A microRNA-dependent program controls p53-independent survival and chemosensitivity in human and murine squamous cell carcinoma

Abstract

The p53 tumor suppressor, a central mediator of chemosensitivity in normal cells, is functionally inactivated in many human cancers. Therefore, a central challenge in human cancer therapy is the identification of pathways that control tumor cell survival and chemosensitivity in the absence of functional p53. The p53-related transcription factors p63 and p73 exhibit distinct functions—p73 mediates chemosensitivity while p63 promotes proliferation and cell survival—and are both overexpressed in squamous cell carcinomas (SCCs). However, how p63 and p73 interact functionally and govern the balance between prosurvival and proapoptotic programs in SCC remains elusive. Here, we identify a microRNA-dependent mechanism of p63/p73 crosstalk that regulates p53-independent survival of both human and murine SCC. We first discovered that a subset of p63-regulated microRNAs target p73 for inhibition. One of these, miR-193a-5p, expression of which was repressed by p63, was activated by proapoptotic p73 isoforms in both normal cells and tumor cells in vivo. Chemotherapy caused p63/p73-dependent induction of this microRNA, thereby limiting chemosensitivity due to microRNA-mediated feedback inhibition of p73. Importantly, inhibiting miR-193a interrupted this feedback and thereby suppressed tumor cell viability and induced dramatic chemosensitivity both in vitro and in vivo. Thus, we have identified a direct, microRNA-dependent regulatory circuit mediating inducible chemoresistance, whose inhibition may provide a new therapeutic opportunity in p53-deficient tumors.

Figure 1. p63-regulated miRs target p73.

(A) Knockdown of endogenous p63 RNA (top) and protein (bottom) by p63-directed or control (Ct) lentiviral shRNA in JHU-029 human SCC cells at 48 hours, in duplicate experiments (Exp). β-Tubulin (β-Tub) loading control. (B) Array analysis showing the fold change and direction of change for all miRs regulated at 1.5-fold or more in p63-ablated versus control samples shown in A. Circles show miRs predicted to target p73. (C) Validation of p63 repression of miRs targeting p73 by real-time quantitative RT-PCR (QRT-PCR) at 72 hours after lentiviral p63-directed or control shRNA expression in JHU-029 cells. (D) The p63-regulated miRs repress gene expression via the p73 3′ UTR. Cotransfection of the indicated miR mimics or control (scrambled) miR, together with the UTR-reporter or control reporter; results show relative luciferase units (RLU) normalized to the control miR. Note that repression correlates with the number of predicted seed-binding sequences (Sites) for each miR. (E) A miR-dependent mechanism for regulation of the p73 3′ UTR by p63. Lentiviral shRNA knockdown of Drosha followed by cotransfection of the UTR or control reporter, together with either a p63 shRNA or ΔNp63α cDNA or their respective controls in JHU-029 cells. RLU values expressed as p63 knockdown/control or ΔNp63α overexpression/control (OV/Ct). Above, immunoblot shows efficient Drosha knockdown. All error bars show SEM for triplicate measurements from representative experiments.

Figure 2. Endogenous p63 and p73 mediate miR-193a regulation by chemotherapy.

