Mutant P53 Gains New Function in Promoting Inflammatory Signals By

ORIGINAL ARTICLE Mutant p53 gains new function in promoting inﬂammatory signals by repression of the secreted interleukin-1 receptor antagonist

V Ubertini1, G Norelli1, D D'Arcangelo2, A Gurtner1, E Cesareo2, S Baldari1, MP Gentileschi1, G Piaggio1, P Nisticò1, S Soddu1, A Facchiano2 and G Bossi3

The TP53 tumor-suppressor gene is frequently mutated in human cancer. Missense mutations can add novel functions (gain-of- function, GOF) that promote tumor malignancy. Here we report that mutant (mut) p53 promotes tumor malignancy by suppressing the expression of a natural occurring anti-inflammatory cytokine, the secreted interleukin-1 receptor antagonist (sIL-1Ra, IL1RN). We show that mutp53 but not wild-type (wt) p53 suppresses the sIL-1Ra production in conditioned media of cancer cells. Moreover, mutp53, but not wtp53, binds physically the sIL-1Ra promoter and the protein–protein interaction with the transcriptional co- repressor MAFF (v-MAF musculoaponeurotic fibrosarcoma oncogene family, protein F) is required for mutp53-induced sIL-1Ra suppression. Remarkably, when exposed to IL-1 beta (IL-1β)inflammatory stimuli, mutp53 sustains a ready-to-be-activated in vitro and in vivo cancer cells’ response through the sIL-1Ra repression. Taken together, these results identify sIL-1Ra as a novel mutp53 target gene, whose suppression might be required to generate a chronic pro-inflammatory tumor microenvironment through which mutp53 promotes tumor malignancy.

Oncogene (2015) 34, 2493–2504; doi:10.1038/onc.2014.191; published online 7 July 2014

INTRODUCTION sIL-1Ra acts as a specific antagonist of the IL-1 pro-inflammatory The mutp53 gain-of-function (GOF) activities are exerted in a cytokines: it binds to both type I and type II IL-1 receptors, with variety of ways, including enhanced proliferation in culture, approximately equal affinity as compared with IL-1α and IL-1β, 10 increased tumorigenicity in vivo and enhanced resistance to a without exerting any agonist activity. The pro-inflammatory IL-1 variety of commonly used anti-cancer drugs.1 The GOF hypothesis is produced by a wide variety of cell types and is implicated in the has recently been reinforced with mutp53 ‘knock-in’ mice, which pathogenesis of acute and chronic inflammatory or autoimmune 11 show a higher frequency of tumor development and increased diseases. Notably, chronic inflammatory processes are etiologi- 12,13 metastatic potential compared with p53-deficient mice.2,3 Further- cally linked to many human cancers, and preclinical studies more, RNA interference (RNAi) studies demonstrated that deple- provide ample support to suggest reduction of IL-1 activity as a 14,15 tion of endogenous mutants renders cancer cells more sensitive to potential therapeutic target in human cancers. DNA-damaging chemotherapeutic agents in vitro and reduces Our results provided evidence of a molecular and functional tumor malignancy.4,5 relation between sIL-1Ra and mutp53 in human cancer cells, We previously showed that conditional depletion of endogen- suggesting that mutp53 by suppressing sIL-1Ra may support a ous mutp53 reduces growth, stromal invasion and angiogenesis pro-inflammatory tumor microenvironment promoting tumor in HT29 xenograft tumors.6 The results were suggesting that malignancy. mutp53 might be putatively involved in the regulation of tumor microenvironment. To determine whether specific secretory RESULTS pathways might be involved in mutp53-driven effects in vivo, we analyzed the cytokines production in a panel of human cancer Mutant but not wild-type (wt) p53 suppresses sIL-1Ra gene lines harboring different p53 mutants. Here we report the expression in tumor cells identification and characterization of the secreted interleukin-1 Based on previous studies,6 we enquired whether mutp53 might receptor antagonist (sIL-1Ra) as a novel mutp53 repressed exert novel, unidentified roles in the regulation of tumor target gene. microenvironment. To address this issue, we investigated the The sIL-1Ra represents the first described specific receptor cytokines pathways in a panel of human cancer cells harboring antagonist of any cytokine or hormone-like molecule. The sIL-1Ra different p53 hot-spot mutations and earlier engineered with a is encoded by the IL1RN gene along with the closely related lentiviral-based TET-OFF inducible RNAi system encoding p53 (sh- intracellular IL-1Ra, whose expression is controlled by the activity p53) or scrambled (sh-scr) shRNA-specific sequences.6,16 Hence, of two separate 5′ regulatory regions.7,8 The sIL-1Ra along with the conditioned media (CMs) were generated from sh-p53 and sh-scr two agonistic proteins IL-1α and β constitutes the IL-1 family.9 The cells upon delivery of tetracycline analogs (doxycycline (DOX)).

1Experimental Oncology Laboratories, Regina Elena National Cancer Institute, Rome, Italy; 2Istituto Dermopatico Dell’immacolata IDI-IRCSS, Rome, Italy and 3Laboratory of Medical Physics and Expert Systems, Regina Elena National Cancer Institute, Rome, Italy. Correspondence: Dr G Bossi, Laboratory of Medical Physics and Expert Systems, Regina Elena National Cancer Institute, Via delle Messi d’Oro 156, Rome 00158, Italy. E-mail: [email protected] Received 2 January 2014; revised 30 April 2014; accepted 28 May 2014; published online 7 July 2014 Mutp53 promotes inﬂammation by sIL-1Ra repression V Ubertini et al 2494

Figure 1. Mutant p53R273H suppresses sIL-1Ra protein production in cancer cells. (a) CMs, generated from either engineered sh-p53 and sh- scr HT29, MDA-MB468, SKBR3 or MDA-MB231 cancer lines (see Materials and methods for details), were analyzed by cytokine arrays. Relative sIL-1Ra production was quantified with respect to controls (sh-scr-DOX) set to 1.0. Results are reported as means and s.d.s. of three independent experiments. Upper panels, western blotting (WB) analyses were performed, on derived protein lysates, with the indicated antibodies. More relevant bands from the same filter at same exposure length are reported. (b) Total RNAs from engineered sh-p53 and sh-scr HT29 and MDA-MB468 were analyzed by reverse transcriptase–PCR (RT–PCR) to assess occurred mutp53 depletion and qPCR to analyze effects on sIL-1Ra gene expression. Relative sIL-1Ra mRNA level were quantified with respect to controls (sh-scr) set to 1.0. Representative data of three independent experiments with similar results are reported. (c) MCF7 cells were treated with either Nut-3 (10 μM) or DMSO, and RNAs and proteins were extracted 48 h later. Upper panel, WB analyses were performed with the indicated antibodies. Relevant bands from the same filter at the same exposure levels are reported. Lower panel, qPCR analyses. Relative sIL-1Ra mRNA level were quantified with respect to DMSO-treated cells set to 1.0. Representative data of three independent experiments with similar results are reported. (d) WBs were performed with the indicated antibody on CMs generated from the indicated cancer cells along with recombinant human sIL-1Ra (Kineret) (positive control). Relevant bands from the same filter at the same exposure levels are reported. Representative data of three independent experiments with similar results are reported. RT–PCR images were acquired by Bio-Rad Universal Hood II gel-imager.

