Mitotic arrest deficient 1 (Mad1) plays a well-characterized role in the mitotic checkpoint. However, interphase roles of Mad1 that do not impact mitotic checkpoint function remain largely uncharacterized. Here we show that upregulation of Mad1, which is common in human breast cancer, prevents stress-induced stabilization of the tumor suppressor p53 in multiple cell types. Upregulated Mad1 localizes to ProMyelocytic Leukemia (PML) nuclear bodies in breast cancer and cultured cells. The C-terminus of Mad1 directly interacts with PML, and this interaction is enhanced by sumoylation. PML stabilizes p53 by sequestering MDM2, an E3 ubiquitin ligase that targets p53 for degradation, to the nucleolus. Upregulated Mad1 displaces MDM2 from PML, freeing it to ubiquitinate p53. Upregulation of Mad1 accelerates growth of orthotopic mammary tumors, which show decreased levels of p53 and its downstream effector p21. These results demonstrate an unexpected interphase role for Mad1 in tumor promotion via p53 destabilization.
Mad1 was initially discovered in a landmark screen demonstrating that mitosis is regulated by a cell cycle checkpoint, termed the mitotic (or spindle assembly) checkpoint1. The mitotic checkpoint ensures accurate chromosome segregation by delaying separation of the replicated sister chromatids until each sister chromatid pair is stably attached to opposite spindle poles through its kinetochores2,3,4,5,6. Mad1 plays an evolutionarily conserved role in the mitotic checkpoint by recruiting its binding partner Mad2 to the kinetochores of unattached chromatids7,8,9. At unattached kinetochores, Mad2 is converted into an active mitotic checkpoint inhibitor that delays sister chromatid separation10,11,12,13. Once the kinetochores of all sister chromatids are stably attached to spindle microtubules, the mitotic checkpoint is satisfied, and Mad1 and Mad2 are no longer recruited. Loss of Mad1 is lethal, and cells with reduced expression of Mad1 missegregate chromosomes to become aneuploid1,14. Thus, Mad1 is essential and plays a highly conserved role in ensuring accurate chromosome segregation during mitosis.
Although Mad1 plays a well-characterized role during mitosis, and expression of many mitotic proteins peaks during mitosis, Mad1 expression levels remain constant throughout the cell cycle2. During interphase, Mad1 recruits Mad2 to nuclear pores at the nuclear envelope, which permits the production of mitotic checkpoint inhibitors during interphase3,15,16,17. Interphase functions of Mad1 that do not affect mitotic checkpoint signaling have remained largely uncharacterized, although it is known that Mad1 functions independently of Mad2 at the Golgi apparatus to promote secretion of α5 integrin18,19. Mad1 is frequently upregulated at both the mRNA and protein level in human breast cancers, where Mad1 upregulation serves as a marker of poor prognosis2,20,21. Mad1 upregulation causes a low rate of chromosome missegregation, which is weakly tumor promoting2,22,23,24. However, whether Mad1 upregulation has additional tumor-promoting activities during interphase has remained unclear.
Upregulated Mad1 localizes to nuclear pores and kinetochores, as expected, but also forms punctate structures2,16. A fraction of these colocalize with markers of annulate lamellae, storage compartments for excess nuclear pore components, which are predominantly cytoplasmic2,16,25. Nuclear Mad1 puncta have remained uncharacterized. Promyelocytic leukemia (PML) nuclear bodies (NBs) represent one prominent source of nuclear puncta. The PML protein, which is fused to retinoic acid receptor alpha (RARα) due to a reciprocal translocation between chromosomes 15 and 17 in >98% of acute PML patients, forms the core of PML NBs26. >100 proteins localize to PML NBs, including proteins involved in cell cycle arrest, apoptosis, transcription, and metabolism27. Though the proteins that localize to PML NBs are functionally diverse, most of these proteins, including PML itself, are sumoylated26,27. Here, we show that upregulated Mad1 localizes to PML NBs.
Protein levels of the p53 tumor suppressor remain low in the absence of cellular stresses due to continuous ubiquitination by MDM2 followed by degradation28,29,30. In response to a variety of cellular stresses including DNA damage, PML sequesters MDM2 in the nucleolus, which physically separates MDM2 from p53 and results in p53 stabilization31,32,33,34. Here, we demonstrate a previously unexpected interphase role for Mad1 in preventing p53 stabilization. The C-terminal domain (CTD) of Mad1 binds PML directly in a manner facilitated by sumoylation of PML. Upregulated Mad1 localizes to PML NBs, and localization is dependent on the SUMO interacting motif (SIM) within the Mad1 CTD. After DNA damage, upregulated Mad1 displaces MDM2 from PML, replaces MDM2 at nucleoli, and increases the interaction of MDM2 with p53. Mad1-YFP promotes orthotopic mammary tumors in a SIM-dependent manner. These data provide molecular insight into a novel interphase role of Mad1 in destabilizing p53 and promoting tumor initiation and growth.
Mad1 accumulates into PML NBs
Upregulated Mad1 localizes to kinetochores and the nuclear envelope, as expected, but also forms nuclear puncta2,16. This fraction of Mad1 does not colocalize with nucleoli (Supplementary Fig. 1a–b), but does show substantial colocalization with Myc and HA tagged SUMO1 and SUMO2 as well as endogenous SUMO1 in MDA-MB-231 breast cancer cells and in HeLa cervical cancer cells (Fig. 1a–c, f, Supplementary Fig. 1b–e). SUMO1 and SUMO2 are highly concentrated in PML NBs34,35,36, suggesting that Mad1 puncta localize there. Indeed, Mad1 nuclear puncta substantially colocalize with endogenous as well as HA-tagged PML during both interphase and mitosis (Fig. 1d–f, Supplementary Fig. 1f–g). Under stressful culture conditions, a small portion of endogenous Mad1 also colocalizes with PML (Supplementary Fig. 1h–j).
Arsenic trioxide is therapeutic in acute PML, since it induces proteasomal degradation of PML-RAR, as well as endogenous PML36,37,38. Arsenic treatment was used to confirm that upregulated Mad1 localizes to PML NBs. 12 and 24 h after arsenic treatment, PML protein levels were decreased and PML NBs substantially dispersed (Fig. 1g, h). Substantially fewer Mad1 puncta formed 12 and 24 h after arsenic treatment, although Mad1 protein levels remained constant (Fig. 1g, h, Supplementary Fig. 1k). These data demonstrate that the formation of Mad1 nuclear puncta is dependent upon PML NBs.