(A) Regulation of human miR-193a-5p (hereafter miR-193a) and murine miR-193* by endogenous p63, measured by QRT-PCR 48 hours after lentiviral p63 knockdown in JHU-029 cells (029), murine SCC cells (mSCC), human immortal primary keratinocytes (OKF6), and control A549 cells, which do not express p63. (B) Repression of endogenous miR-193a by p63 requires DNA binding. Retroviral expression of WT or non–DNA-binding ΔNp63α mutant (R304W) was followed by QRT-PCR. (C) Regulation of murine miR-193* by endogenous p63 in vivo. Homozygous p63^flox mice with or without expression of the K14-Cre/ER transgene (Tg) were treated with tamoxifen (Tam) or vehicle control for 5 days, and epidermal cells were collected 21 days later for RNA analysis by QRT-PCR. (D) Inverse correlation between p63 expression and miR-193a expression in human primary HNSCC specimens, assessed by QRT-PCR. df, degrees of freedom. (E) Rapid upregulation of endogenous PUMA (top) and miR-193a (bottom) following tetracycline-induced (Tet) TAp73 expression in JHU-029 cells. Middle panels show induced TAp73 expression by immunoblot. (F) Endogenous p73-dependent induction of miR-193* in murine SCC cells by cisplatin. Following preinfection with lentiviral TAp73 shRNA or control, cells were treated with cisplatin (Cis, 4 μM, 24 hours) or vehicle control (Veh) prior to RNA analysis. (G) Regulation of miR-193* is independent of p53. WT or p53^–/– E1A-immortalized MEFs were treated with doxorubicin (Dox) (0.2 μM, 12 hours) or vehicle control prior to RNA analysis. Bax and Noxa are induced in a p53-dependent manner, whereas miR-193* induction correlates only with TAp73. All error bars show SEM for triplicate measurements.

Figure 3. miR-193a is a direct transcriptional target of p63 and p73.

(A) Direct binding of p63 within a conserved CpG island upstream of miR-193a. Top: cross-species Vista alignment; peaks show degree of homology; arrows at top show position and orientation of miR-193a; black bars below indicate CpG islands; arrowheads show location of ChIP fragments. Bottom: ChIP results in JHU-029 cells for the indicated fragments. (B) ChIP in primary human foreskin keratinocytes (HFK) for the fragments indicated in A. (C) ChIP showing specific binding of WT ΔNp63α but not the DNA-binding–defective mutant to the conserved region B in JHU-029 cells. (D) ChIP showing binding of tetracycline-induced p73 to region B. Binding in the absence of tetracycline results from leaky p73 expression. (E) Transcriptional activation of a miR-193a regulatory region requires a conserved p63/p73 DNA-binding motif (site 1). Top: schematic of ChIP-binding region B with putative binding motifs, their sequences, and introduced point mutant (Mut) sequence indicated. Bottom: cotransfection of TAp73β with a luciferase reporter containing the indicated fragments. (F) Dose-dependent induction of the miR-193a reporter by cisplatin (Cis) (24 hours) requires the p63/p73 DNA-binding motif. JHU-029 cells were treated as indicated 12 hours after reporter transfection. (G) p63 represses p73-dependent activation of the miR-193a reporter. Cotransfection of the indicated reporter with vector, TAp73β (p73) and vector, or TAp73β and ΔNp63α (p63) in JHU-029 cells. All error bars show SEM for triplicate measurements.

Figure 4. Direct inhibition of p73 by miR-193a is opposed by p63.

(A) Endogenous p73 mRNA is inhibited 48 hours following transfection of JHU-029 cells with a miR-193a mimic compared with a scrambled control miR. (B) Endogenous p73 mRNA is derepressed 48 hours following transfection of JHU-029 cells with a miR-193a antagomir (anti-miR) compared with scrambled control (anti-Ct) antagomir. (C) Schematic of the p73 3′ UTR reporter construct showing putative miR-193a seed-binding sequences (WT seed BS) and nucleotide changes introduced in the mutant reporter (mutant seed BS). (D) UTR-dependent regulation of p73 protein requires miR-193a seed-binding sequences. Immunoblots of lysates following cotransfection of a TAp73β cDNA linked to the WT (left 2 panels) or mutant (mut, right panels) 3′ UTR, together with either a miR-193a mimic, a specific antagomir (anti-miR), or respective controls. Bar graphs below show densitometry. Blots are representative of triplicate experiments. (E) Endogenous p73 mRNA is regulated by p63. Retroviral ΔNp63α or vector control were expressed in JHU-029 cells followed by QRT-PCR for ΔNp63 or TAp73. (F) Endogenous p63-mediated regulation of p73 mRNA. Infection of JHU-029 cells with lentiviral p63 shRNA or scrambled control, followed by QRT-PCR for ΔNp63 or TAp73. (G) Regulation of p73 protein by p63. Retroviral ΔNp63α expression as in E was followed by IP/immunoblot (p73) or immunoblot (p63). Right, densitometry analysis for p73. Heavy chain (HC) loading control. (H) Endogenous p63-mediated regulation of p73 protein. Lentiviral p63 knockdown as in F, followed by analysis as in G. Error bars show SEM for triplicate measurements.