Analyses of cytokines showed a significantly higher sIL-1Ra genotoxic wtp53 activator. Activated wtp53, as confirmed by production in CMs from HT29 and MDA-MB468 cancer cells upon induced p21 expression (Figure 1c), increases significantly the depletion of endogenous p53R273H (sh-p53) with respect to sIL-1Ra mRNA (Figure 1c, lower panel) and secreted protein controls (sh-p53 minus DOX; sh-scr) (Figure 1a). Of interest, no (Figure 1d) with respect to cells treated only with the Nut-3 significant effects were observed in MDA-MB231 and SKBR3 cells solvent, dimethyl sulfoxide (DMSO). These results show that upon depletion of their R280K and R175H mutants, respectively p53R273H represses, whereas wtp53 induces sIL-1Ra expression, fi (Figure 1a). Further analyses con rmed the higher sIL-1Ra mRNA suggesting a novel mutp53 GOF activity. (Figure 1b) and protein production (Figure 1d) in sh-p53 HT29 and MDA-MB468 cancer cells with respect to controls (sh-scr). To explore whether the wtp53 protein might have roles in Mutant but not wtp53 reduces sIL-1Ra promoter activity sIL-1Ra gene expression, we performed experiments with MCF7 To investigate whether the increased sIL-1Ra gene expression cells upon delivery of Nutlin-3 (Nut-3), a well-known non- upon either p53R273H depletion or wtp53 activation might

Figure 2. Mutant p53 inhibits, whereas wtp53 increases the sIL-1Ra gene expression modulating its promoter activity. (a) The activities of sIL-1Ra regulatory region (−1680;+1 with respect to transcriptional start site) were analyzed in either: (i) early lentiviral-transduced sh-scr and sh-p53 HT29 and MDA-MB468 cells; and (ii) Nut-3 or DMSO-treated MCF7 cells. Outcomes were normalized to transfection efficiency (βGAL) and protein quantity. Relative Luc activity was quantified with respect to controls (sh-scr or DMSO) set to 1.0. Results are reported as mean and s.d. of three independent experiments. (b) Endogenous sIL-1Ra promoter activities. Chromatins were isolated from sh-scr and sh-p53 HT29 cells, and ChIP assays performed to analyzed the recruitments of histone H3 pan-acetylated (H3Ac pan); H3 acetylated at lysine 9 (H3K9ac), H3 acetylated at lysine 14 (H3K14ac) and H3 tri-methylated at lysine 4 (H3K4me3). Representative data of three independent experiments with similar results are reported. (c) HEPG2 cells were transfected with pRA1680-Luc vector along with either empty, wtp53, p53R273H or p53R175H expressing vectors. Cells were processed 48 h later after 24 h of LPS (30 ng/ml) treatments. Upper panel: western blotting analyses performed with the indicated antibodies. Lower panel: the sIL-1Ra promoter activities, analyzed by luciferase assays, are reported as means and s.d.s. of three independent experiments. Relative Luc activities were quantified with respect to controls (empty vector) set to 1.0. (d) Engineered sh-scr and sh-p53 HT29 sublines, after 72h incubation with DOX (1.0 μg/ml) were plated in 60 mm dishes. Twentyfour hours later sh-p53 cells were transfected either with p53R175H or p53R273H carrying vectors. Forty-eight hours later, total RNAs were isolated and analyzed by PCR (GAPDH, p53) and QPCR (sIL-1Ra). Relative sIL-1Ra mRNA levels were quantified with respect to sh-scr RNA set to 1.0. Representative data of three independent experiments with similar results are reported. No Ab, no antibody. occur by modulating the sIL-1Ra promoter activity, assays MCF7 cells with respect to their controls (sh-scr and DMSO) were performed with the sIL-1Ra regulatory region driving a (Figure 2a). To further corroborate these results, we analyzed luciferase reporter (pRA1680-Luc; − 1680 bp with respect to the transcriptional activity of the endogenous sIL-1Ra promoter. transcriptional start site), which exhibits patterns of As shown in Figure 2b, an increased occupancy of active expression and induction similar to that of the endogenous chromatin markers was detected in sh-p53 HT29 cells with respect gene.7 Reporter assays were performed with sh-p53 and sh-scr to controls (sh-scr). These results indicate that, under our HT29 and MDA-MB468 cells and with Nut-3- and DMSO-treated experimental conditions, p53R273H suppresses sIL-1Ra gene MCF7 cells. Significantly higher sIL-1Ra promoter activity was expression by reducing its promoter activity, whereas wtp53 found in sh-p53 HT29 and MDA-MB468 cells and Nut-3-treated increases promoter activity.