Mad1 recruits Mad2 to unattached kinetochores and nuclear pores3,16,39, and Mad1 interacts with Mad2 throughout the cell cycle3,40. Mad2 is expressed in molar excess of Mad1, and the only known pool of Mad1 that does not recruit Mad2 is at the Golgi apparatus41. To determine whether Mad1 recruits Mad2 to PML NBs, Mad1-3xFLAG and Mad2-GFP were expressed singly or in combination. Mad1-3xFLAG recruited Mad2-GFP to nuclear puncta, which Mad2-GFP did not form when expressed alone (Supplementary Fig. 1l). Thus, Mad1 recruits Mad2 to PML NBs, as it does to unattached kinetochores and nuclear pores.
We have previously shown that Mad1 protein levels are commonly increased in breast cancer patients, and that upregulated Mad1 forms punctate structures in breast cancer2. To determine if Mad1 puncta represent PML NBs, sections from nine primary breast cancers were costained with antibodies to Mad1 and SUMO1. Indeed Mad1 puncta frequently colocalized with SUMO1 (Fig. 1i, j). Together, these results demonstrate that upregulated Mad1 localizes to PML NBs in cultured cells and in human cancer.
Mad1 interacts with PML through its C terminal domain (CTD)
As an initial test of whether Mad1 localization to PML NBs was due to an interaction with PML or with another component of PML NBs, Mad1-3xFLAG was immunoprecipitated from HEK293T cells. HA-PML was co-immunoprecipitated with Mad1-3xFLAG, supporting an interaction between these proteins (Supplementary Fig. 1m). To determine the region of Mad1 responsible for the interaction with PML, 10 deletion mutants of Mad1 were generated (Fig. 2a, Supplementary Fig. 2a). Immunoprecipitation experiments revealed that all constructs containing the CTD of Mad1 (aa 597–718) co-precipitated PML, while fragments lacking the Mad1 CTD did not (Fig. 2b). A reciprocal immunoprecipitation experiment showed that immunoprecipitation of HA-PML co-precipitated all Mad1 fragments containing the CTD, but no fragments lacking the Mad1 CTD (Fig. 2c). Thus, Mad1 interacts with PML in cell extracts through its CTD.
We then tested the subcellular localization of the Mad1 fragments (Fig. 2d–f, Supplementary Fig. 2b–d). While full length Mad1 localized to PML NBs, a fragment of Mad1 lacking the CTD (aa 1–596) was excluded from PML NBs (Fig. 2d, e), consistent with the requirement of the Mad1 CTD for the interaction of Mad1 with PML. A fragment of Mad1 that contains the CTD and co-immunoprecipitates with PML, aa 180–718, is cytoplasmic and therefore unable to localize to PML NBs, presumably because it lacks the endogenous nuclear localization sequence (NLS; Fig. 2e, Supplementary Fig. 2b). The addition of the NLS from SV40 restored PML localization of aa 180–718 (Fig. 2e), but not of fragments lacking the CTD containing aa 180–480 or aa 180–596 (Supplementary Fig. 2d). The Mad1 CTD (aa 597–718) domain is not sufficient to support localization to PML NBs, at least in part due to lack of an NLS (Fig. 2f). Unexpectedly, the addition of the SV40 NLS drove localization of the CTD to the nuclear envelope, but not to PML NBs; fusion with the endogenous nuclear pore targeting domain (NPD; aa 1–2743) was necessary to localize the Mad1 CTD to PML NBs (Fig. 2f).
To define the domain of PML that interacts with Mad1, four truncations of PML were generated (Fig. 2g). Immunoprecipitation of HA-tagged full length and aa 1–229 of PML co-precipitated Mad1-3xFLAG (Fig. 2h). Reciprocal immunoprecipitation of Mad1 confirmed that Mad1-3xFLAG interacts with aa 1–229 of PML (Fig. 2i). Together, these data demonstrate that the C-terminus of Mad1 interacts with the N-terminus of PML.
A conserved Mad1 SIM is critical for localization with PML
The N-terminal domain of PML contains two sumoylation sites (Fig. 2g)42,43. SUMO chains at these sites interact with the SIM of other proteins in PML NBs42. To determine whether the interaction of Mad1 with PML involved sumoylation, we performed bioinformatic analysis of the structure of Mad1. Using the GPS–SUMO algorithm44,45 we identified a putative SIM in the CTD of Mad1 (aa 689–699; Fig. 3a). The putative SIM in the Mad1 CTD is conserved across multiple species (Fig. 3a) and similar to the bona fide SIMs in PML and the PML interaction partner DAXX (Fig. 3b). The SIM consists of a hydrophobic core separated from an acidic stretch by a short spacer region46,47. Structural analysis and visualization of the Mad1 CTD using SWISS-MODEL Workspace (4dzo.1) and PyMOL software48, revealed a hydrophobic surface localized within the putative SIM at aa 689–692 (Supplementary Fig. 3a). This result is consistent with the presence of the predicted SIM in the CTD of Mad1.
To determine the importance of the SIM in the CTD of Mad1 for localization to PML NBs, the leucine (aa 689) and isoleucine (aa 690) were mutated into lysine residues to abolish the hydrophobic core (Fig. 3b). Mutation of LIEV to KKEV (hereafter SIM mutant) substantially altered Mad1 localization, dispersing it from PML NBs to numerous smaller puncta (Fig. 3c), suggesting that the SIM is essential for Mad1 to localize to PML NBs. Consistent with this, mutation of the SIM in the CTD of Mad1 substantially decreased the interaction of Mad1 with PML in reciprocal immunoprecipitation experiments (Fig. 3d). These experiments demonstrate that mutation of two residues within the CTD of Mad1 largely disrupts its interaction with PML.
