|Home | About | Journals | Submit | Contact Us | Français|
The universal cyclin-Cdk inhibitor p27Kip1 functions as a tumor suppressor and reduced levels of p27Kip1 connote poor prognosis in several human malignancies. p27Kip1 levels are predominately regulated by ubiquitin-mediated turnover of the protein, which is marked for destruction by the E3 ubiquitin ligase SCFSkp2 complex following its phosphorylation by the cyclin E-Cdk2 complex. Binding of phospho-p27Kip1 is directed by the Skp2 F-box protein, and this is greatly augmented by its allosteric regulator Cks1. We have established that programmed expression of c-Myc in the B cells of Eμ-Myc transgenic mice triggers p27Kip1 destruction by inducing Cks1, that this response controls Myc-driven proliferation, and that loss of Cks1 markedly delays Myc-induced lymphomagenesis and cancels the dissemination of these tumors. Here, we report that elevated levels of Skp2 are a characteristic of Eμ-Myc lymphomas and of human Burkitt lymphoma that bear MYC/immunoglobulin chromosomal translocations. As expected, Myc-mediated suppression of p27Kip1 was abolished in Skp2-null Eμ-Myc B cells. However, the impact of Skp2 loss on Myc-driven proliferation and lymphomagenesis was surprisingly modest compared to the effects of Cks1 loss. Collectively these findings suggest that Cks1 targets in addition to p27Kip1 are critical for Myc-driven proliferation and tumorigenesis.
The cyclin-dependent kinase (Cdk) inhibitor p27Kip1 binds to and inactivates cyclin-Cdk complexes to restrict the traverse of cells through the G1 and S phases of the cell cycle (1). p27Kip1 overexpression triggers cell cycle arrest in G1 (2), while loss of p27Kip1 in mice increases rates of cell proliferation (3–4). Patients with tumors having low or undetectable levels of p27Kip1 protein have a very poor outcome (5–6), yet unlike other tumor suppressors p27Kip1 is only rarely directly mutated. Further, mice heterozygous for p27Kip1 develop spontaneous tumors late in life, yet these retain and still express the normal p27Kip1 allele (7). Finally the subcellular localization of p27Kip also has prognostic significance where high cytoplasmic p27Kip1, which is driven by activated Akt, is associated with poor outcome (Liang et al., 2002).
Signals that control p27Kip1 protein levels include its phosphorylation on Threonine-187 by the cyclin E-Cdk2 complex in S phase (8–10). Threonine-187 phosphorylated p27Kip1 is targeted to the proteasome by the SCFSkp2 ubiquitin ligase complex that is comprised of Skp1, Cullin-1 (Cul1), Rbx1, Cks1 and the F-box protein Skp2 (11). Cks1 and Skp2 form the recognition element of the SCFSkp2 complex for phospho-p27Kip1 (11–14), and their binding then leads to the ubiquitylation and destruction of p27Kip1. Accordingly, elevated levels of Skp2 in human cancer correlates with low p27Kip1 levels (15), and enforced Skp2 expression in transgenic mice reduces p27Kip1 levels and induces proliferation (16). By contrast, the targeted deletion of Skp2 leads to p27Kip1 accumulation, reduced proliferation and nuclear abnormalities (17), which are also features of Cks1 loss (14).
Myc oncoproteins that are activated in cancer are members of a basic/helix-loop-helix/leucine zipper (bHLHZip) transcription factor family that coordinates cell growth, division, and metabolism, as well as differentiation, cell migration, and apoptosis (18–19). Accordingly, in normal cells Myc levels are tightly regulated, and this occurs at the levels of transcription and the turnover of its mRNA and protein, as well as at the level of translation (20–21). Myc genes are overexpressed in ~70% of all rapidly dividing tumors, by virtue of chromosomal amplifications or translocations, or through mutations in pathways that normally restrict Myc expression (22). Expression of Myc at levels found in cancer cells is sufficient to drive primary quiescent cells into S phase (23), to accelerates rates of cell proliferation (24), and to prevent withdrawal from the cell cycle (25–26). However, these aberrant proliferative responses are harnessed by apoptotic checkpoints that are induced by Myc, including the Arf-p53 tumor suppressor pathway (27–28) and the Bcl2 family of apoptotic regulators (28). Accordingly, mutations that inactivate these apoptotic checkpoints are found in most tumors induced by Myc (28–29).
