacts with an Mtor promoter CRE and downregulates Mtor expression in vivo
To test whether MZF1 was recruited to the CRE including the Ϫ6 nt site of the Mtor promoter, ChIP analysis was performed with PCT (TEPC1165) cells. ChIP-enriched DNA carrying CRE1B (MZF1 CRE in BALB promoter) was amplified by PCR (Fig. S3, A and B) and quantified by qPCR ( Fig. 4A and Anti-MZF1 antibody, but not IgG antibody (negative control), precipitated the MZF1 CRE1B from TEPC1165 cells. MZF1 bound the MZF1 CRE1B in the nt Ϫ22 to ϩ8 region of the Mtor promoter ex vivo.

BALB/cAnPt Mtor promoter is sensitive to repression by MZF1
were reduced to an even greater extent in cells containing the BALB T allele, compared with the C allele at nt Ϫ6. ChIP-qPCR analyses of DNA from NIH3T3 and BALB3T3 cells revealed a 13-fold increase in the recruitment of MZF1-CRE1B in BALB3T3 cells compared with MZF1-CRE1D in NIH3T3 cells (Fig. 4E).
Consistent with MZF1 regulation of the Mtor promoter, siRNA-mediated knockdown of Mzf1 resulted in increased Mtor mRNA (Fig. 4F) and MTOR protein (Fig. 4G) levels. These findings are consistent with MZF1 functioning as a repressor of Mtor transcription by binding to its CRE surrounding the Ϫ6 nt polymorphic site.

Discussion
Promoter activity studies involving a series of Mtor promoter region deletion and chimeric plasmids, coupled with DNAbinding assays, identified MZF1 as a transcription factor capable of repressing Mtor transcription. The MZF1 CRE at nt Ϫ22 to ϩ8 encompassed a variant SNP (T in BALB and C in DBA) at nt Ϫ6. The BALB variant (T) at the Ϫ6 position from the Tssr 40273 start site was a direct match with the MZF1 consensus binding site, and MZF1 demonstrated preferential binding to the BALB allele. ChIP assays confirmed MZF1 recruitment to the Mtor promoter. Our studies provide insight into under-standing the genetic differences between BALB and DBA Mtor promoters. Further analyses would be required to gain a complete understanding of Mtor regulation.
Pristane treatment, used to induce PCTs, caused Mzf1 levels to increase and Mtor levels to decrease. Importantly, pristaneinduced Mzf1 levels were similar in BALB and DBA mice, whereas the pristane-induced decrease in Mtor was greater in BALB and the PCT-susceptible congenic strain, Pnd7B, than in DBA. Our in vitro studies to overexpress or silence Mzf1 demonstrated concomitant changes in Mtor expression. When Mzf1 was overexpressed, BALB Mtor promoter activity was repressed, whereas siRNA experiments that decreased Mzf1 levels resulted in increased Mtor expression. Thus, our studies have identified MZF1 as a new transcriptional repressor of Mtor in cells exposed to the inflammatory agent pristane.
MZF1 is a member of the SCAN-zinc finger family of transcription factors (11) and can function both as trans-activator and trans-repressor depending on the intracellular environment (15)(16)(17)(18)(19)(20)(21). Although the role of MZF1 was initially studied in myeloid differentiation and leukemia, the factor now appears to be involved in the pathogenesis of several solid tumors (22). The contribution of MZF1 to tumorigenesis is diverse, because it may induce oncogenic or tumor suppressor effects in hema-