The CTD of Mad1 interacts directly with PML
To determine whether the interaction between Mad1 and PML is direct, MBP tagged full length Mad1 or the Mad1 CTD were co-expressed with recombinant PML in bacteria (Fig. 3e, Supplementary Fig. 3b). MBP alone, MBP-Mad1 (full length), or MBP-Mad1(CTD) was isolated from cell lysates with amylose resin and binding to PML was evaluated by SDS-PAGE and western blot. This analysis revealed that full length Mad1 and the Mad1 CTD both interact directly with PML (Fig. 3f). To determine the importance of sumoylation in the interaction between Mad1 and PML, three copies of SUMO-2 were linked to the N-terminus of PML (Fig. 3e, Supplementary Fig. 3b). The addition of 3xSUMO-2 enhanced binding of both full length Mad1 and the CTD of Mad1 to PML (Fig. 3f). These data demonstrate that Mad1 binds directly to PML and that this interaction is enhanced by sumoylation.
Mad1 upregulation prevents stabilization of p53
PML plays an important role in stabilizing the p53 tumor suppressor in response to cellular stress31,34,43. Considering that Mad1 is frequently upregulated in human breast cancers2, and that upregulated Mad1 localizes to PML NBs, we tested whether Mad1 upregulation affects p53 stabilization. Tetracycline (tet)-inducible expression of Mad1 prevented the increase in p53 protein levels observed in response to DNA damage caused by the topoisomerase II inhibitor doxorubicin in control cells in multiple cancer cell types (MDA-MB-231, HeLa, DLD1; Fig. 4a, b, Supplementary Fig. 4a). These cell lines all have an impaired p53 pathway, due to an R280K mutation in p53, expression of HPV E6, or S241F mutation, respectively49,50,51. Upregulation of Mad1 also prevented stabilization of p53 in response to DNA damage in nontransformed MCF10A breast cells, which express wild type p5351,52 (Fig. 4c).
p21 is a critical mediator of p53-dependent cell cycle arrest in response to DNA damage53,54,55, and induction of p21 upon DNA damage is regulated by PML31,34,56. As an initial test of whether upregulation of Mad1 affects p53 downstream signaling, p21 expression levels were assessed. Mad1 upregulation prevented p21 accumulation in nontransformed MCF10A (Fig. 4c) as well as cancerous HeLa (Fig. 4d, Supplementary Fig. 4b) cells.
To determine if the role of Mad1 in regulating the p53 signaling pathway is dependent on its interaction with PML, Mad1 lacking the CTD or the SIM was tested. Expression of Mad1 lacking the CTD restored stabilization of p53 and p21 in response to DNA damage (Fig. 4b, Supplementary Fig. 4b). Similarly, both p53 and p21 accumulated in cells expressing the Mad1 SIM mutant, similar to control cells (Fig. 4d). These data demonstrate that the interaction with PML is necessary for upregulated Mad1 to prevent stabilization of p53.
To further evaluate the effect of Mad1 upregulation on p53 signaling, cell death in response to DNA damage was assessed. We generated cell lines stably expressing 3xFLAG-tagged full length wild type Mad1, Mad1 lacking the CTD, or SIM mutant Mad1 in a tet-inducible manner. DNA damage induced cell death in control cells (Fig. 4e, Supplementary Fig. 4c). Expression of full length Mad1 substantially reduced cell death, while expression of Mad1 lacking the CTD or SIM did not (Fig. 4e, Supplementary Fig. 4c). Thus, upregulation of Mad1 impairs p53 stabilization and downstream signaling, and the interaction of Mad1 with PML is necessary for these effects.
Importantly, upregulation of endogenous Mad1 also prevents p53 stabilization in a breast cancer cell line that expresses high levels of Mad1 protein without experimental manipulation (SKBR3; Supplementary Fig. 5a). Consistent with its localization in primary breast tumors and after induced expression, Mad1 localizes to PML NBs in SKBR3 cells (Supplementary Fig. 5b). Partial depletion of Mad1 with four different shRNA sequences increases the level of p53 in response to DNA damage (Supplementary Fig. 5c). Thus, upregulation of endogenous or exogenous Mad1 results in reduced protein levels of p53.
Mad1 displaces MDM2 from nucleoli after DNA damage
p53 levels are normally low due to ubiquitination by MDM2 followed by degradation. In response to DNA damage, PML sequesters MDM2 in the nucleolus, which physically separates MDM2 from p53 and allows p53 protein levels to accumulate31. To determine whether Mad1 destabilizes p53 by disrupting MDM2 translocation into nucleoli, we examined MDM2 localization in cells that stably express Mad1 in response to tet. After DNA damage, a substantial portion of MDM2 colocalizes with nucleolin (Fig. 5a, b). In contrast, in cells expressing full length wild type Mad1 in response to tet, MDM2 localization to nucleoli is substantially reduced (Fig. 5a, b). Conversely, while Mad1 expressed in response to tet does not colocalize with nucleolin in the absence of DNA damage, in the presence of doxorubicin a majority of cells contain nucleolar Mad1 (Fig. 5c, d). Because the SIM of Mad1 contributes to its interaction with PML (Fig. 3c–f), we predicted that SIM mutant Mad1 would be unable to displace MDM2 from nucleoli. Indeed, expression of SIM mutant Mad1 did not affect localization of MDM2 to nucleoli in the presence of DNA damage (Fig. 5e, f), nor did SIM mutant Mad1 localize to nuclei after doxorubicin treatment (Fig. 5g, h). These data demonstrate that upregulated Mad1 displaces MDM2 from nucleoli in response to DNA damage in a SIM-dependent fashion.
Mad1 regulates the interaction of MDM2 with PML and p53
In response to DNA damage, PML escorts MDM2 to nucleoli through a direct interaction, which separates MDM2 from p53 and permits p53 protein levels to accumulate31,33. To determine whether Mad1 upregulation prevents p53 stabilization by displacing MDM2 from PML, MDM2 was immunoprecipitated in the presence of increasing amounts of full length Mad1 or Mad1 lacking the CTD. Expression of full length Mad1 decreased the co-immunoprecipitation of PML with MDM2 in a dose-dependent fashion (Fig. 6a). Importantly, the expression of Mad1 lacking the CTD to a similar level as full length Mad1 did not affect the ability of MDM2 to co-immunoprecipitate PML (Fig. 6a). Reciprocal experiments, in which PML was immunoprecipitated, showed that co-immunoprecipitation of MDM2 with PML was substantially impaired by full length Mad1 but not by Mad1 lacking the CTD (Fig. 6b). While full length Mad1 was co-immunoprecipitated with PML, Mad1 lacking the CTD was not, despite similar expression levels (Fig. 6b). These results demonstrate that Mad1 binding displaces MDM2 from PML.