Figure 5. miR-193a regulates cell viability and chemosensitivity through p73.

(A) p73-dependent inhibition of viable cells following miR-193* antagomir (anti-miR) treatment. Murine SCC (mSCC) cells were infected with lentiviral p73 shRNA or control, then treated with a miR-193* antagomir or scrambled control followed by cell counts at 72 hours. (B) miR-193a abrogates cisplatin-induced growth inhibition. JHU-029 cells were pretreated with miR-193a mimic or scrambled control (miR Ct) for 24 hours followed by 1 hour cisplatin treatment, and cell counts were determined by XTT assay 72 hours later. (C) Enhancement of cisplatin sensitivity by the miR-193a antagomir is p73-dependent. JHU-029 cells were infected with lentiviral TAp73 shRNA or control, then treated for 24 hours with a miR-193a antagomir or scrambled control (anti–miR Ct) prior to 1 hour cisplatin treatment and cell counts 72 hours later. (D) Endogenous miR-193a controls cisplatin-mediated clonogenic suppression. JHU-029 cells were treated for 24 hours with a miR-193a antagomir (anti–miR-193a) or scrambled control (anti–miR Ct) prior to cisplatin treatment (0.5 μm, 1 hour), then plated at clonal density for colony counts. Scale bar: 1 cm. *34.7% repression; **64.5% repression. (E) Control of p73 activity by endogenous miR-193a. Cells treated as in D were harvested at 24 hours after cisplatin for analysis of the p73 target gene NOXA by QRT-PCR. Error bars show SEM.

Figure 6. In vivo control of tumor progression and chemosensitivity by miR-193a.

(A) Inverse correlation between miR-193a and p73 expression in primary human HNSCC specimens, assessed by QRT-PCR. (B) Murine SCCs resemble human SCC, assessed by histology (H&E) and p63 immunohistochemical staining. Original magnification, ×200. (C) Anti–miR-193* inhibits tumor growth in the absence of cisplatin, and it enhances cisplatin sensitivity. Primary disaggregated SCC was treated with miR-193* antagomir or scrambled control, then implanted into nude mice, followed by cisplatin treatment (5 mg/kg) or vehicle. n = 10 per arm. *P < 0.05 by 2-way repeated measures ANOVA. (D) Increased expression of activated caspase-3 (brown staining) in vivo following antagomir-mediated miR-193* inhibition compared with control antagomir treatment, both in the presence and absence of cisplatin treatment. Original magnification, ×200. The percentage of cleaved caspase-3–positive cells ± SEM is shown below each panel, based on 100 cells counted in 3 representative fields. (E) Representative mice following cisplatin treatment.

Figure 7. Model for miR-dependent mechanism mediating p63/p73 crosstalk, tumor cell survival, and chemosensitivity.

Top: in SCC, p63 is an inhibitor of p73 and a transcriptional repressor of miRs that target p73. The miR-193a/193* is a direct transcriptional target of both p63 and p73 and a direct feedback inhibitor of p73. Middle: cisplatin treatment induces p63 degradation and p73 activation, thereby inducing this miR, which then limits p73-dependent chemosensitivity through direct feedback inhibition. Bottom: this feedback loop is disrupted by miR inhibition, which increases p73 levels and enhances chemosensitivity.