© 2015 Macmillan Publishers Limited Oncogene (2015) 2493 – 2504 Mutp53 promotes inflammation by sIL-1Ra repression V Ubertini et al 2496 We next asked whether the sIL-1Ra suppression occurs in a MAF subgroup of proteins that have been identified as a p53R273H-specific fashion or if other p53 mutants might similarly molecular switch capable of regulating gene expression in a affect the sIL-1Ra gene expression. Thus experiments were positive or negative manner depending on their abundance.25 performed in isogenic conditions with HEPG2 cells, a line widely Indeed, small MAFs act as heterodimers with CNC (cap’n’ collar) used to study the production of acute-phase proteins (APPs), as proteins functioning as transcriptional activators, whereas as sIL-1Ra was described as an APP.17 The sIL-1Ra promoter activity homodimer they function as transcriptional repressors. Interest- was evaluated upon transient ectopic expression of wtp53, ingly, the MAFF promoter occupancy was significantly affected p53R273H or p53R175H encoding cDNAs. The results showed upon p53R273H depletion (Figure 3b, right panels) without that, with respect to control (empty vector), exogenous wtp53 disturbing the MAFF gene expression in either tested lines at significantly increases the sIL-1Ra promoter activity, whereas both the mRNA (Figure 3c, upper panel) and protein levels (Figure 3c, ectopically expressed mutants inhibit its activity (Figure 2c). lower panel). These data suggest that mutp53 could act at the Confirmatory experiments were carried out assessing sIL-1Ra gene level of MAFF recruitment on the sIL-1Ra promoter regulating expression with DOX-treated sh-p53 HT29 cells upon exogenous gene expression. expression of either p53R175H or p53R273H mutants. As shown in We also explored whether activated wtp53 might modulate Figure 2d, the increased sIL-1Ra mRNA level upon depletion of MAFF gene expression and thus sIL-1Ra production. To this aim, endogenous p53R273H drops consistently upon p53R175H or analyses were performed with Nut-3- or DMSO-treated MCF7 cells. p53R273H ectopic expression. These results show that mutp53- As shown in Figure 3d, activated wtp53 affects MAFF gene induced sIL-1Ra suppression might not occur in a mutation- expression, reducing both mRNA (left panel) and protein levels specific fashion, and the discrepancy with SKBR3 results might be (right panel) and, importantly, reducing MAFF recruitment to the related to cell specificity of mutp53 GOF. sIL-1Ra promoter (Supplementary Figure S1). Overall, these results identify MAFF as a relevant player in the Mutant but not wtp53 is physically recruited to the sIL-1Ra control of sIL-1Ra gene expression in both mutant and wtp53 promoter carrying cells. To explore whether mutp53 suppresses sIL-1Ra through the physical interaction with its promoter, chromatin immunoprecipi- MAFF works as a common player in the regulation of sIL-1Ra gene tation (ChIP) assays were performed with chromatins derived from expression by mutant and wtp53 proteins sh-scr and sh-p53 HT29 and MDA-MB468 cells. To identify putative To assess whether the mutp53 and MAFF co-recruitment on the sIL-1Ra regulatory regions involved in mutp53 promoter occu- sIL-1Ra promoter might be a consequence of a protein–protein pancy, primers were designed to dissect the sIL-1Ra promoter interaction, co-immunoprecipitation (co-IP) assays were per- (−1680) in four regions (I–IV) (Figure 3a, upper panel). A significant formed with HT29 and MDA-MB468 cells. A 53-KDa protein p53R273H recruitment was found in sIL-1Ra regions I and III with recognized by an anti-p53 antibody was detected in MAFF both lines (Figure 3a). The enrichment was specific as vanished in immunoprecipitated lysates but not in the controls (Rabbit-IgG), mutp53-depleted chromatins (Figure 3a). We next enquired demonstrating that MAFF complexes to p53R273H in both cancer whether wtp53 could interact with the sIL-1Ra promoter. As lines (Figure 4a). Comparable results were obtained with shown in Figure 3a, no significant enrichment was observed in p53R175H-carrying HEPG2 cells (Supplementary Figure S2). either MCF7 or HCT116 chromatins, indicating that activated We next explored whether MAFF might have a role in the wtp53 is not recruited on the sIL-1Ra promoter. These results are mutp53 recruitment on the sIL-1Ra promoter. To this aim, ChIP in agreement with in silico analyses (MatIspector), which showed assays were performed with HT29 cells upon lentiviral-based RNAi no wtp53 consensus binding site in the sIL-1Ra promoter depletion of endogenous MAFF. Efficient MAFF depletion (sh- − ( 1680 bp) examined (data not shown). MAFF) (Figure 4b, upper panel) significantly hampers the Taken together, these results suggest that while mutp53 could p53R273H recruitment to the sIL-1Ra promoter compared with be directly involved in the sIL-1Ra suppression, the wtp53 might controls (sh-con) (Figure 4b, lower panel). Thus the results suggest fi induce sIL-1Ra expression indirectly, through unde ned molecular that through protein–protein interactions MAFF contributes to players. engage p53R273H on the sIL-1Ra promoter. As the small MAFs abundance has been identified as a fine tuning molecular switch MAFF is involved in sIL-1Ra gene expression in both mutant and regulating positively or negatively gene expression,25 we asked wtp53 carrying cells whether the modulation of MAFF abundance might affect the Currently, two not mutually exclusive possibilities are considered sIL-1Ra gene expression. To modulate MAFF protein levels, HT29 to explain the mutp53-driven transcription of its target genes: cells were transduced with increasing multiplicities of infection (i) mutp53 retains residual transcriptional activity and acts as a (MOIs) of either sh-MAFF or sh-con carrying lentiviruses, and regulator of transcription;18,19 and (ii) mutp53 can no longer bind sIL-1Ra gene expression was measured by quantitative PCR DNA but interacts with other transcription factors (TFs) and (qPCR). It was found that efficient MAFF depletion (15–20 MOI) modulates their activities.20,21 Accordingly, we asked whether inhibits sIL-1Ra gene expression, whereas a mild reduction in specific TFs with repressor activity might be involved in the MAFF (30%, 10 MOI) significantly increased sIL-1Ra mRNA levels mutp53-driven sIL-1Ra suppression. To this end, in silico analyses (Figure 4c) with respect to controls (sh-con). In agreement, mild probing for TF consensus sequence (MatIspector) and binding reduction in MAFF increased the sIL-1Ra promoter activity peaks (ENCODE) on the sIL-1Ra promoter (−1680 bp) identified (Figure 4d, left panel) and the production of secreted protein in (i) growth factor independent 1 (GFI1),22 (ii) hairy-related CM (Figure 4d, right panel) of HT29 cells. Further confirmatory transcription factor-1 (HRT1, HEY1),23 and (iii) v-MAF musculoapo- experiments were carried out using Stealth siRNAs specificto neurotic fibrosarcoma oncogene family, protein F (MAFF).24 To MAFF (Supplementary Figure S3A). To explore whether MAFF determine whether in our cells the identified TFs are physically might have a role in wtp53-induced sIL-1Ra expression, we recruited to the sIL-1Ra promoter, ChIP assays were performed measured the sIL-1Ra gene expression in MCF7 cells upon MAFF with sh-p53 and sh-scr HT29 cells. As shown in Figure 3b (left RNAi. At RNAi conditions that mimic the MAFF modulation upon panels), all analyzed TFs were found recruited to various extents wtp53 activation, a significantly higher sIL-1Ra mRNA was seen on the sIL-1Ra regulatory regions. Of interest, only the MAFF compared with controls (sh-con) (Supplementary Figure S3B). occupancy was in accordance with mutp53 recruitment on These results indicate MAFF as a common player in the regions I and III (Figure 3b, left panel). MAFF belongs to the small regulation of sIL-1Ra gene expression by mutant and wtp53

Figure 3. Mutant but not wtp53 is physically recruited on the sIL-1Ra promoter along with the MAFF transcription factor. (a) ChIP assays were performed with chromatins isolated either from sh-p53 and sh-scr HT29 and MDA-MB468 cells, DMSO or Nut-3-treated MCF7 cells or Adriamycin (Adr)-treated HCT116 cells. ChIP-enriched DNA were quantified by QPCR with set of primers (upper cartoon) designed to dissect the − 1680 sIL-1Ra promoter in four regions (I–IV). Relative enrichments were quantified with respect to ‘No Ab’ set to 1.0. Representative data of three independent experiments with similar results are reported. (b) Occupancy of transcription factors with repressor activity (Gfi1, HRT1, MAFF) on sIL-1Ra promoter. ChIP assays were performed with chromatins from sh-p53 and sh-scr HT29 cells. ChIP-enriched DNA were measured by qPCR with sets of primers specific to sIL-1Ra regulatory regions I and III, and outcomes quantified with respect to ‘No Ab’ set to 1.0. Results are reported as means and s.d.s. of three independent experiments. (c) MAFF gene expression in mutp53-expressing cells. sh-scr and sh-p53 HT29 and MDA-MB468 cancer cells were cultured 72 h in the presence of DOX (1,0 μg/ml) before processing for RNA and protein extraction. Upper panel: Relative MAFF mRNA levels were analyzed by qPCR, normalized to b-actin housekeeping gene and quantified with respect to sh-scr cells set to 1.0. Lower panel: western blotting analyses performed with the indicated antibodies. (d) MAFF gene expression in wtp53-expressing cells. MCF7 cells were either treated with Nut-3 (10 μM) or DMSO (vehicle control). Forty-eight hours later, cells were processed for RNA and protein extraction. Left panel: Relative MAFF mRNA levels were analyzed by qPCR, normalized to b-actin housekeeping gene and quantified with respect to sh-scr cells set to 1.0. Right panel: western blotting analyses performed with the indicated antibodies. Densitometry was performed with the ImageJ software, values were normalized to actin and folds quantified with respect to controls (DMSO) set to 1.0. Representative data of three independent experiments with similar results are reported.

Figure 4. MAFF interacts physically with mutp53 proteins and has relevant roles in mutp53-driven sIL-1Ra suppression. (a) Co-IP assays were performed with endogenous proteins from HT29 and MDA-MB468 total cell lysate. Specific anti-MAFF or control Rabbit-IgG antibodies were used for IP, and anti-p53- and anti-MAFF-specific antibodies were used for immunoblotting (IB). Representative images of three independent experiments are reported. Relevant bands from the same filter at the same exposure levels are shown. (b) HT29 cells were transduced with either sh-MAFF or sh-control (sh-con) carrying lentiviruses (20 MOI). Seventy-two hours later, cells were collected and processed. Upper panel: western blotting analyses were performed with the indicated antibodies. Lower panel: ChIP assays. Gray, dashed, square boxes show the p53R273H and MAFF recruitment on sIL-1Ra promoter. Relevant bands from the same images at the same exposure levels are reported. Representative results of three independent experiments with similar outcomes are reported. (c) HT29 cells were transduced with increasing MOIs (10, 15 and 20) of either sh-MAFF or sh-con carrying lentivuses. Seventy-two hours later, cells were collected and processed to isolated total RNA. MAFF and sIL-1Ra gene expression were analyzed by qPCR. Relative MAFF and sIL-1Ra mRNA levels were normalized to b-actin housekeeping gene and quantified with respect to sh-con RNA set to 1.0. Representative results of two independent experiments in triplicate are reported. (d) HT29 cells were infected with either sh-MAFF or sh-con carrying lentiviruses at MOI 10. Seventy-two hours later, cells were collected and processed. Left panel: Relative Luc activities normalized with transfection efficiency (βGaL) and protein quantity were quantified with respect to controls (sh-con) set to 1.0. Results are reported as means and s.d.s. of three independent experiments. Right panel: Western blottings were performed with protein lysates and CMs with the indicated antibodies. Representative results of three independent experiments are reported.