MDM2 immunoprecipitations were also used to determine whether the MDM2 displaced by Mad1 can interact with p53. Expression of full length Mad1, but not Mad1 lacking the CTD, impaired co-precipitation of PML with MDM2 (Fig. 6c), consistent with the previous result (Fig. 6a). Importantly, expression of full length Mad1 increased the amount of p53 co-precipitated with MDM2, while expression of Mad1 lacking the CTD did not (Fig. 6c). These results demonstrate that Mad1 binding to PML displaces MDM2 from PML, and that the displaced MDM2 binds p53.
Mad1 upregulation promotes tumors by inhibiting p53 and p21
Mad1 is commonly upregulated at both the mRNA and protein levels in primary breast cancer2. To determine whether increased expression of Mad1 is sufficient to promote tumor initiation, nontransformed MCF10A cells were engineered to stably express Mad1-mNeonGreen in response to tet. Parental cells that expressed the tet repressor but not Mad1-mNeonGreen were used as controls. Cells were injected orthotopically into mammary fat pads. All animals received chow containing the tet analog doxycycline (dox) to induce expression of Mad1-mNeonGreen or control for the effects of dox. Parental MCF10A cells formed tumors at 0 of 12 injection sites, while Mad1-mNeonGreen expressing MCF10A cells formed tumors at 9 of 12 injection sites (Fig. 7a, b). These data show that upregulation of Mad1 is sufficient for transformation and tumorigenesis in MCF10A cells.
To determine whether Mad1 upregulation promotes tumor growth and progression, MDA-MB-231 cells that express Mad1-YFP in response to tet were injected orthotopically into mammary fat pads. Parental MDA-MB-231 cells that express the tet repressor but not Mad1-YFP were used as controls. All animals received chow containing dox. As expected, parental MDA-MB-231 cells formed tumors at 6 of 6 injection sites. However, 6 of 6 MDA-MB-231 tumors expressing Mad1-YFP grew at a faster rate and to a larger size than all 6 of the parental tumors (Fig. 7c, d). In a separate experiment, MDA-MB-231 cells expressing SIM mutant Mad1 in response to tet grew at the same rate as parental cells, while all 6 tumors expressing Mad1-YFP substantially outgrew both the 6 parental and 6 SIM mutant Mad1-expressing tumors (Fig. 7e, f). Interestingly, while parental MDA-MB-231 tumors and tumors expressing SIM mutant Mad1 showed detectable levels of p53 and p21 expression, these were notably reduced in MDA-MB-231 tumors expressing Mad1-YFP (Fig. 7g, Supplementary Fig. 4d). These data demonstrate that upregulation of Mad1 destabilizes p53 and p21 in vivo and that p53 destabilization is central to the ability of Mad1 to promote orthotopic mammary tumor growth.
Mad1 plays a well-characterized role in regulating chromosome segregation during mitosis2,3,57,6. Interphase roles of Mad1 remain less well studied. Here we demonstrate that upregulated Mad1 prevents stabilization of p53 during interphase through direct binding to PML. p53 levels are normally kept low due to continuous ubiquitination by MDM2 in PML NBs (Fig. 7h, top)28,30,58. In response to DNA damage, PML sequesters MDM2 into nucleoli (Fig. 7h, middle)31,33,34,56. This spatially separates MDM2 from p53, stabilizing p53 protein and allowing it to regulate gene transcription, resulting in p21 expression, cell cycle arrest, and/or cell death. Our data support a model in which upregulated Mad1 directly binds PML through its CTD in a manner promoted by sumoylation. Mad1 binding displaces MDM2 from PML. PML then sequesters Mad1 into nucleoli (Fig. 7h, bottom). Displaced MDM2 remains in close physical proximity to p53, allowing it to bind and ubiquitinate p53, leading to continuous p53 degradation and tumor promotion. Importantly, this mechanism regulates the expression of both wild type and mutant p53. These data provide molecular insight into an unexpected interphase role of Mad1.
Mad1 is commonly upregulated in human breast cancer at both the mRNA and protein levels, where it serves as a marker of poor prognosis2,59. Mad1 upregulation enhances anchorage-independent growth in culture2 and promotes both tumor initiation and tumor growth in orthotopic models of breast cancer (Fig. 7a–f). Upregulation of Mad1 to levels similar to those found in human breast cancer causes a low rate of chromosome missegregation, which is weakly tumor promoting2,22,23,24,60. However, p53 is the most commonly altered tumor suppressor in human cancers, and destabilization of p53 substantially contributes to the tumor-promoting effects of Mad1 upregulation (Fig. 7e, f).
It is interesting that Mad1 expression regulates p53 levels not only in cells with a wild type p53 pathway (MCF10A), but also in cells in which p53 function is altered due to mutation (S241F in DLD1; R280K in MDA-MB-231) or by human papillomavirus E6 (HeLa). Our results are in agreement with previous studies showing that even mutant p53 is stabilized in response to DNA damage61,62. This suggests that Mad1 destabilization of p53 is robust and widespread, and is likely to occur in most tumor cells in which Mad1 is upregulated, independent of p53 mutation status.
Numerous mouse models with altered expression of mitotic checkpoint genes have been used as presumptive tests of the impact of aneuploidy and chromosomal instability (CIN) on tumor initiation14,63,64,65. Taken together, these studies support the overarching conclusion that low rates of CIN are weakly tumor promoting while high rates of CIN lead to cell death and tumor suppression22,23,24,60. However, the tumor phenotypes of mouse models with distinct genetic manipulations vary widely despite similar levels of aneuploidy. Like Mad1, most mitotic checkpoint genes are expressed throughout the cell cycle and have been implicated in diverse cellular functions during interphase22,66. The previously unsuspected role of Mad1 in influencing p53 stability has a substantial impact on the orthotopic tumor phenotype caused by Mad1 upregulation. Thus, the wide variety of tumor phenotypes caused by alteration of “mitotic” genes is likely to be due in large part to the known as well as the undiscovered interphase functions of these genes.
The importance of sumoylation in chromosome segregation during mitosis is emerging, with roles for sumoylation identified on several critical mitotic regulators, including CENP-E, BubR1, and APC/C46,47,67,68,69. Here we demonstrate a functional role for the Mad1 SIM during interphase. Whether sumoylation or SUMO interaction affects the interphase function of additional “mitotic” genes as well as whether sumoylation and binding affect the mitotic function of Mad1 remain important unanswered questions.