BALB/cAnPt Mtor promoter is sensitive to repression by MZF1
topoietic and nonhematopoietic cells. MZF1 has been implicated in mediating the migration and invasion of cancer cells by suppressing the activity of certain gene promoter regions in vivo and in vitro (17). The relative oncogenic activity of MZF1 is determined by the aggregated effects produced by the increases and decreases in gene expression, phosphorylation and SUMOylation modifications, and co-activating and co-repressing molecules (22).
The MTOR signaling pathway controls key cellular processes, such as metabolism, growth, motility, and survival, and is a frequently dysregulated pathway in cancer (3). Aberrant activation of the pathway via multiple mechanisms has been detected in a range of tumor types, including diverse genomic alterations of the pathway's components (23). Somatic cancerassociated mutations in MTOR are often hyperactivating (24), and at least one promoter polymorphism (rs2295080 at nt Ϫ64) has been associated with an altered risk of developing renal (25), gastric (26), and genitourinary cancers, as well as acute childhood leukemia (27). Previously, only one transcription factor, NFE2L2 (nuclear factor erythroid 2-related factor, also called NRF2), has been reported to regulate Mtor and only when the phosphatidylinositol 3-kinase pathway is intact (28). The promoter sequence surrounding the rs2295080 polymorphism does not have putative consensus binding sites for either NFE2L2 or MZF1.
Our studies suggest that the amount of MZF1 protein is an important driver in the differential regulation of Mtor promoter variants, as evidenced by the ability of low-dose overexpression of Mzf1 to preferentially lower Mtor promoter levels associated with the BALB T allele. The Mtor promoter carrying the DBA C allele was only repressed by high-dose overexpression of Mzf1, likely because the DBA promoter does not have the optimal consensus binding site present in the BALB promoter. These data, coupled with our previous studies on Mtor in PCT development, reveal that BALB mice have both a coding region polymorphism (nt C1977T; AA R628C) (5), which lowers its kinase activity, and now at least one identified functional SNP that can, in addition, contribute to lower Mtor gene expression in BALB mice.
This compound allelic variation in both Mtor coding and regulatory sequences, along with previously reported allelic variants in Cdkn2a (exon 2: p16) coding (29) and promoter (30) sequences, contributes to the complex genetics associated with PCT susceptibility in BALB mice (6). Thus, hypomorphic activity of BALB Cdkn2a (p16) and Mtor alleles is associated with tumorigenesis, suggesting that both p16 and Mtor can act as tumor suppressors in PCT development in an allele-dependent manner. Although this may be expected for p16, which is widely recognized as a tumor suppressor, it is counter to the prevailing notion that Mtor activity is always growth-promoting. Interestingly, the G allele at rs2295080 in the human MTOR promoter is associated simultaneously with reduced promoter activity and increased risk for childhood acute leukemia but decreased risk for genitourinary cancers (27), thus highlighting the possibility of differential, tissue-specific susceptibility. Recently, Villar et al. (31,32) have shown that the activation of mTORC1 by glutaminolysis during nutritional imbalance inhibits autophagy and induces apoptosis in cancer cells, thus supporting a tumor suppressor role under stress conditions. Furthermore, Villar et al. (31,32) showed that the MTOR inhibitor, rapamycin, can actually block glutaminolysis-mediated apoptosis, leading to improved cancer cell survival. Thus, as in the murine pristaneinduced PCT system, there appear to be human tumors where MTOR may play an inhibitory role.

Mice
Mice were maintained under conventional, closed barrier conditions at the National Cancer Institute under animal protocol LG-009. Experiments were conducted according to the institutional ethical guidelines for animal experiments and safety guidelines for gene manipulation experiments. Mesenteric lymph nodes were collected from 8-week-old BALB/cAn and DBA/2N mice, 18 days postinjection with pristane (0.5 ml). Pristane induces chronic inflammation in BALB/c mice (33). BALB/c congenic mice, C.D2-Pnd7A mice, and C.D2-Pnd7B mice were generated by introgressive backcrossing (N12) of segments of DBA DNA from Chr 4 onto the BALB background (34). C.D2-Pnd7A mice are relatively resistant to plasmacytoma induction and carry a segment of DBA DNA encompassing the Mtor locus; in contrast, C.D2-Pnd7B mice are as susceptible as BALB/c mice and carry a segment of DBA DNA adjacent to, but not including the Mtor locus (34). Splenic B cells were prepared from spleens of DBA and BALB/c congenic mice using the protocol for a CD45R(B220) MicroBeads kit (catalog no. 130-049-501, Miltenyi Biotec).

Reagents and antibodies
All the reagents were of analytical grade and purchased from Merck or Sigma-Aldrich. All primers and oligonucleotides were ordered from Eurofins Genomics. Restriction enzymes were purchased from New England Biolabs. The anti-MZF1 antibodies (sc-66991 for Western blotting and sc-66991x for ChIP assay plus control rabbit IgG; sc-2027) were purchased from Santa Cruz Biotechnology. Another anti-MZF1 antibody (Western blotting) was purchased from Abcam (ab64866). The anti-mTOR antibody (catalog no. 2972) and anti-␤-actin antibody (catalog no. 4970) were from Cell Signaling Technology.