proteins. Indeed, while activated wtp53 induces sIL-1Ra gene We first established that both HT29 and MDA-MB468 lines possess expression through MAFF repression, mutp53 does not affect the IL-1 receptor type 1 (IL-1R1) and that expression is not affected MAFF gene expression but through protein–protein interaction by mutp53 silencing (Supplementary Figure S4). Because cytokine might increase MAFF abundance, induce homodimers and thus array and reverse transcriptase–PCR analyses showed that both sIL-1Ra suppression. lines do not produce IL-1β in unstimulated conditions (data not shown), we explored cell response to recombinant human (rh) β β fi β IL-1 . As positive control of IL-1 blockade ef ciency, experiments Upon mutp53 depletion, de-repressed sIL-1Ra blocks IL-1 were carried out by pretreating a subset of control cells (sh-scr) response with recombinant rhsIL-1Ra (Kineret), which was delivered at The sIL-1Ra specifically antagonize the IL-1 pro-inflammatory dosages revealed in CMs of mutp53-depleted HT29 and MDA- cytokine. Thus, to assess whether the mutp53-mediated sIL-1Ra MB468 cells (0.2 and 0.5 μg/ml, respectively). Because the IL-1β suppression might contribute to a ready-to-be-activated IL-1 cell response can be studied in vitro by evaluating the IL-8 gene response, conventional in vitro and in vivo assays were carried out. expression,26 we found that rhIL-1β delivery rises efficiently the

Figure 5. Mutant p53 sustains a prompt cell response to IL-1β by suppressing the sIL-1Ra gene expression in the studied cancer lines. (a) IL-1β- induces IL8 gene expression. Engineered sh-p53 and sh-scr HT29 and MDA-MB468 cells after treatments (Material and methods) were collected and processed for RNA extraction. Semi-quantitative PCR (upper panels) were performed to verify mutp53 depletion, whereas IL-8 gene expression was quantiﬁed by qPCR (lower panel). Images were acquired with Bio-Rad Universal Hood II gel-imager. Relative IL-8 mRNA levels were normalized with housekeeping gene beta-actin and quantiﬁed with respect to controls (sh-scr+IL-1β) set to 1.0. Results are reported as means and s.d.s. of two independent experiments in triplicate. (b) IL-1β-induced endothelial cell injury. HUVEC cell monolayers, cultured 24 h with CMs generated from sh-scr and sh-p53 HT29 cells along with treatments (Materials and methods), were collected and endothelial cell injury assessed by trypan blue exclusion analyses. Results are reported as means and s.d.s. of three independent experiments, each in triplicate. (c) IL-1β-induced cancer cells’ proliferation. Cell proliferation of engineered sh-p53 and sh-scr HT29 (left diagram) and MDA- MB468 (right diagram) cancer cells along treatments (Materials and methods) was assessed by BrdU incorporation assays. Results are reported as means and s.d.s. of three independent experiments. Kineret: recombinant human sIL-1Ra; rhIL-1β: recombinant human IL-1β.

IL-8 expression in both the tested lines (Figure 5a: lanes 2 and 4 represents an hallmark of early angiogenesis.26,27 To explore and 8 and 10), whose effects were significantly inhibited by whether de-repressed sIL-1Ra upon mutp53 depletion might mutp53 depletion (Figure 5a: lanes 3 and 9) or Kineret efficiently counteract the IL-1β-induced angiogenic phenotype, pretreatments (Figure 5a: lanes 5 and 6 and 11 and 12). These we evaluated injuries of HUVEC cell monolayers upon co-culture results indicate that in both lines mutp53 sustains the functional with CMs generated from IL-1β-treated sh-scr and sh-p53 HT29 IL-1β signal cascade through the sIL-1Ra suppression. cells as reported above. As shown in Figure 5b, CM from sh-scr We next explored the functionality of de-repressed sIL-1Ra in cells delivered with rhIL-1β induced significantly higher endothe- non-isogenic cell system. To this aim, it is well established that lial cell death with respect to CM of vehicle-treated cells IL-1β-induced permeability of endothelial cell monolayer (Figure 5b, P = 0.014), whereas rhIL-1β effects were significantly