Antibodies used in this study include affinity-purified rabbit anti-Mad1 antibodies prepared against amino acids 333–617 of human Mad12 and diluted 1:3000 for western blots and 1:2000 for immunofluorescence, mouse anti-SUMO1 (21C7, Developmental Studies Hybridoma Bank) diluted 1:250, mouse anti-Myc (9E10, Developmental Studies Hybridoma Bank) diluted 1:200, mouse anti-tubulin (12G10, Developmental Studies Hybridoma Bank) diluted 1:5000, rabbit anti-GFP (#2555S, Cell Signaling), mouse anti-p53 (DO-1, #sc-126, Santa Cruz), rabbit anti-p53 (#NB200-171, Novusbio), mouse anti-actin (JLA20, Developmental Studies Hybridoma Bank) diluted 1:500, rabbit anti-tubulin (#2144S, Cell Signaling), rabbit anti-p21 (#ab188224, Abcam), rabbit anti histone H3 (# 9715S, Cell Signaling), mouse anti-FLAG M2 (Sigma) diluted 1:2500, rabbit anti-PML (sc-5621) diluted 1:500, and mouse anti-HA (#901501, Biolegend), and pan-cytokeratin antibody (AE1 + AE3 conjugated to Alexa Fluor® 647, Novus, cat. no. NBP2-33200AF647).
HeLa (ATCC, CCL-2) and DLD1 (ATCC CCL-221) cells were grown in DMEM with 10% FBS, 100 U/mL penicillin and 100 mg/mL streptomycin and cultured in 5% CO2 at 37 °C. MDA-MB-231 (ATCC HTB-26) cells were grown in DMEM with 10% FBS supplemented with 100 U/mL penicillin and 100 mg/mL streptomycin and cultured in 10% CO2 at 37 °C. MCF10A cells (ATCC CRL-10317) were grown in DMEM/F12 with 5% horse serum, 100 U/mL penicillin, 100 mg/mL streptomycin, 20 ng/mL EGF, 0.5 mg/mL hydrocortisone, 100 ng/mL cholera toxin and 10 μg/mL insulin and cultured in 5% CO2 at 37 °C. SKBR3 (ATCC HTB-3) cells were grown in ATCC-formulated McCoy’s 5a Medium Modified with 10% FBS supplemented with 100 U/mL penicillin and 100 mg/mL streptomycin and cultured in 5% CO2 at 37 °C. Stable cell lines were generated by transducing cells stably expressing the tet repressor with retroviruses expressing full length wild type Mad1-3xFLAG, Mad1(ΔCTD)-3xFLAG, or Mad1(SIM mutant)-3xFLAG under a tet-inducible promoter. Stable integrants were selected with 4 μg/mL puromycin and 200 µg/mL blasticidin and validated for inducible expression of Mad1 upon tet addition. Virus expressing shRNA against Mad1 (5′-TGAGATCTTTGAACAACTT-3′ for Mad1-shRNA#40, 5′-AGCGATTGTGAAGAACATG-3′ for Mad1-shRNA#717, 5′-GCTTGCCTTGAAGGACAAG-3′ for shRNA#1036, 5′-GCGATTGTGAAGAACATGA-3′ for ad1- shRNA#718) was made from pSUPERIOR.retro.puro. SKBR3 cells were infected by virus containing shRNA sequence against Mad1 and stable integrants were selected with 4 μg/mL puromycin.
Plasmids, mutagenesis, cloning, and virus production
Cloning was performed using Gibson assembly. The plasmids and primers used in this study are listed in Supplementary Table 1. Full length and truncated Mad1 with 3xFLAG or HA tag were cloned into the pRetro-CMV2-TO-puromycin vector. The SIM mutation of Mad1 was generated by site-directed mutagenesis. The pCMV-HA-PML-IV plasmid was a kind gift from Drs. Moon Hee Lee and Shigeki Miyamoto. Nucleolin-GFP and 3xFLAG-MDM2 cDNAs linked by P2A were cloned into pRetro-CMV2-TO-Hygromycin vector. Retroviruses were first generated by transient transfection of HEK 293T cells with the pReto-CMV-TO and separate plasmids that express Gag-Pol, Rev, Tat, and VSV-G. Supernatants were clarified by filtration. Nucleolin-mNeonGreen and mScarlet-MDM2 cDNAs linked by P2A, Mad1-mNeonGreen, and mNeonGreen were cloned into pENTR1A vector (Invitrogen), which contains a CMV promoter, and SV40 polyA signal, and introduced into the pAd/PL-DEST vector by using LR Clonase (Invitrogen). HEK293A cells were transfected with pAd/PL-DEST vectors after linearization with PacI. Viral particles were isolated by three freeze–thaw cycles and amplified by reinfection of HEK293A cells.
Immunostaining of tumor tissues
De-identified formalin-fixed paraffin-embedded human breast cancer tissue sections were obtained from the Translational Research Initiatives in Pathology Lab at UW-Madison (IRB exempt). Samples were deparaffinized and rehydrated by sequential incubation in xylene (3 × 5 min), 100% ethanol (3 × 5 min), 95% ethanol (3 × 5 min), 75% ethanol (2 × 5 min), and H2O (2 × 5 min). Antigen retrieval was performed by heating for 30 min at 95 °C in 10 mM citrate buffer (pH 6.0). Samples were cooled for at least 30 min and washed 3 × 5 min with 1× TBS/0.3% Triton X-100. Tissues were delineated using a hydrophobic barrier pen. Tissues were blocked for 24 h at 4 °C with 1× TBS/5% goat serum and 5% BSA, then incubated overnight at 4 °C with primary antibodies diluted in 5% goat serum in 1× TBS/0.3% Triton X-100. Antibodies used were anti-Mad1 rabbit polyclonal (1:2000) and anti-SUMO1 mouse monoclonal (1:250). Tissues were washed 3 × 10 min in 1× TBS/0.3% Triton X-100 followed by incubation with FITC and TRITC labeled secondary antibodies for 1 h. Tissues were then washed 5 × 10 min in 1× TBS/0.3% Triton X-100 and incubated overnight at 4 °C with pan-cytokeratin antibody. Tissues were washed 3 × 10 min with 1× TBS/0.3% Triton X-100 before mounting with DAPI.