Myc accelerates the rates of cell proliferation, at least in part, through its ability to down-regulate the expression of the Cdk inhibitor p27Kip1 (30–31) which inactivates cyclin E-Cdk2 and cyclin A-Cdk2 complexes that are orchestrate entry and progression through S phase (1, 6, 32). Myc suppresses p27Kip1 expression at the level of transcription (33), but its effects on p27Kip1 protein levels in vivo are more profound (31). First, Myc induces the expression of E2f1 (34), which then promotes cyclin E transcription (35), thus activating cyclin E-Cdk2 complexes (30). Moreover, Myc induces the transcription of both cyclin D2 and its catalytic partner Cdk4, and this holoenzyme sequesters p27Kip1, thus relieving inhibition of cyclin E-Cdk2 complexes (36–37). Under either scenario activated cyclin E-Cdk2 complexes then phosphorylates p27Kip1 on Thr-187, allowing its recognition by the SCFSkp2 complex, ubiquitination and degradation by the proteasome (8–9, 38). Finally, Myc induces the expression of some of the components of the SCFSkp2 complex, including Cul1 (39) and Cks1, and at least the latter is required for down-regulation of p27Kip1 (31).
p27Kip1 is a key regulator of Myc-induced proliferation and tumorigenesis. First, loss of p27Kip1 accelerates lymphoma development in Eμ-Myc transgenic mice (40), a mouse model of human Burkitt lymphoma (41). Further, loss or heterozygosity of E2f1, or loss of Cks1, effectively cancels Myc’s ability to suppress p27Kip1 protein (but not p27Kip1 mRNA) levels, impairs Myc-induced proliferation, and markedly delays lymphoma onset and triples the lifespan of Eμ-Myc mice (31, 34). Here we report that Myc also induces the expression of the Skp2 F-box component of the SCFSkp2 complex in B cells and fibroblasts, and that Skp2 is expressed at high levels in Myc-driven lymphomas of mice and man. As expected, Skp2 loss abolishes the suppression of p27Kip1 protein in Eμ-Myc B cells. However, quite surprisingly, the effects of the Skp2 deficiency on Myc-induced proliferation and tumorigenesis are at most modest, suggesting that Cks1 has targets in addition to p27Kip1 that contribute to lymphomagenesis.
Myc suppresses p27Kip1 expression primarily by provoking ubiquitin-mediated destruction of p27Kip1 protein (8). Mechanistically this occurs through Myc-mediated induction of upstream activators of the cyclin E-Cdk2-to- p27Kip1 pathway such as E2f1 (Baudino et al., 2003) and of the Cul1 and Cks1 components of the SCFSkp2 complex that directs p27Kip1 degradation (31, 39). Specifically, Cks1 levels are markedly elevated in the pre-malignant B cells of Eμ-Myc transgenic mice, whereas the expression of the Skp1, Rbx and Cul1 components of the SCFSkp2 complex are similar to those expressed in B cells from wild type littermates (31). However, the expression of the Skp2 F-box protein that binds to phospho-p27Kip1 is also elevated in pre-cancerous Eμ-Myc B cells (Fig. 1A, B). Furthermore, Skp2 levels are also markedly elevated in lymphomas that arise in Eμ-Myc mice (Fig. 1C and (31)) and in human Burkitt lymphoma (BL), where 12 of 14 BL samples analyzed expressed elevated levels of SKP2 mRNA and protein compared to control B cells (Fig. 1D, E). Increased levels of Skp2 were not due to Skp2 amplification as previously noted in lung cancers (42), as assessed by Southern blot analyses of Eμ-Myc lymphomas (data not shown). Thus, Skp2 expression is augmented by Myc in vivo, and high levels of Skp2 are a hallmark of Myc-driven lymphoma.