RNA extraction and qPCR
Total RNA was prepared using TRIzol reagent (Invitrogen) and purified using RNeasy PowerClean Pro cleanup kit (Qia-

BALB/cAnPt Mtor promoter is sensitive to repression by MZF1
gen). cDNAs were made with TaqMan RT reagents kit (Applied Biosystems: N808-0234). Expression of 18S rRNA from the same sample under analysis was used as an endogenous control. qPCR was performed using SYBR Green Master PCR Mix (part no. 4309155, ABI 7500, Applied Biosystems). The results are expressed as the ratios of values for treated samples to untreated samples and normalized to their 18S levels. Threshold values were converted to relative fold change using the 2 Ϫ⌬⌬CT method. The primer sets used for estimation of transcript levels for different genes are described in Table S1.

Cloning and sequencing of Mtor promoter
BALB/c and DBA/2N Mtor promoters were cloned following PCR amplification of tissue DNA from BALB/cAn and DBA/2N mice, respectively. The following primers were used to generate Mtor promoter fragments: forward, GCTCCGGCTATTGCCC-GTTG, and reverse, GAGCAGATCCGCCAGCCTG. Fragments

Plasmid constructs
To functionally characterize the 5Ј-flanking region of the BALB/cAn and DBA/2N Mtor gene, a series of 5Ј deletions or chimeric sequences in the Ϫ674 to ϩ28 mostly upstream sequence were amplified using PCR from the subcloned pCR-TOPO-Mtor-1p-BALB/c and pCR-TOPO-Mtor-1p-DBA/2N plasmids containing Mtor promoter region sequences. The upstream primers contained a 5Ј-flanking MluI restriction enzyme site, whereas the downstream primers were flanked by a BglII site. The fragments of the 5Ј region or chimeric sequences of the mTOR were inserted into the luciferase reporter plasmid, pGL3-basic vector (Promega). All plasmids were analyzed and confirmed by restriction digestion and DNA sequencing. The strategy and primer sets used for producing mTOR promoter reporter plasmids are described in Fig. S2 and Table S2.

Luciferase assays
Mouse plasmacytoma cell lines were transiently transfected by electroporation (Amaxa nucleofector kit from Lonza Biosci-ence). NIH3T3 and BALB3T3 cells, grown in 24-well plates, were transfected using Lipofectamine 2000 or 3000 (Invitrogen). The cells were transfected with different reporter constructs carrying the Mtor promoter variants (driven by firefly luciferase, Luc) and harvested 24 h post-transfection for luciferase assays following Dual-Luciferase reporter assay system (Sigma) protocols; activity was measured with a FLUOstar Omega microplate reader (BMG Labtech). The pRL-SV40 (Renilla luciferase reporter, Rluc) was used as an internal transfection control and co-transfected with the reporter plasmids. Promoter activities are the means Ϯ S.E. of 3-8 transfections (luciferase expression activity relative to Renilla luciferase activity, Luc/Rluc).

Western blots
Protein lysates were made with radioimmune precipitation assay lysis buffer (150 mM NaCl, 1.0% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 50 mM Tris, pH 8.0). Protein concentrations were determined by BCA reagent (Pierce). Proteins were mixed with 4ϫ loading buffer (LDS sample buffer, Novex) containing 5% ␤-mercaptoethanol, and separated on 4 -20% Tris-glycine polyacrylamide SDS-PAGE gels (Invitrogen); proteins were transferred to polyvinylidene difluoride membranes using the iBlot 2 dry blotting system (Invitrogen). The membranes were blocked with 5% (w/v) dry milk in TBST at room temperature for 1 h and then incubated with antibody with gentle rocking at 4°C for overnight. After washing three times with TBST for 5 min each, the membranes were incubated with horseradish peroxidase-conjugated goat anti-rabbit IgG secondary antibody with gentle rocking at room temperature for 1 h. Membranes were washed five times with TBST for 5 min each and then detected with SuperSignal West Dura extended duration substrate (Thermo Fisher). In all experiments, ␤-actin was used as an internal reference, and ImageJ software was used for quantifying fluorescence of area integrated intensity.