5 × 106 HEK-293T cells were seeded in 10 cm cell culture dishes and grown for 24 h. Cells were transiently transfected using calcium phosphate with 25 µg plasmid. Cells were scraped in RIPA buffer (50 mM Tris–HCl, pH 8.0, 1 mM EDTA, 150 mM NaCl, 20% glycerol, 1% NP-40, 0.5% sodium deoxycholate, and 1× PMSF) and cell lysates kept on ice for 30 min. Cell lysates were cleared by centrifugation at 10,000 × g for 1 min and immunoprecipitations were performed with anti-FLAG M2 affinity gel (No. A2220-5ML, Sigma-Aldrich) or anti-HA magnetic beads (No. B26202, Bimake) for 5 h at 4 °C. Samples were washed 5 × 30 min in RIPA buffer. Proteins of immunoprecipitates and total cell lysates were separated by 12% SDS-PAGE, transferred to nitrocellulose membranes, blocked with 5% milk in TBST and analyzed by immunoblotting with indicated antibodies.
Immunofluorescence and immunoblotting
Cells were rinsed in PBS, fixed for 10 min in 4% formaldehyde, and permeabilized for 7 min in 0.25% Triton X-100 in PBS. Unless otherwise specified, images were acquired on a Nikon Ti‐E inverted microscope using a CoolSNAP HQ2 camera driven by Nikon Elements software and subsequently deconvolved using the AQI 3D Deconvolution module in Nikon Elements. 2D maximum projections assembled in Elements are shown. Overlays were generated in Photoshop. For immunoblotting, equal numbers of cells were lysed in 2× sample buffer. Proteins were separated by 12% SDS-PAGE, transferred to nitrocellulose, blocked with 5% milk in TBST and probed with primary and secondary antibodies.
MBP pull-down experiments
Protein expression was carried out in BL21 Escherichia coli cotransformed with pMAL-c2x (Mad-1, Mad-1 CTD, MBP alone) and pET-28a (Smt3p-PML, Smt3p-3xSUMO2-PML). The cells were sonicated in lysis buffer (50 mM HEPES pH 7.6, 100 mM NaCl, 1 mM DTT) and the lysate obtained after centrifugation was incubated with SUMO protease overnight at 4 °C to remove the Smt3p tag. The lysate was then incubated with amylose resin for 1 h at 4 °C. After washing with lysis buffer, bound proteins were eluted from the resin with elution buffer (50 mM HEPES pH 7.6, 100 mM NaCl, 1 mM DTT, 10 mM maltose) and analyzed via immunoblot.
Nuclear and cytoplasmic protein fractionation
Cells from a 10 cm dish were harvested and incubated in Buffer A (200 mM HEPES pH 7.9, 400 mM NaCl, 1 mM EDTA, 1 mM EGTA, pH 7.4) with protease inhibitors for 10–15 min on ice. NP-40 was added to a final concentration of 0.5%. Lysate was centrifuged at 2000 × g for 5 min to separate supernatant (cytoplasm) and pellet (nuclei).
Orthotopic mouse model
All animal studies were performed in compliance with all relevant ethical regulations for animal testing and research. The study was approved by the Institutional Animal Care and Use Committee of the University of Wisconsin-Madison. 2.5 × 106 MDA-MB-231 or 5 × 106 MCF10A cells were injected into mammary fat pads of 5-week-old female athymic nude mice. Mice were fed Teklad TD.120769, which contains 625 mg/kg doxycycline. Tumor size was measured every 2 days using calipers. Tumor volumes were calculated using the formula v = width2 × length/2. After 7 weeks, necropsy was performed, and tumors were harvested for protein isolation.
The Sen–Adichie test calculated in MTSAT (https://mcardle.wisc.edu/mstat/), version 6.4.2 was used to compare tumor volumes. For all other experiments, significant differences were determined using a two-tailed Student t test calculated using the R commander (Rcmdr) package in R (https://socialsciences.mcmaster.ca/jfox/Misc/Rcmdr/) version 2.4-4. Results are presented as mean ± SD unless otherwise specified.
Further information on experimental design is available in the Nature Research Reporting Summary linked to this article.
The authors declare that all data supporting the findings of this study are available within the paper and its Supplementary Information files. The source data underlying Figs. 1f, g, i, 2b, c, h, i, 3d, f, 4a–e, 5b, d, f, 6a–c, 7b, d f, g and Supplementary Figs. 1b, 1e, 1i-m, 2a-b, 3b, 4a-d and 5b-c are provided as a Source Data file.
Hoyt, M. A., Totis, L. & Roberts, B. T. S. cerevisiae genes required for cell cycle arrest in response to loss of microtubule function. Cell 66, 507–517 (1991).
Ryan, S. D. et al. Up-regulation of the mitotic checkpoint component Mad1 causes chromosomal instability and resistance to microtubule poisons. Proc. Natl. Acad. Sci. USA 109, E2205–E2214 (2012).
Rodriguez-Bravo, V. et al. Nuclear pores protect genome integrity by assembling a premitotic and mad1-dependent anaphase inhibitor. Cell 156, 1017–1031 (2014).
Moyle, M. W. et al. A Bub1–Mad1 interaction targets the Mad1–Mad2 complex to unattached kinetochores to initiate the spindle checkpoint. J. Cell Biol. 204, 647–657 (2014).
London, N. & Biggins, S. Mad1 kinetochore recruitment by Mps1-mediated phosphorylation of Bub1 signals the spindle checkpoint. Genes Dev. 28, 140–152 (2014).
Chung, E. & Chen, R. H. Spindle checkpoint requires Mad1-bound and Mad1-free Mad2. Mol. Biol. Cell 13, 1501–1511 (2002).
Chen, R. H., Shevchenko, A., Mann, M. & Murray, A. W. Spindle checkpoint protein Xmad1 recruits Xmad2 to unattached kinetochores. J. Cell Biol. 143, 283–295 (1998).
Sironi, L. et al. Mad2 binding to Mad1 and Cdc20, rather than oligomerization, is required for the spindle checkpoint. EMBO J. 20, 6371–6382 (2001).
Seeley, T. W., Wang, L. & Zhen, J. Y. Phosphorylation of human MAD1 by the BUB1 kinase in vitro. Biochem. Biophys. Res. Commun. 257, 589–595 (1999).