© 2015 Macmillan Publishers Limited Oncogene (2015) 2493 – 2504 Mutp53 promotes inflammation by sIL-1Ra repression V Ubertini et al 2500 inhibited when HUVEC cells were delivered with CM of either inflammation has been shown to have important roles in tumor Kineret pretreated sh-scr (Figure 5b, 0.2 μg/ml, P = 0.007 and biology by influencing tumor growth, invasion and metastasis. 0.5 μg/ml, P = 0.015) or sh-p53 (Figure 5b, P = 0.018) cells. The Among them, IL-1β is a highly active and pleiotropic pro- results demonstrated that upon mutp53 depletion de-repressed inflammatory cytokine implicated in the pathogenesis of many sIL-1Ra abrogated IL-1β effects with efficiency similar to the inflammation-associated diseases. recombinant sIL-1Ra protein. Missense p53 mutations are reported to add novel oncogenic GOF activities through which mutated proteins actively contribute Mutp53 sustains IL-1β-driven tumor malignancy through sIL-1Ra in tumor malignancy. Most of the identified mutp53 GOF effects suppression were shown to be linked to the ability of mutp53 to modulate 38,39 Previous studies have shown that IL-1β promotes proliferation of the expression of several genes uncovering novel cell- breast,28 ovarian29 and colon30 cancer cells in vitro. Thus we autonomous functions. Importantly, emerging evidences under- explored whether mutp53, through sIL-1Ra repression, might lined the existence of non-cell-autonomous wtp53 functions to contribute to sustain IL-1β-induced cell proliferation. As shown in suppress tumorigenesis by promoting an antitumor Figure 5c, rhIL-1β delivery increases 5-bromo-2′-deoxyuridine microenvironment,40,41 and to our knowledge, only one study (BrdU) incorporation in both sh-scr HT29 and MDA-MB468 cells, reported non-cell-autonomous mutp53 GOF.42 Our studies show whose effects were either rescued efficiently upon Kineret that mutp53, through sIL-1Ra suppression, contributes to maintain pretreatments or mutp53 depletion (sh-p53). a prompt IL-1β cancer cell response contributing to add novel Patients with IL-1 producing solid tumors (breast, colon, lung, insights for the identification of novel non-cell-autonomous head and neck and melanomas) have a generally worst mutp53 GOF in human cancer. Indeed, results suggest that 31,32 prognosis. Likewise, preclinical studies demonstrated that mutp53-driven sIL-1Ra repression might contribute in sustaining a fl local IL-1 production in uences tumor growth and metastasis pro-inflammatory tumor microenvironment further promoting either through direct proliferative effects or by promoting host tumor malignancy. In agreement, a molecular study reported fl 33 in ammatory and angiogenic pathways. To explore whether mutp53 to sustain cancer progression by augmenting nuclear β mutp53, through the sIL-1Ra repression, might contribute to IL-1 - factor κB (NFκB) activation in the context of chronic inflammation induced tumor malignancy, we investigated the growth of in vitro and in vivo.43,44 Notably, IL-1β induces the expression of xenograft tumors in lipopolysaccharide (LPS)-treated mice. As canonical IL-1 target genes through the activation of NFκB LPS rapidly induces IL-1 production after in vivo injection,34 it signaling pathway. Accordingly, we speculate that the mutp53- constitutes a useful approach for studying the role of IL-1 and the 35 driven sIL-1Ra suppression could constitute a novel autocrine antagonistic effects in different biological responses. Xenograft mechanism through which mutp53 might sustain the IL-1β- tumors were generated either with sh-p53 or sh-scr HT29 and κ MDA-MB468 cells and LPS delivered to all mice after tumor nodule induced NF B activation. formation. To explore antagonistic efficiency in vivo, Kineret was Overall, our results provide important insights into mutp53 GOF delivered in peritumoral sites in a subset of sh-scr tumor-bearing activities and might aid in the development of new prognostic mice. As shown in Figure 6a, either mutp53 depletion (sh-p53) or tools and therapeutic strategies against tumors carrying p53 Kineret delivery (sh-scr+Kineret) significantly inhibited tumor mutations. Thus we strongly suggest that the targeting of IL-1 growth when compared with control mice (sh-scr) with both activity might constitute a valid anti-tumor therapeutic interven- lines. We next explored the IL-8 gene expression in excised tumors tion, in combination with existing clinical treatments, in mutp53 to evaluate IL-1 effects in vivo. Significantly higher IL-8 mRNA was carrying tumors. found upon LPS delivery with respect to vehicle-treated mice (Figure 6b, HT29: lane 1 vs 2; MDA-MB468: lane 2 vs 1). By contrast, MATERIALS AND METHODS either Kineret delivery or mutp53 depletion abolished LPS- induced IL-8 mRNA (Figure 6b, HT29: lane 2 vs lanes 3, 4 and 5, Cell culture 6; MDA-MB468: lane 1 vs lanes 3, 4 and 5–7). Human lines HT29 (colon adenocarcinoma), MDA-MB468 (breast adeno- The overall results demonstrate that de-repressed sIL-1Ra, upon carcinoma), SKBR3 and MDA-MB231 (breast carcinoma), engineered with mutp53 depletion, inhibits the IL-1β cancer cell response in vitro an lentiviral-based TET-OFF inducible RNAi system carrying sh-RNA sequences specific for p53 (sh-p53) or control scrambled (sh-scr), are and in vivo. 6,16 To further support our studies, we evaluated IL-1Ra expression previously described. HEPG2 (hepatocellular carcinoma) and MCF7 (breast carcinoma) cells were