To determine if Skp2 is also induced by Myc in other cell contexts, wild type mouse embryo fibroblasts (MEFs) were infected with an MSCV-based retrovirus encoding the conditional Myc-ERTAM transgene and the puromycin-resistance (PuroR) gene (43). As a control MEFs were infected with a retrovirus only expressing the PuroR gene. Puromycin-resistant cells were expanded in culture and were then treated with the estrogen receptor (ER) agonist 4-hydroxytamoxifen (4-HT), which selectively activates Myc-ERTAM. As expected, Myc activation led to the induction of the direct Myc target gene Ornithine decarboxylase (Odc) (44) and also led to the induction of Skp2, although the magnitude of the Skp2 response was not as robust (Fig. 2A).
Myc activates the majority of its transcription targets by binding, in conjunction with its requisite dimerization partner Max, to E-box elements harboring CACGTG or CACATG recognition elements (45–46). The mouse (and human) Skp2 promoter-regulatory regions lack such sites, suggesting that Myc might regulate Skp2 expression in an indirect fashion. To address this issue, Myc-ERTAM-expressing MEFs were pre-treated (for 30 min) with cycloheximide (Chx) to block de novo protein synthesis. Activation of Myc-ERTAM failed to induce Skp2 mRNA in the presence of Chx, whereas p27Kip1 transcripts were still suppressed (Fig. 2B). To assess the possibility that elevated Myc levels affect the half-life of Skp2 transcript or protein, we analyzed Myc-expressing early passage MEFs that were treated with Actinomycin D which blocks RNA synthesis or Chx to block de novo protein synthesis. No increase in RNA half-life was detected (Supplementary Fig. S1A), while Skp2 protein half-life was significantly prolonged upon ectopic Myc expression (Fig. 2C). Therefore, the regulation of Skp2 by Myc is indirect and involves transcriptional as well as post-translational mechanisms.
E2f1 is necessary for Myc to suppress p27Kip1 protein levels and E2f1 is induced by Myc (34). Skp2 and E2f1 are both elevated in Ras-induced lymphomas (47) and Skp2 has been identified as an E2f1 transcription target (48). Indeed, in wild type MEFs infected with a retrovirus encoding ER-E2f1, a conditionally activate-able ER fusion of E2f1 (49), treatment with 4-HT induced Skp2 transcripts as well as the well-characterized E2f1 target genes Thymidine kinase (Tk) and Dihydrofolate reductase (Dhfr) (Supplementary Fig. S1B; (35)). Furthermore, Skp2 promoter activity was significantly induced following co-transfection of an E2f1 expression plasmid (Supplementary Fig. S1C). Thus, we predicted that Myc would induce Skp2 via the agency of E2f1 and tested this hypothesis by evaluating the expression of Skp2 in the pre-cancerous B220+ B cells of Eμ-Myc;E2f1+/+ versus Eμ-Myc;E2f1−/− littermates. As expected, E2f1 transcripts were elevated in pre-cancerous Eμ-Myc B cells (34), and Skp2 mRNA levels were elevated 3–4-fold in Eμ-Myc B cells compared to levels expressed in the B cells of non-transgenic littermates. Surprisingly, similarly increased levels of Skp2 transcripts were evident in Eμ-Myc;E2f1−/− B cells (Fig. 3A); thus, the induction of Skp2 expression by Myc, at least in this cell context, is E2f1-independent. Furthermore, the lymphomas that arose in Eμ-Myc;E2f1−/− mice actually expressed somewhat higher levels of Skp2 protein than those expressed in Eμ-Myc;E2f1+/+ lymphomas (data not shown). Finally, Skp2 protein levels were similarly elevated in E2f1+/+ versus E2f1−/− MEFs transduced with MSCV-Myc-IRES-GFP retrovirus (Fig. 3B). Therefore, the induction of Skp2 by Myc is E2f1-independent.
Cks1 loss triples the lifespan of Eμ-Myc transgenics (Keller et al., 2007). The markedly increased levels of Skp2 in Eμ-Myc transgenic B cells and Myc-driven lymphomas (Fig. 1) suggested that Skp2 might also play critical roles in Myc-induced tumorigenesis. To test this hypothesis, we initially co-expressed Skp2 or Cks1 with Myc in immortalized BALB/c-3T3 fibroblasts. However, there were no appreciable effects of enforced expression of either Skp2 or Cks1 on Myc-induced colony formation in soft agar (Supplementary Fig. S2A–C).