Hewitt, L. et al. Sustained Mps1 activity is required in mitosis to recruit O-Mad2 to the Mad1-C-Mad2 core complex. J. Cell Biol. 190, 25–34 (2010).
De Antoni, A. et al. The Mad1/Mad2 complex as a template for Mad2 activation in the spindle assembly checkpoint. Curr. Biol. 15, 214–225 (2005).
Cleveland, D. W., Mao, Y. & Sullivan, K. F. Centromeres and kinetochores: from epigenetics to mitotic checkpoint signaling. Cell 112, 407–421 (2003).
Musacchio, A. & Desai, A. A molecular view of kinetochore assembly and function. Biology 6, E5 (2017).
Iwanaga, Y. et al. Heterozygous deletion of mitotic arrest-deficient protein 1 (MAD1) increases the incidence of tumors in mice. Cancer Res. 67, 160–166 (2007).
Iouk, T., Kerscher, O., Scott, R. J., Basrai, M. A. & Wozniak, R. W. The yeast nuclear pore complex functionally interacts with components of the spindle assembly checkpoint. J. Cell Biol. 159, 807–819 (2002).
Campbell, M. S., Chan, G. K. & Yen, T. J. Mitotic checkpoint proteins HsMAD1 and HsMAD2 are associated with nuclear pore complexes in interphase. J. Cell Sci. 114, 953–963 (2001).
Cairo, L. V., Ptak, C. & Wozniak, R. W. Mitosis-specific regulation of nuclear transport by the spindle assembly checkpoint protein Mad1p. Mol. Cell 49, 109–120 (2013).
Wan, J. et al. A Golgi-localized pool of the mitotic checkpoint component Mad1 controls integrin secretion and cell migration. Curr. Biol. 24, 2687–2692 (2014).
Wan, J. & Weaver, B. A. Mad1 at the Golgi apparatus: a story beyond kinetochores. Cell Cycle 14, 1763–1764 (2015).
Han, S. et al. Clinical implication of altered expression of Mad1 protein in human breast carcinoma. Cancer 88, 1623–1632 (2000).
Sun, Q. et al. Increased expression of mitotic arrest deficient-like 1 (MAD1L1) is associated with poor prognosis and insensitive to Taxol treatment in breast cancer. Breast Cancer Res. Treat. 140, 323–330 (2013).
Funk, L. C., Zasadil, L. M. & Weaver, B. A. Living in CIN: mitotic infidelity and its consequences for tumor promotion and suppression. Dev. Cell 39, 638–652 (2016).
Zasadil, L. M., Britigan, E. M. & Weaver, B. A. 2n or not 2n: Aneuploidy, polyploidy and chromosomal instability in primary and tumor cells. Semin. Cell Dev. Biol. 24, 370–379 (2013).
Holland, A. J. & Cleveland, D. W. Boveri revisited: chromosomal instability, aneuploidy and tumorigenesis. Nat. Rev. Mol. Cell Biol. 10, 478–487 (2009).
Weberruss, M. & Antonin, W. Perforating the nuclear boundary—how nuclear pore complexes assemble. J. Cell Sci. 129, 4439–4447 (2016).
de Thé, H., Le Bras, M. & Lallemand-Breitenbach, V. The cell biology of disease: acute promyelocytic leukemia, arsenic, and PML bodies. J. Cell Biol. 198, 11–21 (2012).
Guan, D. & Kao, H. Y. The function, regulation and therapeutic implications of the tumor suppressor protein, PML. Cell Biosci. 5, 60 (2015).
Kubbutat, M. H., Jones, S. N. & Vousden, K. H. Regulation of p53 stability by Mdm2. Nature 387, 299–303 (1997).
Lakin, N. D. & Jackson, S. P. Regulation of p53 in response to DNA damage. Oncogene 18, 7644–7655 (1999).
Haupt, Y., Maya, R., Kazaz, A. & Oren, M. Mdm2 promotes the rapid degradation of p53. Nature 387, 296–299 (1997).
Bernardi, R. et al. PML regulates p53 stability by sequestering Mdm2 to the nucleolus. Nat. Cell Biol. 6, 665–672 (2004).
Louria-Hayon, I. et al. The promyelocytic leukemia protein protects p53 from Mdm2-mediated inhibition and degradation. J. Biol. Chem. 278, 33134–33141 (2003).
Wei, X. et al. Physical and functional interactions between PML and MDM2. J. Biol. Chem. 278, 29288–29297 (2003).
Gottifredi, V. & Prives, C. P53 and PML: new partners in tumor suppression. Trends Cell Biol. 11, 184–187 (2001).
Fu, C. et al. Stabilization of PML nuclear localization by conjugation and oligomerization of SUMO-3. Oncogene 24, 5401–5413 (2005).
Tatham, M. H. et al. RNF4 is a poly-SUMO-specific E3 ubiquitin ligase required for arsenic-induced PML degradation. Nat. Cell Biol. 10, 538–546 (2008).
Kakizuka, A. et al. Chromosomal translocation t(15;17) in human acute promyelocytic leukemia fuses RAR alpha with a novel putative transcription factor, PML. Cell 66, 663–674 (1991).
Torok, D., Ching, R. W. & Bazett-Jones, D. P. PML nuclear bodies as sites of epigenetic regulation. Front Biosci. 14, 1325–1336 (2009).
Chen, R. H., Brady, D. M., Smith, D., Murray, A. W. & Hardwick, K. G. The spindle checkpoint of budding yeast depends on a tight complex between the Mad1 and Mad2 proteins. Mol. Biol. Cell 10, 2607–2618 (1999).
Lee, S. H., Sterling, H., Burlingame, A. & McCormick, F. Tpr directly binds to Mad1 and Mad2 and is important for the Mad1–Mad2-mediated mitotic spindle checkpoint. Genes Dev. 22, 2926–2931 (2008).
Wan, J. et al. A Golgi-localized pool of the mitotic checkpoint component Mad1 controls integrin secretion and cell migration. Curr. Biol. 24, 2687–2692 (2014).
Lallemand-Breitenbach, V. & de Thé, H. PML nuclear bodies. Cold Spring Harb. Perspect. Biol. 2, a000661 (2010).
Nisole, S., Maroui, M. A., Mascle, X. H., Aubry, M. & Chelbi-Alix, M. K. Differential roles of PML isoforms. Front. Oncol. 3, 125 (2013).
