Gfi1 upregulates c-Myc expression and promotes c-Myc-driven cell proliferation

Gfi1 is a zinc-finger transcriptional repressor that plays an important role in hematopoiesis. When aberrantly activated, Gfi1 may function as a weak oncoprotein in the lymphoid system, but collaborates strongly with c-Myc in lymphomagenesis. The mechanism by which Gfi1 collaborates with c-Myc in lymphomagenesis is incompletely understood. We show here that Gfi1 augmented the expression of c-Myc protein in cells transfected with c-Myc expression constructs. The N-terminal SNAG domain and C-terminal ZF domains of Gfi1, but not its transcriptional repression and DNA binding activities, were required for c-Myc upregulation. We further show that Gfi1 overexpression led to reduced polyubiquitination and increased stability of c-Myc protein. Interestingly, the levels of endogenous c-Myc mRNA and protein were augmented upon Gfi1 overexpression, but reduced following Gfi1 knockdown or knockout, which was associated with a decline in the expression of c-Myc-activated target genes. Consistent with its role in the regulation of c-Myc expression, Gfi1 promoted Myc-driven cell cycle progression and proliferation. Together, these data reveal a novel mechanism by which Gfi1 augments the biological function of c-Myc and may have implications for understanding the functional collaboration between Gfi1 and c-Myc in lymphomagenesis.


Results
Gfi1 upregulates c-Myc protein level. The expression of c-Myc is tightly regulated in normal lymphoid cells. Notably, most lymphomas with MYC overexpression originate from cells that do not normally express MYC, suggesting that these tumors have developed additional oncogenic events to disrupt the MYC regulatory mechanisms 18 . In our previous studies 15, 16 , we noticed that Gfi1 significantly increased the protein level of c-Myc transiently expressed in Hela cells. As Gfi1 collaborates with c-Myc in lymphomagenesis, we further examined the role of Gfi1 in regulating c-Myc expression. Hela cells were transiently transfected with increasing amounts of c-Myc without or with fixed amount of Gfi1 or vice versa. The expression of c-Myc protein was examined by Western blot analysis. As shown in Fig. 1A,B, Gfi1 markedly augmented the level of c-Myc protein in all the experimental conditions, indicating that Gfi1 was required for efficient expression of c-Myc protein in Hela cells. The effect of Gfi1 on c-Myc expression was specific as Gfi1 did not upregulate the expression of other transcription factors, including AML1, C/EBPε, STAT5 and Miz-1 (supplementary Fig. 1S).
We next addressed whether Gfi1 augmented c-Myc expression in hematopoietic cells. We generated Ba/F3 (BaF/MycER/Gfi1) and 32D (32D/MycER/Gfi1) cells that stably expressed c-MYC/estrogen receptor ligand binding domain fusion protein (MycER) and inducibly expressed Gfi1 in response to doxycycline (Dox). The Dox-induced expression of Gfi1 in Ba/F3 cells was lower than the expression of endogenous Gfi1 in human  www.nature.com/scientificreports/ myeloid leukemic HL60 and U937 cells, which expressed relatively high levels of Gfi1 (supplementary Fig. 2S). BaF/MycER/Gfi1 and 32D/MycER/Gfi1 cells were treated with 4-hydroxytamoxifen (4-HT), Dox or both for 24 h prior to evaluation of protein expression by Western blot analysis. As expected, treatment with 4-HT increased the MycER protein levels in the two cell lines (Fig. 1C, D). Irrespective of 4-HT treatment, induction of Gfi1 expression with Dox resulted in a significant increase in MycER protein level in the nuclei of the cells. Dox and 4-HT had no significant effect on the expression of MycER mRNA in Ba/F3 although 4-HT appeared to inhibit MycER mRNA expression in 32D cells. Together, these results demonstrated that Gfi1 upregulated MycER protein expression in hematopoietic cells.