To directly assess the role of Skp2 in Myc-induced tumorigenesis, Eμ-Myc transgenics (C57Bl/6) were mated to Skp2−/− mice (17) and Eμ-Myc;Skp2+/− F1 offspring were bred to Skp2+/− mice to obtain the desired Eμ-Myc;Skp2+/+, Eμ-Myc;Skp2+/− and Eμ-Myc;Skp2−/− cohort. These littermates were followed for lymphoma onset and 4 week-old mice were assessed for hallmarks of the pre-cancerous phase of the disease, including lymphocytosis and splenomegaly. White blood cell (WBC) numbers and spleen weights of Skp2−/− mice were similar to those of Skp2+/+ littermates (data not shown). As expected, Eμ-Myc;Skp2+/+ mice showed elevated numbers of WBC and lymphocytes, as well as obvious splenomegaly. Notably, there were moderate reductions in total WBC numbers in Skp2-null Eμ-Myc transgenics (Eμ-Myc;Skp2−/−, 8.0±3.2 ×103/μl vs. Eμ-Myc;Skp2+/+, 11.6±2.6 ×103/μl, Fig. 4A, left panel) and there were corresponding reductions in lymphocyte numbers (Eμ-Myc;Skp2−/−, 4.3±1.4 ×103/μl vs. Eμ-Myc;Skp2+/+, 7.6±1.0 ×103/μl, Fig. 4A, middle panel, p<0.05). Finally, the spleens of Eμ-Myc;Skp2−/− mice were smaller than those of Eμ-Myc;Skp2+/+ littermates (spleen sizes 174±29 mg vs. 123±24 mg for Eμ-Myc;Skp2+/+ vs. Eμ-Myc;Skp2−/− cohorts, Fig. 4A, right panel). There were essentially no effects of Skp2 heterozygosity on these parameters (data not shown). Therefore, loss of Skp2 moderately attenuates the pre-cancerous phase of disease in Eμ-Myc transgenic mice.
The pre-malignant B220+ B cells of Eμ-Myc transgenics have high proliferative indices, but this response is counterbalanced by the activation of apoptotic checkpoints in these cells (29, 34). There was no difference in the apoptotic indices of pre-cancerous B220+ Eμ-Myc;Skp2+/+ and Eμ-Myc;Skp2−/− B cells in vivo by Annexin-V+ FACS analyses (data not shown). Further, the apoptotic indices of Eμ-Myc;Skp2+/+ and Eμ-Myc;Skp2−/− B cells cultured ex vivo in medium supplemented with interleukin-7 was similar (Fig. 4B). Loss of Cks1 markedly impairs the hyper-proliferative response of Eμ-Myc B cells (31). Thus, we predicted that Skp2 loss would similarly affect Myc’s proliferative response. Indeed, Skp2-deficient Eμ-Myc B cells had significantly slower growth indices than B cells derived from the bone marrow of their wild type transgenic littermates when cultured ex vivo (Fig. 4C). However, these differences were not manifest in vivo, where the proliferative indices of Eμ-Myc;Skp2+/+ and Eμ-Myc;Skp2−/− B220+ B cells were similar (Fig. 4D). Therefore, unlike Cks1 (31), Skp2 does not contribute to Myc’s proliferative response in B cells in vivo.
Eμ-Myc transgenic mice succumb to aggressive, disseminating pre-B/immature B cell lymphoma, generally within 4 months of age (41). Quite remarkably Cks1 loss nearly triples the lifespan of Eμ-Myc mice (31). Non-transgenic littermates lacking Skp2 showed no signs of tumor development throughout their lifespan. Surprisingly, Eμ-Myc;Skp2−/− transgenic mice showed an only moderately delayed course of lymphoma development, with a median survival of 143 days compared to 97 days median survival of their Eμ-Myc;Skp2+/+ littermates (Fig. 5, p=0.405, not significant). There was no effect of Skp2 heterozygozity on survival (95 days median survival, Fig. 5). The lymphomas that arose in Skp2-null Eμ-Myc transgenics were phenotypically identical (pre-B and immature B cell lymphomas) to those that arose in wild type Eμ-Myc transgenic littermates (data not shown). Thus, in sharp contrast to Cks1 loss, Skp2 loss has very moderate, statistically non-significant effects on Myc-driven lymphomagenesis.