purchased from ATCC (Manassas, VA, USA). in primary tumors public gene expression data repositories ’ fi (Oncomine). Analyses of data sets from specimens of primary All but MDA-MB231 and MCF7 cells, cultured in Dulbecco s modi ed Eagle’s medium (DMEM)-F12 1:1, were cultured in DMEM supplemented tumors with respect to normal tissues revealed a significantly with 10% fetal bovine serum (GIBCO-BRL, Grand Island, NY, USA), lower IL-1Ra and consistently higher IL-1β mRNA in the colon 36 37 L-glutamine (2 mM) and Penicillin/Streptomycin (100 U/ml) (Life Technolo- (Figure 7a) and breast (Figure 7b). Noteworthy, analyses based gies Inc., Eggenstein, Germany). HUVEC cells (Lonza, Basel, Switzerland), fi on p53 mutational status revealed a signi cantly lower IL-1Ra in cultured in endothelial cell basal medium (EBM-2) supplemented with colon tumors carrying p53 mutations in comparison to those with endothelial cell Bullet Kit (Lonza), were used between passages 4 and 5 for 37 wtp53 (Figure 7c). experiments.45 All cells were grown at 37 °C in a humidified atmosphere fi Taken together, our studies provide the rst evidence of a with 5% CO2. molecular and functional link between mutp53 and sIL-1Ra and provide new insights on the role of p53 mutants in oncogenic GOF Cytokine arrays activity in human cancer. To generate CMs, engineered sh-p53 and sh-scr HT29, MDA-MB468, SKBR3 or MDA-MB231 cells were cultured 72 h with plus/minus (± ) tetracycline- DISCUSSION derived DOX (1.0 μg/ml) (D9891, Sigma-Aldrich, Milan, Italy) to induce sh- RNA expression. Then, cells were seeded (6.0 ×105/60 mm dish) and 24 h It is widely accepted that tumor growth is the result of a complex later washed (3 × ) in phosphate-buffered saline (PBS) before refed with bidirectional interaction between cells that progressively acquire fresh DMEM ± DOX. Seventy-two hours later, CMs were collected and molecular alterations, and a transformed phenotype, and the analyzed with non-magnetic BIOPLEX-HU-27-PLEX kit (171A11127, Bio-Rad, surrounding host cells that may inhibit or promote autonomous Hercules, CA, USA) and Bio-Plex200 instrument, equipped with the Bio-Plex cancer cell growth and expansion. Chemokine-mediated chronic Manager Software 4.1 following the manufacturer’s guidelines.

Figure 6. Mutant p53 sustains LPS-induced HT29 and MDA-MB468 xenograft tumor growth through the sIL-1Ra suppression. (a) Xenograft tumors were generated with sh-scr and sh-p53 engineered HT29 and MDA-MB468 cancer cells in nude mice (CD1/SWISS). After tumor nodule formation, LPS (1.0 μg/mouse, intraperitoneal) and DOX (2.0 g/l, tap water) were delivered to all mice. To assess sIL-1Ra effects in vivo, Kineret was delivered in peritumoral regions to a subgroup of sh-scr tumor-bearing mice (sh-scr+kineret). Tumor growth was followed by caliper measurements twice a week. Representative data of two independent experiments with similar results are reported. (b) Xenograft tumors generated in panel (a) were excised, and RNA was isolated. Lower panel: Semi-quantitative PCRs were performed to assess mutp53 depletion in vivo. Upper panel: qPCRs were performed to analyze the human IL-8 gene expression along with treatments in vivo. Relative IL-8 mRNA levels were normalized with housekeeping gene human beta-actin and quantiﬁed with respect to control tumors (sh-scr) set to 1.0. Results are reported as means and s.d.s. of two independent analyses in triplicate.

Semi-quantitative and quantitative reverse transcriptase–PCR -3′, REV5′-ATGATGGAGTTGAAGGTAGTTTCGT-3′). All reactions were per- TRIzol (15596-026; Invitrogen, Monza, Italy) extracted RNAs from cells or formed in triplicate in a final volume of 20 μl. Dissociation curves were run tumors were retro-transcribed with Moloney-Murine-Leukemia virus to confirm that single products were amplified in each reaction. QPCR data reverse-transcriptase (M-MLV-RT, Invitrogen) following the manufacturer’s were analyzed using the 2 − ΔCt method: 2 − ΔCt = 2 − (ΔCt target − ΔCt instruction. reference) where Ct target and Ct reference (hBeta-actin) are mean For semiquantitative PCR, cDNAs were amplified by Hot-Master Taq threshold cycles of PCR done in triplicates on the same cDNA samples. The (5PRIME) with: hGAPDH (human glyceraldehyde 3-phosphate dehydro- relative mRNA levels with respect to control samples (set to 1.0) were genase; FOR5′-ATGACATCAAGAACGTGGTG-3′, REV5′-CATACCAGGAAAT obtained by the ratio 2 − ΔCt sample/2 − ΔCt control sample. GAGCTTG-3′); and hp53 (FOR5′-GTCTGGGCTTCTTGCATTCT-3′, REV5′-A ATCAACCCACAGCTGCAC-3′). For qPCR, cDNAs were amplified with SYBR reagent in a real-time PCR Western blotting 16 machine (ABI 7900; Applied Biosystems, Foster City, CA, USA) with 50 cycles Cells, rinsed (2 × ) with ice-cold PBS, were lysed in 1 × RIPA buffer of two-step amplification with the following primers: hsIL1RA (FOR5′- supplemented with protease/phosphatase inhibitors cocktail (Sigma- TTCCTGTTCCATTCAGAGACGAT-3′, REV5′-CCAGATTCTGAAGGCTTGCAT-3′); Aldrich). Lysates (30μg/lane) were resolved on 10% or 13% SDS–PAGE hIL-8 (FOR5′-CTCTGTCTGGACCCCAAGGA-3′, REV5′-TGAATTCTCAGCCC (sodium dodecyl sulfate–polyacrylamide gel electrophoresis), and filters TCTTCAAAA-3′); hMAFF (FOR5′-TGCCCAGGTCCCATTTCTC-3′, REV5′-GGCCC were immuno-reacted with the following antibodies: mouse anti-p53 ACGAAGGGAATGT-3′); and hBeta-actin (FOR5′-GCTGCCCTGAGGCACTCTT (DOI),6 rabbit anti-MAFF (M8194, Sigma-Aldrich), mouse anti-actin (Ab-1,