Electrophoretic mobility shift assay
NEs were prepared from TEPC1165 cells and splenic B cells with NE buffer (NE-PER nuclear and cytoplasmic extraction kit; Pierce) supplemented with protease and phosphatase inhibitors. EMSA was performed using manufacturer's protocols (LightShift Chemiluminescent EMSA kit; Pierce). NEs were incubated for 20 min at room temperature in a final 20-l reaction volume containing 10 mM Tris-HCl (pH 7.5), 50 mM KCl, 1 mM DTT, 0.2 mM EDTA, 2.5% (v/v) glycerol, 50 ng of poly(dI-dC), 0.05% Nonidet P-40, and 20 fmol of Biotin 3Ј-labeled oligonucleotide probe in the absence or presence of various competitors (200-fold cold probes). The reaction products were electrophoresed on a 6% nondenaturing polyacrylamide gel in TBS buffer. The gels were transferred to nylon membranes and cross-linked at 120 mJ/cm 2 using a UV cross-linker equipped with 254-nm bulbs and visualized using detection of Biotin-labeled DNA according to the manufacturer's protocols (LightShift Chemiluminescent EMSA kit (Pierce)). For quantitation, the free probe band was used as an internal reference, and ImageJ software was used to quantify the fluorescence of area-integrated intensity.

ChIP assay
ChIP experiments were performed according to manufacturer's instructions (Pierce agarose ChIP kit) using anti-MZF1 antibody and normal rabbit IgG as a negative control with TEPC1165, NIH3T3, and BALB3T3 cells. In brief, the ChIP experiment was performed to detect the effects of MZF1 on the promoter activity of Mtor in the cell lines. The cells were crosslinked with 1% formaldehyde for 10 min at room temperature. Glycine was added to terminate the fixation, and cells were washed and collected with ice-cold PBS. The cells were collected, and the cell pellets were broken up with membrane extraction buffer containing protease/phosphatase inhibitors. The nuclei were collected and digested with MNase. Digested chromatin was collected by centrifugation. An immunoprecipitation experiment was conducted following the manufacturer's instructions for the Pierce agarose ChIP kit (Thermo Scientific). Five g of either rabbit anti-MZF1 antibody or normal rabbit IgG were used in immunoprecipitation reactions for 2 h at 4°C with mixing. The antibody-protein-DNA complex was enriched by incubation with ChIP grade protein A/G-agarose beads for 1.5 h at 4°C with mixing. The bound DNA was purified following the kit's instructions. Immunoprecipitated genomic DNA fragments were amplified and quantified by qPCR using the primers (forward, 5Ј-GCCTGAAGACAGTG-GGGAAGG-3Ј; and reverse, 5Ј-CCACACTGGCTGGGAGT-CTA-3Ј; the fragment is between Ϫ20 and ϩ145).

Cell transfection and retroviral particle production and transduction
NIH 3T3 and BALB3T3 cells were transfected with 20 g of pMYs-MZF1-IRES-EGFP or pMYs-IRES-EGFP plasmid by Lipofectamine 3000 (Invitrogen) in 10-cm dishes when the cells were 50 -60% confluent. DNA samples (for ChIP analysis), protein lysates, and total RNAs were made 48 h post-transfection. Retroviruses expressing MZF1-GFP or GFP were prepared by adapting previously described methods (35). PlatE cells were transfected with pMYs-MZF1-IRES-EGFP or pMYs-IRES-EGFP plasmids by Lipofectamine 3000 (Invitrogen), and recombinant retroviral particles were collected 24 -36 h after transfection. T1165 cells were transduced with retroviral particles by spin inoculation (2500 rpm for 90 min), cultured for 48 h, and then sorted for GFPpositive cells with BD FACSAria Violet (Flow Cytometry Core Facility, NCI, National Institutes of Health).

siRNA knockdown
siRNA-mediated knockdown of Mzf1 was performed using the DharmaFECT 1 transfection reagent (horizon/Dharmacon, catalogue no. T-2001-02) according to the provided protocol. NIH 3T3 and BALB3T3 cells were transfected with mouse Mzf1 siRNA (Thermo Fisher, siRNA code s99636, catalogue no. 4390771) at 25 nM final concentration in 6-well plates. The transfection medium was replaced with complete medium after 6 -8 h, and incubation continued for an additional 48 h. At this point cells were collected and analyzed for mRNA and protein expression.

Bioinformatics and statistical analysis
Putative MZF1 CREs in the promoter of Mtor were predicted using MatInspector (Genomatix), and NRLB was used for generating the recognition model of the ⌬⌬G/RT coefficients for MZF1 cis-regulatory elements. All experiments were performed at least three times. Statistical significance of the differences between the control and treatment was analyzed by Student's t test. The data are presented as means Ϯ S.D. Error bars represent the means Ϯ S.E. The values were considered to be statistically significant when p Ͻ 0.05 (*) and p Ͻ 0.01 (**).