Given Myc’s ability to induce Skp2 expression whilst repressing p27Kip1 protein levels (Fig. 1), and the well-established role of the SCFSkp2 complex in directing p27Kip1 degradation (12, 17), we evaluated p27Kip1 RNA and protein levels in splenic B220+ B cells from pre-cancerous Eμ-Myc;Skp2+/+ and Eμ-Myc;Skp2−/− littermates. Notably, Myc’s ability to suppress p27Kip1 protein levels was essentially cancelled in Skp2-deficient Eμ-Myc B cells, both in vivo (Fig. 6A) and ex vivo (Fig. 6B). By contrast, p27Kip1 mRNA was suppressed in all Eμ-Myc B cells, regardless of their Skp2 status (Fig. 6C). Thus, like Cks1 (31), Skp2 is specifically required for Myc-mediated down-regulation of p27Kip1 protein levels. Finally, unlike lymphomas that arose in Eμ-Myc;Skp2+/+ littermates, nearly all Eμ-Myc;Skp2−/− lymphomas maintained high levels of p27Kip1 protein expression (data not shown).
Skp2 has been suggested to regulate Myc ubiquitination and stability and to function as an essential co-activator of Myc-mediated transactivation (50–51). If Skp2 were to play essentials role in regulating Myc protein levels, a prediction was that Myc protein levels would be elevated in Skp2-deficient Eμ-Myc transgenic B cells. This was clearly not the case, as Skp2 loss had essentially no effect on the steady state levels of Myc protein in Eμ-Myc transgenic B cells, either in vivo or ex vivo (Fig. 6A, B). We also addressed if Skp2 affected Myc’s transcriptional activity analyzing the expression of the established Myc target genes Cad (52) and Rcl (53) in pre-cancerous B220+ B cells from Eμ-Myc;Skp2+/+ versus Eμ-Myc;Skp2−/− littermates. There were essentially no changes in the levels of Rcl transcript in this cohort, and, if anything, the levels of Cad mRNA were elevated by Skp2 heterozygosity or loss (Supplementary Fig. S3). Therefore, at least the induction of these two bona fide transcription targets of Myc is independent of Skp2.
Myc promotes cell cycle entry and accelerates the rates of proliferation by suppressing the levels of p27Kip1, a key cell cycle inhibitor (Baudino et al., 2003; Keller et al., 2007; Martins and Berns, 2002). The SCFSkp2 allosteric regulator Cks1 is a target induced by Myc that clearly plays major roles in Myc’s proliferative response and in Myc-driven tumorigenesis in the Eμ-Myc transgenic mouse model of human B cell lymphoma. Further, Cks1 overexpression is a hallmark of Myc-driven lymphomas in mouse and man, and is absolutely required for Myc to suppress p27Kip1 protein levels in vivo (31). Here, we report Skp2 as yet another component of the SCFSkp2 ubiquitin ligase complex that is regulated by Myc, and in other cell contexts the SCFSkp2 Cul1 scaffold protein is induced by Myc (39). Thus, Myc orchestrates the ubiquitin-mediated degradation of p27Kip1 by affecting highly specific (Skp2 and Cks1, (13–14, 17, 31)) and rather ubiquitously expressed components (Cul1, (39)) of the SCFSkp2 ubiquitin ligase.
Skp2 was revealed, as predicted, to be essential for Myc-mediated suppression of p27Kip1 levels in Eμ-Myc B cells. These findings, along with those showing that p27Kip1 deficiency accelerates lymphoma onset in Eμ-Myc transgenics (Martins and Berns, 2002) and that Cks1 loss impairs Myc-induced proliferation and lymphomagenesis (Keller et al., 2007), strongly suggested that Skp2 loss would cancel Myc’s proliferative response and thus impair Myc-induced lymphoma development. Surprisingly, this was not the case, where despite fully restoring p27Kip1 levels the effects of Skp2 loss on Myc-induced proliferation and lymphoma development were at most moderate compared to those manifest in the Cks1 deficiency (31). These findings suggest that at least in this context Cks1 has functions other than that as a regulator of p27Kip1 that also contribute to Myc-induced proliferation and tumorigenesis. Indeed, Cks1 has functions as a regulator of transcription in yeast (54–55) and human cancer cells (56) and appears to have Skp2/p27Kip1-independent functions in controlling human multiple myeloma cell growth and survival (57).