Figure 7. Public gene expression data repositories (Oncomine) of mammary and colon primary tumors. The box plots revealed the relative mRNA levels of sIL-1Ra and IL-1β in primary tumors of colon (a) and breast (b) with respect to normal tissue. (c) The box plots indicates the IL-1Ra mRNA levels in primary colon tumors carrying either wild-type or mutant p53.

Calbiochem, Billerica, MA, USA), rabbit anti-hIL-1R1 (EP409Y, Epitomic, supplemented with loading dye (Tris/HCl pH6.8, Glycerol and Bromophe- Burlingame, CA, USA); and rabbit anti-p21 (sc-397, Santa Cruz Biotechnol- nol-Blue) and resolved by 13% SDS–PAGE. Filters were immuno-reacted ogy, Dallas, TX, USA). with rabbit anti-hsIL-1Ra (ab2573, Abcam, Cambridge, UK). To detect sIL-1Ra production, CMs, concentrated with Centricon-10 Secondary HRP-conjugated anti-mouse (Santa Cruz Biotechnology) or (4205-Amicon) following the manufacturer’s instruction, were anti-rabbit (Calbiochem) antibodies and ECL kit (Amersham Biosciences,

Oncogene (2015) 2493 – 2504 © 2015 Macmillan Publishers Limited Mutp53 promotes inflammation by sIL-1Ra repression V Ubertini et al 2503 Glattbrugg, Switzerland) were used to detect immuno-reactions. Images AGCTTTGCTGGATAGa-3′), were annealed and cloned into MluI/ClaI (Roche were acquired with the EPSON Expression 10000 XL scanner (Epson, Long Applied Science, Basilea, Switzerland) digested pLV-THM vector generating Beach, CA, USA), and densitometry was performed with the ImageJ pLV-THsh/MAFF. Lentivirus was produced and titered as described.6 For software (NIH, Bethesda, MD, USA). constitutive RNAi, cells were transduced at required MOI with polybrene (8.0 μg/ml) (Sigma-Aldrich, H-9268). After 16 h, cells were washed and Transfection and luciferase reporter assay replenished with fresh medium and cultured until processing. 16 Stealth RNAi si-RNA duplex oligo ribonucleotides (Invitrogen Life Early transduced sh-scr and sh-p53 HT29 and MDA-MB468 cells were fi seeded (5 × 104cells/6 well plate) and 24 h later transfected with pRA-1680- Technologies) MAFF-speci c or negative control were transduced with ’ Luc17 (0.8 μg/well) and pRSVβ-GAL16 (0.2 μg/well) vectors with lipofecta- INTERFERin (Polyplus transfection) following the manufacture s guidelines. mine/plus reagents (Invitrogen) following the manufacturer’s guidelines. Cell were processed 48 h later. Cells were processed 48 h later. MCF7 cells were seeded (2 × 105/60 mm dish) and transfected with pRa-1680-Luc (8 μg/dish) and pRSVβ-GAL (2 μg/ dish) by calcium phosphate procedure. After 24 h, cells refed with fresh IL-8 gene expression medium were either treated with Nut-3 (10 μM, 10004372 Cayman) or Sh-scr and sh-p53 HT29 and MDA-MB468 cells, cultured 72 h with DOX 5 DMSO and collected 48 h later. HEPG2 cells were seeded (2 × 105/60 mm (1 μg/ml), were seeded (2.5 × 10 cells/six-well plate) and 24 h later washed dish) and co-transfected with pRA-1680-Luc (7 μg/dish) and pRSVβ-GAL 3 × ) in PBS before refed with DMEM+DOX. Cells were cultured 72 h for (2 μg/dish) along with either empty, wtp53, p53R273H or p53R175H medium conditioning and then incubated (1 h) with either recombinant encoding vectors (1 μg/dish). Twenty-four hours after medium human IL-1β (rhIL-1β; 400-002; RELIATech GmbH) or vehicle solution (PBS replacement, cells were treated with LPS (30 ng/ml; Escherichia Coli 055: +0.1% bovine serum albumin). For neutralization assays, before rhIL-1β B5, Sigma-Aldrich) and processed 48 h posttransfection. Luciferase and delivery sh-scr cells were pretreated (2 h) with recombinant non- 16 β-galactosidase assays were performed as reported earlier. glycosylated human sIL-1Ra (0.2 and 0.5 μg/ml; anakinra, Kineret, SOBI, Stockholm, Sweden). After incubations, IL-8 gene expression was verified ChIP assays by qPCR. Cells upon treatments were cross-linked as reported.16 Sonicated chromatins were incubated with the following antibodies: rabbit Endothelial cell injury anti-p53 (6 μl/reaction; ab-7 no. PC35, Oncogene, Billerica, MA, USA); μ Sh-scr and sh-p53 HT29 and MDA-MB468 cells, cultured 72 h with DOX, rabbit anti-MAFF (8 l/reaction; M8194, Sigma-Aldrich); goat anti-GFI-1 6 (4 μg/reaction; sc-8558, Santa Cruz Biotechnology); goat anti-HRT1 (4 μg/ were seeded (1.5 × 10 cells/100 mm dish) and 24 h later washed (×3) in reaction; sc-16424, Santa Cruz Biotechnology); anti-PAN-H3ac (10 μl/ PBS before refed with DMEM+DOX. After 72 h, all cells were incubated (6 h) reaction; no. 06-599, Millipore, Billerica, MA, USA); anti-H3K9ac with either rhIL-1β or vehicle solution, along with Kineret pretreatment as (4 μl/reaction; no. 07-352, Upstate, Billerica, MA, USA); anti-H3K14ac (4 μl/ described above. After incubations, CMs were collected and delivered (1:10 reaction; no. 07-353, Upstate); and anti-H3K4me3 (8 μl/reaction; no. 07-473, diluted in serum-free EBM-2) to 24-h cultured HUVEC cells (1.5 × 105cells/ Upstate). 60 mm dish). Cell viability was assessed 24 h later by trypan-blue. ChIP-enriched DNA was determined with semi-quantitative and qPCR. Semi-quantitative PCRs were performed with Hot-Master Taq (5PRIME) and the following primers: hsIL-1Ra-I (FOR5′-CTGGGATTACAGGCACATGC Bromodeoxyuridine incorporation assays -3′; and REV5′-TGTCTCCTTGGCCCTCAAAG-3′). Sh-scr and sh-p53 HT29 and MDA-MB468 cells, cultured 72 h with DOX, QPCR were performed with SYBR reagent in a real-time PCR machine were seeded (1 × 105cells/35 mm dish) and 96 h later incubated (6 h) with (ABI 7900; Applied Biosystems) with 50 cycles of two-step amplification either rhIL-1β or vehicle solution, along with Kineret pretreatments as ′ ′ and the following primers: hsIL-1Ra-I (FOR5 -CCAGCCCAGCCATCATTTT-3 ; described above. After BrdU (20 μM; Sigma-Aldrich) incubation (2 h) cells ′ ′ ′ REV5 -TTGGCCCTCAAAGGAAGACA-3 ); hsIL-1Ra-II (FOR5 -GGGTGGCACA were processed as described.4 The percentage of BrdU-positive nuclei ′ ′ ′ AGGCAAGT-3 ; REV5 -AACTCAGCATTTGGACAGGAATG-3 ); hsIL-1Ra-III were estimated by counting 500 cells per field, five fields for each (FOR5′-CCAAGGCTGTCCATTTTTCAA-3′; REV5′-GATAGGGCTCCCTGCACATG experimental condition. -3′); and hsIL-1Ra-IV (FOR5′-GCTTGGGTGAGTGACTATTTCTTTATAA-3′; REV5′-TCCATTCTGTGACTGCAGCAA-3′). All reactions were performed in triplicate in a final volume of 20 μl. Dissociation curves were run to confirm In vivo assay that single products were amplified in each reaction. QPCR data were 5 − ΔΔ Exponentially growing sh-scr and sh-p53 HT29 (5 × 10 cell/mouse) or analyzed using the 2 Ct method (fold enrichment relative to the 6 NoAb); this includes normalization for both background levels and input MDA-MB468 (5 × 10 cells/mouse) cells were either injected (8 mice/group) – chromatin: 2 − ΔΔCt = 2 − (ΔCtIP − ΔCtnoAb), where ΔCtIP = CtIP − Ctinput subcutaneously in 45-days-old (20 23 gr) female nude mice (CD1/Swiss, and ΔCtnoAb = CtnoAb-Ctinput. CtIP and CtnoAb are mean threshold Charles River, Lecco, Italy). After tumor appearance, all mice received DOX 6 cycles of PCR done in triplicates on DNA samples immunoprecipitated with (2.0 g/l) as reported. LPS (1 μg/mouse, intraperitoneal) was delivered in specific antibody and control (noAb) defined as 1.0.16 physiological water to all mice, whereas Kineret (0.2 μg/mouse in DMEM) was delivered in peri-tumoral region to a subgroup of sh-scr tumor-bearing mice 1 h after LPS administration. Treatments were performed three times Co-IP a week till the end of the experiment. Tumor growth was followed by Cells, washed (2 × ) with cold PBS, were collected with prechilled lysis caliper measurements twice a week and tumor volumes were (TV) buffer (PBS pH8.3; 10 mM EDTA; 0.1% Tween20) supplemented with estimated by the formula: TV = a ×(b2)/2, where a and b are tumor length protease/phosphatase inhibitors and lysate on ice by passing cell and width, respectively. All the procedures involving animals and their care suspension through a 26-G needle several times. Lysates (1 mg/sample) were approved by the Ethical Committee of the Regina Elena Cancer were clarified (16 000 g, 10 min at 4 °C), precleared with Pierce Protein-G Agarose (20399, Thermo-Scientific, Waltham, MA, USA) and incubated Institute (CE/532/12) and were conformed to the relevant regulatory (overnight at +4 °C) with either rabbit anti-MAFF (4 μl/sample, M8194 standards in accordance with the Italian legislation. Sigma-Aldrich) or rabbit IgG (2729, Cell Signaling, Danver, MA, USA) μ antibodies. The day after, washed Protein G slurry (50 l/sample) was Statistical analysis added, and samples were incubated for 45 min at 4 °C. After washes (2 × ) in cold lysis buffer, samples were resolved in 10% SDS–PAGE, and filters All experiments were performed in triplicate. Numerical data are reported were incubated with p53 (DO1) and MAFF-specific antibodies. as means ± s.d.s. Significances were assessed by the Student’s t-test analyses. MAFF RNAi Human MAFF shRNA-specific sequence (RNAi-codex-portal/database)16 (5′-cgcgtccCTATCCAGCAAAGCTCTAAttcaagaga-TTAGAGCTTTGCTGGATAG CONFLICT OF INTEREST tttttggaaat-3′;5′-cgatttccaaaaaCTATCCAGCAAAGCTCTA-AtctcttgaaTTAG The authors declare no conflict of interest.