Zhao, Q. et al. GPS-SUMO: a tool for the prediction of sumoylation sites and SUMO-interaction motifs. Nucleic Acids Res. 42, W325–W330 (2014).
Ren, J. et al. Systematic study of protein sumoylation: development of a site-specific predictor of SUMOsp 2.0. Proteomics 9, 3409–3412 (2009).
Wan, J., Subramonian, D. & Zhang, X. D. SUMOylation in control of accurate chromosome segregation during mitosis. Curr. Protein Pept. Sci. 13, 467–481 (2012).
Zhang, X. D. et al. SUMO-2/3 modification and binding regulate the association of CENP-E with kinetochores and progression through mitosis. Mol. Cell 29, 729–741 (2008).
Osipovitch, M. et al. Automated protein motif generation in the structure-based protein function prediction tool ProMOL. J. Struct. Funct. Genomics 16, 101–111 (2015).
Bamford, S. et al. The COSMIC (Catalogue of Somatic Mutations in Cancer) database and website. Br. J. Cancer 91, 355–358 (2004).
Jia, L. Q. et al. Screening the p53 status of human cell lines using a yeast functional assay. Mol. Carcinog. 19, 243–253 (1997).
Leroy, B. et al. Analysis of TP53 mutation status in human cancer cell lines: a reassessment. Hum. Mutat. 35, 756–765 (2014).
Merlo, G. R., Basolo, F., Fiore, L., Duboc, L. & Hynes, N. E. p53-dependent and p53-independent activation of apoptosis in mammary epithelial cells reveals a survival function of EGF and insulin. J. Cell Biol. 128, 1185–1196 (1995).
He, G. et al. Induction of p21 by p53 following DNA damage inhibits both Cdk4 and Cdk2 activities. Oncogene 24, 2929–2943 (2005).
Kachnic, L. A. et al. The ability of p53 to activate downstream genes p21(WAF1/cip1) and MDM2, and cell cycle arrest following DNA damage is delayed and attenuated in scid cells deficient in the DNA-dependent protein kinase. J. Biol. Chem. 274, 13111–13117 (1999).
Krenning, L., Feringa, F. M., Shaltiel, I. A., van den Berg, J. & Medema, R. H. Transient activation of p53 in G2 phase is sufficient to induce senescence. Mol. Cell 55, 59–72 (2014).
Guo, A. et al. The function of PML in p53-dependent apoptosis. Nat. Cell Biol. 2, 730–736 (2000).
Sironi, L. et al. Crystal structure of the tetrameric Mad1–Mad2 core complex: implications of a ‘safety belt’ binding mechanism for the spindle checkpoint. EMBO J. 21, 2496–2506 (2002).
Honda, R., Tanaka, H. & Yasuda, H. Oncoprotein MDM2 is a ubiquitin ligase E3 for tumor suppressor p53. FEBS Lett. 420, 25–27 (1997).
Yuan, B. et al. Increased expression of mitotic checkpoint genes in breast cancer cells with chromosomal instability. Clin. Cancer Res. 12, 405–410 (2006).
Ricke, R. M., van Ree, J. H. & van Deursen, J. M. Whole chromosome instability and cancer: a complex relationship. Trends Genet. 24, 457–466 (2008).
Carroll, B. L., Pulkoski-Gross, M. J., Hannun, Y. A. & Obeid, L. M. CHK1 regulates NF-κB signaling upon DNA damage in p53-deficient cells and associated tumor-derived microvesicles. Oncotarget 7, 18159–18170 (2016).
Xu, J. & Loo, G. Different effects of genistein on molecular markers related to apoptosis in two phenotypically dissimilar breast cancer cell lines. J. Cell. Biochem. 82, 78–88 (2001).
Ricke, R. M., Jeganathan, K. B., Malureanu, L., Harrison, A. M. & van Deursen, J. M. Bub1 kinase activity drives error correction and mitotic checkpoint control but not tumor suppression. J. Cell Biol. 199, 931–949 (2012).
Kalitsis, P. et al. Increased chromosome instability but not cancer predisposition in haploinsufficient Bub3 mice. Genes Chromosomes Cancer 44, 29–36 (2005).
Rowald, K. et al. Negative selection and chromosome instability induced by Mad2 overexpression delay breast cancer but facilitate oncogene-independent outgrowth. Cell Rep. 15, 2679–2691 (2016).
Li, F. et al. The BUB3–BUB1 complex promotes telomere DNA replication. Mol. Cell 70, 395–407 (2018).
Cubeñas-Potts, C. et al. Identification of SUMO-2/3-modified proteins associated with mitotic chromosomes. Proteomics 15, 763–772 (2015).
Yang, F. et al. BubR1 is modified by sumoylation during mitotic progression. J. Biol. Chem. 287, 4875–4882 (2012).
Lee, C. C., Li, B., Yu, H. & Matunis, M. J. Sumoylation promotes optimal APC/C activation and timely anaphase. eLife 7, e29539 (2018).
Kim, S., Sun, H., Tomchick, D. R., Yu, H. & Luo, X. Structure of human Mad1 C-terminal domain reveals its involvement in kinetochore targeting. Proc. Natl. Acad. Sci. USA 109, 6549–6554 (2012).
The authors thank Dr. Moon Hee Lee and Dr. Shigeki Miyamoto for the PML-IV cDNA, the Biomedical Research Model Services Mouse Breeding Core for use of its facilities and services, and members of the Burkard and Weaver laboratories for useful discussions. The authors also thank the University of Wisconsin Translational Research Initiatives in Pathology laboratory, in part supported by the UW Department of Pathology and Laboratory Medicine and UWCCC Grant P30 CA014520, for use of its facilities and services. Beth A. Weaver was supported by a Research Scholar Grant, RSG-15-006-01 - CCG, from the American Cancer Society. This work was also supported, in part, by predoctoral fellowship 16PRE29650011 from the American Heart Association (J.W.) and NIH Grants GM088151 (A.A.) and T32CA009135 (C.S.).
The authors declare no competing interests.
Journal peer review information: Nature Communications thanks Patrick Meraldi and the other anonymous reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Wan, J., Block, S., Scribano, C.M. et al. Mad1 destabilizes p53 by preventing PML from sequestering MDM2. Nat Commun 10, 1540 (2019). https://doi.org/10.1038/s41467-019-09471-9