The importance of Skp2 in cancer has be documented, where Skp2 expression is highly elevated in a number of malignancies, and where this is associated with reduced p27Kip1 levels, high proliferative rates, and poor outcome (15, 58). Furthermore, Skp2 cooperates with oncogenic N-Ras in promoting anchorage-independent growth of rodent fibroblasts in vitro, and in promoting lymphomagenesis in vivo (47). In contrast, Skp2 does not augment Myc-induced soft agar growth of fibroblasts (Supplementary Fig. S2) and Skp2 loss has no significant effects on Myc-driven lymphomagenesis, despite canceling Myc’s ability to suppress p27Kip1 protein levels. We conclude that there are context specific effects of Skp2 in tumorigenesis.
Myc oncoproteins are short-lived, and Myc turnover occurs through the ubiquitin-proteasome pathway (59). Skp2 has been suggested to bind to c-Myc, to promote its ubiquitination and degradation, and to also augment its transactivation functions (50–51). If this scenario were operational in B cells then Skp2 loss should have at least led to increased levels of Myc protein in Eμ-Myc B cells and thus perhaps to accelerated disease, as homozygous Eμ-Myc transgenics develop more rapid lymphomas than hemizygous Eμ-Myc littermates (60). Neither of these responses was, however, evident in Eμ-Myc;Skp2−/− mice, and the expression of at least some established Myc targets was also unaffected in Skp2-deficient Eμ-Myc B cells. Our findings are thus more in accord with those of others indicating that the F-box proteins Fbw7 (20, 61) or HectH9 (62) regulate Myc turnover.
In normal cells, Myc’s ability to accelerate proliferation is harnessed by the activation of apoptotic pathways and disabling this response, by loss-of-function mutations in the Arf-p53 tumor suppressor pathway (29), dramatically accelerates the course of Myc-induced malignancies. The effects of biallelic loss of p27Kip1 on lymphoma onset in Eμ-Myc mice are less dramatic (40), yet the Skp2/Cks1-p27Kip1 pathway is affected in all Myc-driven lymphomas (this report and (31)). Myc-mediated induction of Skp2 is indirect, suggesting that Myc may work through the agency of other transcription factors to induce Skp2. One candidate was E2f1, as Myc induces E2f1 and since E2f1 is required for Myc-mediated repression of p27Kip1 in Eμ-Myc B cells (34). Further, in immortal fibroblasts and some tumor cell lines E2f1 promotes p27Kip1 degradation via its induction of Skp2 (48), and in pancreatic cancer, the malignant phenotype is associated with E2f1-dependent induction of Skp2 (63). The finding that Myc-induced expression of Skp2 is independent of E2f1 was thus surprising, and a role for FoxM1, another activator of Skp2 transcription (64), seems also unlikely, as FoxM1 expression is reduced in Eμ-Myc B cells and lymphomas (31). Thus, other transcriptional regulators downstream of Myc must control Skp2 expression. The complexity of Skp2 regulation is further documented by an increase in Skp2 protein half-life that points to post-translational effects of Myc. The fact that the effects of Skp2 loss on Myc-induced lymphomagenesis are moderate at best however strongly point towards Cks1 roles besides SCFSkp2 and p27Kip1 control.
Skp2 overexpression in cancer has heretofore been linked to SKP2 gene amplification, E2f1 and FoxM1. Our findings strongly suggest that Myc regulates Skp2 expression to control p27Kip1 levels. The impact of Skp2 loss on Myc-driven proliferation and lymphomagenesis was surprisingly modest. Given differences in the magnitude of the effects of Cks1 versus Skp2 loss on Myc-mediated tumorigenesis suggest other Cks1 targets (e.g., not linked to SCFSkp2) that can be exploited in cancer therapeutics.