© 2015 Macmillan Publishers Limited Oncogene (2015) 2493 – 2504 Mutp53 promotes inflammation by sIL-1Ra repression V Ubertini et al 2504 ACKNOWLEDGEMENTS 23 Fischer A, Klattig J, Kneitz B, Diez H, Maier M, Holtmann B et al. Hey basic helix- We thank Dr S Bacchetti, Dr FL Graham and Dr G D’Orazi for critical reading of the loop-helix transcription factors are repressors of GATA4 and GATA6 and restrict manuscript; Dr C Gabay for providing pRa-1680-Luc vector; and Dr G Tolstonog for expression of the GATA target gene ANF in fetal hearts. Mol Cell Biol 2005; 25: ENCODE in silico analyses on the sIL-1Ra promoter. This work was supported with 8960–8970. grants from Associazione Italiana Ricerca sul Cancro (AICR) to GB (IG no. 8804), AG 24 Kimura T, Ivell R, Rust W, Mizumoto Y, Ogita K, Kusui C et al. Molecular cloning of a (MFAG no. 11752), PG (IG no. 13234), and AF (RC 3.5–2013). We thank the Proteomic human MAFF homologue, which specifically binds to the oxytocin receptor gene Facility for Complex Potein Muxture (CPM) Analisys at ISS, Rome. in term myometrium. Biochem Biophys Res Commun 1999; 264:86–92. 25 Blank V, Small MAF. proteins in mammalian gene control: mere dimerization partners or dynamic transcriptional regulators? J Mol Biol 2008; 376:913–925. REFERENCES 26 Elaraj DM, Weinreich DM, Varghese S, Puhlmann M, Hewitt SM, Carroll NM et al. The role of Interleukin 1 in growth and metastasis in human cancer xenograft. 1 Aas T, Børresen AL, Geisler S, Smith-Sørensen B, Johnsen H, Varhaug JE et al. Clin Cancer Res 2006; 12: 1088–1096. Specific P53 mutations are associated with de novo resistance to DOXorubicin in 27 Puhlmann M, Weinreich DM, Farma JM, Carroll NM, Turner EM, Alexander Jr HR. breast cancer patients. Nature Med 1996; 2:811–814. Interleukin-1beta induced vascular permeability is dependent on induction of 2 Lang GA, Iwakuma T, Suh YA, Liu G, Rao VA, Parant JM et al. Gain of function of a endothelial tissue factor (TF) activity. J Transl Med 2005; 3:37. p53 hot spot mutation in a mouse model of Li-Fraumeni syndrome. Cell 2004; 28 Streicher KL, Willmarth NE, Garcia J, Boerner JL, Dewey TG, Ethier SP. Activation of 199:861–872. a nuclear factor kappaB/interleukin-1 positive feedback loop by amphiregulin in 3 Olive KP, Tuveson DA, Ruhe ZC, Yin B, Willis NA, Bronson RT et al. Mutp53 gain of human breast cancer cells. Mol Cancer Res 2007; 5:847–861. function in two mouse models of Li-Fraumeni syndrome. Cell 2004; 119:847–860. 29 Wu S, Meeker WA, Wiener JR, Berchuck A, Bast Jr RC, Boyer CM. Transfection of 4 Bossi G, Lapi E, Strano S, Rinaldo C, Blandino G, Sacchi A. Mutp53 gain of function: ovarian cancer cells with tumor necrosis factor-alpha (TNF-alpha) antisense mRNA reduction of tumor malignancy of human cancer cell lines through abrogation of mutp53 expression. Oncogene 2006; 25: 304–309. abolishes the proliferative response to interleukin-1 (IL-1) but not TNF-alpha. 53 – 5 Weisz L, Zalcenstein A, Stambolsky P, Cohen Y, Goldfinger N, Oren M, Rotter V. Gynecol Oncol 1994; :59 63. β Transactivation of the EGR1 gene contributes to mutp53 gain of function. Cancer 30 Li Y, Wang L, Pappan L, Galliher-Beckley A, Shi J. IL-1 promotes stemness Res 2004; 64: 8318–8327. and invasiveness of colon cancer cells through Zeb1 activation. Mol Cancer 2012; 11 6 Bossi G, Marampon F, Maor-Aloni R, Zani B, Rotter V, Oren M et al. Conditional :87. RNA interference in vivo to study mutp53 oncogenic gain of function on tumor 31 Germano G, Allavena P, Mantovani A. Cytokines as a key component of cancer- fl 43 – malignancy. Cell Cycle 2008; 7:1870–1879. related in ammation. Cytokine 2008; : 374 379. 7 Smith Jr MF, Eidlen D, Brewer MT, Eisenberg SP, Arend WP, Gutierrez-Hartmann A. 32 Lewis AM, Varghese S, Xu H, Alexander HR. Interleukin-1 and cancer progression: Human IL-1 receptor antagonist promoter. Cell type-specific activity and the emerging role of inteleukin-1 receptor antagonist as a novel therapeutic 4 identification of regulatory regions. J Immunol 1992; 149: 2000–2007. agent in cancer treatment. J Transl Med 2006; : 48. fl 8 Jenkins JK, Drong RF, Shuck ME, Bienkowski MJ, Slightom JL, Arend WP et al. 33 Saijo Y, Tanaka M, Miki M, Usui K, Suzuki T, Maemondo M et al. Proin ammatory Intracellular IL-1 receptor antagonist promoter: cell type-specific and inducible cytokine IL-1h promotes tumor growth of Lewis lung carcinoma by induction of regulatory regions. J Immunol 1997; 158: 748–755. angiogenic factors: in vivo analysis of tumor-stromal interaction. J Immunol 2002; 169 – 9 Dunn E, Sims JE, Nicklin MJ, O’Neill LA. Annotating genes with potential roles in :469 475. the immune system: six new members of the IL-1 family. Trends Immunol 2001; 34 Ulich TR, Guo K, Yin S, del Castillo J, Yi ES, Thompson RC et al. Endotoxin-induced Ÿ 22:533–536. cytokine gene expression in vivo. IV. Expression of interleukin-la/Ã and 10 Schreuder H, Tardif C, Trump-Kallmeyer S, Soffientini A, Sarubbi E, Akeson A et al. interleukin-1 receptor antagonist mRNA during endotoxemia and during fl 141 – A new cytokine-receptor binding mode revealed by the crystal structure of the endotoxin-initiated local acute in ammation. Am J Pathol 1992; :61 68. IL-1 receptor with an antagonist. Nature 1997; 386:194–200. 35 Chirivi RG, Garofalo A, Padura IM, Mantovani A, Giavazzi R. Interleukin 1 receptor 11 Apte RN, Voronov E. Is interleukin-1 a good or bad 'guy' in tumor immunobiology antagonist inhibits the augmentation of metastasis induced by interleukin 1 or and immunotherapy? Immunol Rev 2008; 222:222–241. lipopolysaccharide in a human melanoma/nude mouse system. Cancer Res 1993; 12 Zitvogel L, Kepp O, Galluzzi L, Kroemer G. ‘Inflammosome in carcinogenesis and 53: 5051–5054. anticancer immune responses’. Nat Immunol 2012; 13:343–351. 36 Skrzypczak M, Goryca K, Rubel T, Paziewska A, Mikula M, Jarosz D et al. 13 Mantovani A, Allavena P, Sica A, Balkwill F. Cancer-related inflammation. Nature Modeling oncogenic signaling in colon tumors by multidirectional analyses of 2008; 454: 436–444. microarray data directed for maximization of analytical reliability. PLoS ONE 2010; 14 Dinarello CA. Why not treat human cancer with Interleukin-1 blockade? Cancer 5: pii e13091. Metastasis Rev 2010; 29: 317–329. 37 Notterman DA, Alon U, Sierk AJ, Levine AJ. Transcriptional gene expression pro- 15 Dinarello CA, Simon A, van der Merr JWM. Treating inflammation by blocking files of colorectal adenoma, adenocarcinoma, and normal tissue examined by interleukin-1 in a broad spectrum of diseases. Nature Rev 2012; 11:633–652. oligonucleotide arrays. Cancer Res 2001; 61: 3124–3130. 16 Gurtner A, Starace G, Norelli G, Piaggio G, Sacchi A, Bossi G. Mutp53-induced 38 Muller PA, Vousden KH. p53 mutations in cancer. Nat Cell Biol 2013; 15:2–8. up-regulation of mitogen-activated protein kinase kinase 3 contributes to gain of 39 Freed-Pastor WA, Prives C. Mutp53: one name, many proteins. Genes Dev 2012; 26: function. J Biol Chem 2010; 285: 14160–14169. 1268–1286. 17 Gabay C, Smith Jr MF, Eidlen D, Arend WP. Interleukin-1 receptor antagonist 40 Lujambio A, Akkari L, Simon J, Grace D, Tschaharganeh DF, Bolden JE et al. Non- (IL-1Ra) is an acute-phase protein. J Clin Invest 1997; 99: 2930–2940. cell-autonomous tumor suppression by p53. Cell 2013; 153: 449–460. 18 Kim E, Deppert W. Transcriptional activities of mutp53: when mutations are more 41 Bar J, Feniger-Barish R, Lukashchuk N, Shaham H, Moskovits N, Goldfinger N et al. than a loss. J Cell Biochem 2004; 93:878–886. Cancer cells suppress p53 in adjacent fibroblasts. Oncogene 2009; 28: 19 Quante T, Otto B, Brázdová M, Kejnovská I, Deppert W, Tolstonog GV. Mutp53 is a 933–936. transcriptional co-factor that binds to G-rich regulatory regions of active genes 42 Addadi Y, Moskovits N, Granot D, Lozano G, Carmi Y, Apte RN et al. p53 status in and generates transcriptional plasticity. Cell Cycle 2012; 11: 3290–3303. stromal fibroblasts modulates tumor growth in an SDF1-dependent manner. 20 Scian MJ, Stagliano KE, Anderson MA, Hassan S, Bowman M, Miles MF et al. Cancer Res 2010; 70: 9650–9658. Tumor-derived p53 muts induce NF-kappaB2 gene expression. Mol Cell Biol 2005; 43 Weisz L, Damalas A, Liontos M, Karakaidos P, Fontemaggi G, Maor-Aloni R et al. 25: 10097–10110. Mutp53 enhances nuclear factor kappaB activation by tumor necrosis factor alpha 21 Di Agostino S, Strano S, Emiliozzi V, Zerbini V, Mottolese M, Sacchi A et al. Gain of in cancer cells. Cancer Res 2007; 67: 2396–2401. function of mutp53: the mutp53/NF-Y protein complex reveals an aberrant 44 Cooks T, Pateras IS, Tarcic O, Solomon H, Schetter AJ, Wilder S et al. Mutant p53 transcriptional mechanism of cell cycle regulation. Cancer Cell 2006; 10:191–202. prolongs NF-κB activation and promotes chronic inflammation and inflammation- 22 Zweidler-Mckay PA, Grimes HL, Flubacher MM, Tsichlis PN. Gfi-1 encodes a associated colorectal cancer. Cancer Cell 2013; 23: 634–646. nuclear zinc finger protein that binds DNA and functions as a transcriptional 45 D'Arcangelo D, Gaetano C, Capogrossi MC. Acidification prevents endothelial cell repressor. Mol Cell Biol 1996; 16:4024–4034. apoptosis by Axl activation. Circ Res 2002; 91:e4–e12.

Supplementary Information accompanies this paper on the Oncogene website (http://www.nature.com/onc)