Gfi1 reduces the polyubiquitination of, and increases the stability of c-Myc protein. c-Myc
protein is highly unstable with a half-life of less than 30 min in non-transformed cells and is rapidly degraded mainly through the ubiquitin-proteasome pathway 19,20 . To examine whether Gfi1 stabilized c-Myc protein, BaF/MycER/Gfi1 cells were treated with Dox for 24 h to induce Gfi1 expression, followed by treatment with cycloheximide (CHX) to block protein synthesis. MycER protein level rapidly declined and became barely detectable at 30 min in BaF/MycER/Gfi1 cells untreated with Dox, but was stable for at least 45 min in Dox-treated cells ( Fig. 2A). To examine whether Gfi1 stabilized endogenous c-Myc, we generated Ba/F3 cells expressing the Dox-inducible Gfi1 (BaF/Gfi1). As shown in Fig. 2A, Dox-induced expression of Gfi1 (see Fig. 5A) led to the significantly increased stability of endogenous c-Myc protein. Interestingly, when Hela cells transfected with c-Myc alone were treated with different amounts of proteasome inhibitor MG132, the expression of c-Myc protein increased to a level similar to that in cells transfected with both c-Myc and Gfi1 (Fig. 2B), suggesting that Gfi1 may inhibit c-Myc degradation mediated by the ubiquitin-proteasome pathway. We further performed ubiquitination assay to assess the effect of Gfi1 on c-Myc polyubiquitination. As shown in Fig. 2C, co-expression The N-terminal SNAG domain and C-terminal ZF domains of Gfi1 are required for c-Myc upregulation. We examined which activities or domains of Gfi1 were required for c-Myc upregulation.
The Gfi1 N382S mutant is defective in DNA binding whereas the P2A mutant lacks the repressor activity [21][22][23] . As shown in Fig. 3A, both Gfi1 mutants upregulated c-Myc expression as efficiently as the wild type (WT) Gfi1 in Hela cells. Gfi1 has an N-terminal SNAG domain and 6 ZF domains at its C-terminus that are involved DNA binding and protein-protein interactions. We generated the Gfi1 mutants lacking the N-terminal SNAG domain or with progressive truncation of the C-terminal ZF domains (Fig. 3B). Deletion of the SNAG domain markedly diminished c-Myc upregulation by Gfi1. Interestingly, progressive truncation of the C-terminal ZF domains resulted in the gradual loss of the ability of Gfi1 to upregulate c-Myc, which was associated with an increase in the expression of the mutant Gfi1 proteins (Fig. 3C, D). Thus, both the SNAG domain and the ZF domains, but not the DNA binding and repressor activities of Gfi1, were required for c-Myc upregulation.
Fbxw7 and Skp2 are not the major E3 ubiquitin ligases involved in c-Myc upregulation by Gfi1. It has been shown that Gfi1 is also polyubiquitinated and degraded through the ubiquitin-proteasome pathway, which takes place at its C-terminal ZF domains 24 , consistent with the augmented expression of the C-terminally truncated Gfi1 mutants. We addressed the possibility that Gfi1 might upregulate c-Myc protein expression by competing with c-Myc for a common E3 ubiquitin ligase(s). A number of c-Myc E3 ubiquitin ligases, including Fbxw7, Skp2, CHIP,FBX29, FBXO32, PirH2, TRIM32 and Truss have been reported to reduce c-Myc protein expression 19,20 . Interestingly, Gfi1 has been identified as a substrate of Fbxw7 25 . c-Myc degradation catalyzed by Fbxw7 is dependent on c-Myc threonine 58 26 . We therefore examined the effect of Gfi1 on the expression of c-Myc T58A, which is resistant to Fbxw7-mediated degradation. As shown in supplementary  Gfi1 stabilizes c-Myc protein independent of Miz-1. We previously showed that Gfi1 indirectly interacts with c-Myc through Miz-1 15,16 . Interestingly, Miz-1 has been shown to stabilize c-Myc 29 , raising the possibility that Gfi1 might stabilize c-Myc protein through Miz-1. We therefore examined the effect of Gfi1 on the expression of c-Myc-V394D mutant, which is defective in Miz-1 interaction, but retains the abilities to dimerize with Max and activate transcription 30,31 . As shown in Fig. 4, Gfi1 upregulated the level of c-Myc-V394D transiently expressed in Hela cells and the level of MycER V394D stably expressed in Ba/F3 cells. Thus, it appears that Gfi1-mediated upregulation of c-Myc was independent of Miz-1.

Gfi1 upregulates the mRNA and protein levels of endogenous c-Myc. Since Gfi1 increased
the stability of endogenous c-Myc protein ( Fig. 2A), we were interested to know whether Gfi1 regulated the expression of endogenous c-Myc. Indeed, Dox treatment of BaF/Gfi1 cells augmented the protein expression of endogenous c-Myc (Fig. 5A). Unexpectedly, however, Gfi1 also increased the mRNA level of endogenous c-Myc. To explore whether Gfi1 regulated c-Myc expression in lymphoma cells, we expressed the Dox-inducible Gfi1 in human Burkitt′s lymphoma Ramos cells (Ramos/Gfi1), which carried the characteristic t(8;14) chromosomal translocation that linked c-MYC to the immunoglobulin heavy-chain gene (IgH). As in BaF/Gfi1 cells, Gfi1 augmented the mRNA and protein levels of endogenous c-Myc in Ramos/Gfi1 cells (Fig. 5B).
We previously showed that human leukemic HL60 and U937 cells expressed high levels of endogenous GFI1 15,16 . We examined the effect of GFI1 knockdown on c-MYC expression in these cells. As shown in Fig. 6A, GFI1 knockdown was associated with the decreased levels of c-MYC mRNA and protein. It appears that the effect of GFI1 knockdown on c-MYC protein expression was more dramatic than on mRNA expression in both cell lines. Furthermore, c-Myc mRNA and protein levels were decreased in Lin -BM cells from Gfi1 -/mice as compared to the levels in Gfi1 +/+ BM cells (Fig. 6B), which was associated with decreased expression of c-Myc-activated target genes in Gfi1 -/-BM cells (supplementary Fig. S5). Together, these data indicate that Gfi1 upregulate c-Myc expression at both mRNA and protein levels.  www.nature.com/scientificreports/ proliferation, Ramos/Gfi1 cells were cultured in the absence or presence of Dox. As shown in Fig. 7B, induction of Gfi1 expression with Dox led to a significant increase in cell proliferation, as determined by the MTS assay and direct cell counting. Together, these data indicate that Gfi1 acted in collaboration with c-Myc to drive cell proliferation.