Skp2 null mice (C57Bl/6) (17) were bred with Eμ-Myc transgenic mice (C57Bl/6) (41). F1 Eμ-Myc;Skp2+/− offspring were bred to Skp2+/− mice obtain Eμ-Myc;Skp2+/+, Eμ-Myc;Skp2+/−, and Eμ-Myc;Skp2−/− littermates. Animals were observed for signs of morbidity and tumor development. Tumors were harvested after sacrifice of mice, snap-frozen in liquid nitrogen, and processed for analysis of DNA, RNA and protein. E2f1 null mice (65), again on a C57BL/6 background) were bred with Eμ-Myc transgenic mice. F1 Eμ-Myc;E2f1+/− offspring were bred to E2f1+/− obtain Eμ-Myc;E2f1+/+, Eμ-Myc;E2f1+/− and Eμ-Myc;E2f1−/− littermates.
With institutional review board approval, and following informed consent, tumors from 14 Burkitt lymphoma patients were banked. RNA and protein were extracted from these tumors. As a control, pooled peripheral blood mononuclear cells from healthy donors were enriched using CD19-MicroBeads according to the manufacturer’s instruction (Miltenyi Biotech) and RNA and protein was prepared.
Primary bone marrow-derived pre-B cells were cultured as described previously (29). MEFs from E13.5-E14.5 embryos were cultured and infected with MSCV-Myc-ER™-IRES-GFP, MSCV-Myc-IRES-GFP, pBabe-Myc-ER™-IRES-Puromycin, pBabe-ER™-E2f1-Puromycin or control retrovirus as described (27). To evaluate consequences of Myc activation cells were treated with 2-μM 4-hydroxytamoxifen (4-HT) and harvested for protein and RNA preparation. To assess whether Myc induction of Skp2 was direct, Myc-ER-expressing cells or control cells were pre-treated with 1μg/ml Cycloheximide (Chx, Sigma Chemicals) for 30 min (which inhibited >95% of protein synthesis), prior to adding 4-HT. For analysis of Skp2 RNA half-life MEFs were cultured in the presence of 1μg/ml Actinomycin D (ActD, Sigma-Aldrich) and harvested at the indicated time. To estimate Skp2 protein half-life MEFs were cultured in the presence of 10μg/ml Chx and harvested at the indicated time.
Rates of proliferation or apoptosis of B cells were determined using a Flow Kit as described by the manufacturer (BD Biosciences Pharmingen). Bone marrow and spleen cells were incubated with B220 MicroBeads and enriched by magnetic cell sorting for B cells according to the manufacturer’s instructions (Miltenyi Biotech) and used for immunoblot or real-time PCR analysis.
RNA was prepared from cultured MEFs, MACS-sorted B cells, or lymphomas using the RNeasy kit (Qiagen). For real-time PCR, cDNA was prepared from 1 μg RNA using the iScript cDNA synthesis kit (Bio-Rad). Real-time PCR was performed using an iCycler machine (Bio-Rad) and the iTaq SYBR green kit (Bio-Rad). Data analyses were performed by comparing Ct values with a control sample set as 1. Sequences for primers are available upon request.
Protein extracts (20 or 50 μg per lane) were separated electrophoretically on a SDS-PAGE gel, transferred to membranes (Protran, Schleicher & Schuell) and blotted with antibodies specific for Skp2 (Zymed Laboratories, Inc.), p27Kip1 (BD Biosciences Transduction Laboratories), c-Myc and E2f1 (Santa Cruz, Inc.) and β-Actin (Sigma Chemicals).
The statistical analysis of survival differences in Eμ-Myc transgenics of Skp2−/− vs. Skp2+/+ genotype was performed using a Cox-Regression analysis with a multiple cohort comparison Bonferroni- adjusted. The statistics performed to analyze differences in the ex vivo and in vivo B cell proliferation and apoptosis indices involved paired t-tests.
SUPPLEMENTAL FIGURE S1. E2f1 induces Skp2. A. Primary early passage MEFs were infected with MSCV-IRES-Puro (Puro) or MSCV-Myc-IRES-Puro (Myc) retroviruses and Puromycin-selected. The left panel shows increased Skp2 RNA levels in Myc over-expressing MEFs (relative expression as compared to Rps21 levels; Rps21 is not significantly regulated by Myc) assessed by quantitative real-time PCR. The right panel shows Skp2 transcript levels in MEFs treated with Actinomycin D (ActD, 1μg/ml) for the indicated time. The levels of the Skp2 transcripts were normalized to Rps21 expression and set to 100 percent (0h). Shown is the percent expression after the indicated time of ActD treatment. B. Primary early passage MEFs were infected with pBabe-Puro (Puro) or pBabe-ER-E2f1-Puro (ER-E2f1) retroviruses. Puromycin-resistant cells were treated for the indicated times with 4-HT to activate ER-E2f1 and assessed for changes in the expression of Skp2, and the E2f1 targets thymidine kinase (Tk) and dihydrofolate reductase (Dhfr), by real-time PCR. Levels of RNA were standardized to the expression of Ubiquitin (Ub). C. NIH-3T3 cells were co-transfected with a Skp2 promoter-reporter construct (firefly luciferase, (63)) and either E2f1, c-Myc or control (ctrl) expression plasmids. Luciferase activity was determined according to the manufacturer’s instruction (Promega, Madison, WI). Relative luciferase activity was determined by calculating the ratio of firefly to co-transfected renilla luciferase activity. Note that E2f1, but not c-Myc, activated the Skp2 promoter.
SUPPLEMENTAL FIGURE S2. Skp2 or Cks1 do not cooperate with Myc in the transformation of immortal fibroblasts. A. Immunoblot analysis of the indicated proteins in BALB/c-3T3 fibroblasts transfected with pBabe-Puromycin (Puro) or pBabe-Myc-Puromycin (Myc-Puro) expression plasmids +/− Cks1 or Skp2 expression (which were expressed in MSCV-IRES-GFP expression plasmids; ctrl: MSCV-IRES-GFP). Transfected cells were selected for growth in puromycin-containing medium (6μg/ml) for 48 hr and were sorted for GFP by FACS. B. Soft-agar colony formation. 500 cells were plated per well in 6-well plates and grown for 14 days. Colonies were then stained with MTT. C. Quantification of colony formation. Shown is a representative experiment performed in duplicate.
SUPPLEMENTAL FIGURE S3. Analysis of Myc target gene expression in non-transgenic (wt) and Eμ-Myc B cells of the indicated Skp2 genotypes. SYBR-green real-time PCR analysis of c-myc, Cad, and Rcl mRNA levels was performed on splenic B220+ B cells from 4 week-old pre-cancerous Eμ-Myc;Skp2+/+, Eμ-Myc;Skp2+/−, and Eμ-Myc;Skp2−/− mice and compared to their expression in B cells from a wild type (wt) littermate. Levels of mRNAs were standardized to the expression of Ubiquitin (Ub).
We thank Sara Norton, Chunying Yang and Elsie White for expert technical assistance, the Animal Resource Center, the Hartwell Center and the FACS Core Facility of SJCRH for animal care and technical support, and Tibor Schuster (Institute for Medical Statistics and Epidemiology, TU München, Munich, Germany) for statistical analyses. We also thank Kristian Helin (Copenhagen, Denmark) for providing ER-E2f1 plasmid, Roland M. Schmid (Munich, Germany) for providing the CMV-E2f1 plasmid and the Skp2-promoter reporter construct, and Michael Deininger (Portland, OR) for providing the Skp2 cDNA. We also are indebted to Drs. Mihaela Onciu and John Sandlund (Memphis, TN) for providing Burkitt lymphoma samples.
Grant Support: This work was supported by NIH grant CA76379 (JLC), by Cancer Center Core Grant CA21765, by the American Lebanese Syrian Associated Charities (ALSAC) of St. Jude Children’s Research Hospital (SJCRH), and by monies from the State of Florida to Scripps Florida. UBK was supported by the Deutsche Forschungsgemeinschaft (SFB TRR54). JBO was supported by NRSA grant F32 CA099478.
Disclosure of Potential Conflict of Interest
The authors indicate no potential conflict of interest.