Discussion
Gfi1 has been shown to act in collaboration with c-Myc in lymphomagenesis, but the underlying molecular mechanisms remain incompletely understood. We previously showed that Gfi1 indirectly interacts with c-Myc through Miz-1 and collaborates with c-Myc in repressing p15 INK4B , p21 Cip1 and p27 Kip115, 16 . In this paper, we have shown that Gfi1 upregulates c-Myc expression at both mRNA and protein levels, and augments the effect of c-Myc on cell proliferation. We have further demonstrated that Gfi1 stabilizes c-Myc protein, likely by reducing its polyubiquitination. The ability of Gfi1 to upregulate c-Myc protein is dependent on its N-terminal SNAG and C-terminal ZF domains, but not on its DNA binding and repressor activities. Notably, although Miz-1 has been shown to stabilize c-Myc, Gfi1 was able to upregulate the expression of the c-Myc V394D mutant, which is defective in Miz-1 interaction, suggesting that Miz-1 may not be involved in Gfi1-mediated stabilization of c-Myc protein. Our data reveal a novel mechanism by which Gfi1 augments the biological function of c-Myc and may have implications for understanding the role of Gfi1 in Myc-mediated lymphomagenesis.
In addition to c-Myc, Gfi1 has been shown to stabilize another transcription factor GATA3, which was associated with reduced GATA3 polyubiquitination 34 . Interestingly, as in the case of c-Myc, the N-terminal SNAG domain of Gfi1, but not its repressor activity, is required for GATA3 stabilization. It has been shown that the SNAG domain of Gfi1 is required for its nuclear localization 23 and to interact with the histone demethylase LSD1 and the corepressor Co-REST 35 . Notably, substitution of the proline at position 2 for alanine (P2A) disrupts Gfi1 interactions with LSD1 and Co-REST without altering its nuclear localization 23 . Thus, it appears that Gfi1 needs to be in the nucleus to stabilize c-Myc and GATA3, which are also nuclear proteins. It was not examined in the previous study whether the C-terminal ZF domains are required for GATA3 stabilization.
It is unknown how Gfi1 influences c-Myc polyubiquitination. The fact that the C-terminal ZF domains of Gfi1, which are required for its polyubiquitination and degradation 24 , are also required for c-Myc upregulation raises the possibility that Gfi1 and c-Myc might compete for binding to a common E3 ubiquitin ligase(s). Interestingly, both Gfi1 and c-Myc have been identified as the substrates of Fbxw7 25 . However, our data indicate that Fbxw7 may not play a key role in Gfi1-mediated regulation of c-Myc protein. Skp2 is another major regulator of c-Myc protein stability. We have shown that Gfi1 efficiently upregulated the expression of c-Myc ΔMBII mutant, which is significantly impaired in its ability to interact with Skp2, making it less likely that Skp2 is the major player in c-Myc upregulation by Gfi1. It remains to be determined whether other E3 ubiquitin ligases or other mechanisms are involved in Gfi1-mediated c-Myc upregulation. www.nature.com/scientificreports/ The mechanism by which Gfi1 upregulates the mRNA level of c-Myc is also unknown. Gfi1 has been shown to function mainly as a transcriptional repressor. However, the possibility cannot be excluded that Gfi1 may bind to c-Myc promoter to activate its transcription. Interestingly, analysis of the ChIP-seq data (GSE31657) submitted by Möröy's research group 12 indicates that Gfi1 may bind to the promoters of c-Myc in murine hematopoietic progenitor cells (data not shown). An alternative possibility is that Gfi1 may repress a negative regulator of c-Myc, such as a transcription factor or microRNA, leading to upregulation of c-Myc mRNA. In this aspect, it is of note that Gfi1 has been shown to repress microRNA-196b (miR-196b) and miR-21 36 , and miR-196b has been shown to suppress c-Myc mRNA expression 37 . Further studies are needed to address how Gfi1 upregulates c-Myc mRNA expression.

Real-time reverse transcription polymerase chain reaction (qRT-PCR).
Total RNA was extracted using TRIzol reagent (Invitrogen) and cDNA was synthesized using the GoScript Reverse Transcription System and Oligo(dT)15 primer (Promega, Madison, WI). qRT-PCR was performed using the SsoFast EvaGreen Supermix kit (Bio-Rad) and the relative levels of mRNAs for the different myeloid differentiation markers were normalized to GAPDH mRNA expression. The following primer sets were used: Statistics. All statistical analyses were performed using GraphPad Prism software (GraphPad Software, La Jolla, CA, USA). Data are presented as mean ± SD with * indicating p < 0.05, ** < 0.01, and *** < 0.001 in all figures.

Data availability
No datasets were generated or analyzed during the current study. www.nature.com/scientificreports/ Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.