Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 25 (2), 139-51

The PRKCI and SOX2 Oncogenes Are Coamplified and Cooperate to Activate Hedgehog Signaling in Lung Squamous Cell Carcinoma

Affiliations

The PRKCI and SOX2 Oncogenes Are Coamplified and Cooperate to Activate Hedgehog Signaling in Lung Squamous Cell Carcinoma

Verline Justilien et al. Cancer Cell.

Abstract

We report that two oncogenes coamplified on chromosome 3q26, PRKCI and SOX2, cooperate to drive a stem-like phenotype in lung squamous cell carcinoma (LSCC). Protein kinase Cι (PKCι) phosphorylates SOX2, a master transcriptional regulator of stemness, and recruits it to the promoter of Hedgehog (Hh) acyltransferase (HHAT) that catalyzes the rate-limiting step in Hh ligand production. PKCι-mediated SOX2 phosphorylation is required for HHAT promoter occupancy, HHAT expression, and maintenance of a stem-like phenotype. Primary LSCC tumors coordinately overexpress PKCι, SOX2, and HHAT and require PKCι-SOX2-HHAT signaling to maintain a stem-like phenotype. Thus, PKCι and SOX2 are genetically, biochemically, and functionally linked in LSCC, and together they drive tumorigenesis by establishing a cell-autonomous Hh signaling axis.

Figures

Figure 1
Figure 1. Lung cancer oncospheres exhibit stem-like characteristics
A) Phase contrast photomicrographs of H1703, ChagoK1 and H1299 parental adherent cells (top panels), oncospheres in low adherence culture (middle panels) and redifferentiated oncosphere cells after return to adherent culture (bottom panels). B) QPCR for putative stem cell markers expressed as fold of parental cells +/−SEM, n = 3; *p<0.05 versus NT adherent cells. Results are representative of five independent experiments. C) Photomicrographs showing clonal expansion of individual cells into oncospheres over a 15 day period. D) Anchorage-independent growth expressed as mean fold-change from parental cells +/− SEM. n = 5, * p<0.05 versus NT adherent cells. Results are representative of five independent experiments. E) Formation of lung orthotopic tumors in immunocompromised mice. 10,000 parental H1299 adherent or oncosphere cells were implanted into the lungs of immune-deficient mice. Lateral and dorsal views of bioluminescent images of a tumor-bearing mouse (a) and corresponding bioluminescence image upon lung dissection (b). H&E of typical parental and oncosphere tumors (c). Oncospheres develop primary tumors at the site of injection (arrows) and multiple lesions to the ipsilateral and contralateral lobes of the lung (arrowheads); parental cells develop either a single, small tumor at the site of injection (shown) or no identifiable tumor. 40X magnification reveals the similar morphology of oncosphere- and parental cell-derived tumors (d). Incidence of tumor formation in limiting dilutions of H1299 oncosphere and parental cells (e). Quantification of tumor growth by IVIS (f); Mean +/−SEM. *p<0.05 n=5, *p<0.05 vs. parental cells. See also Figure S1.
Figure 2
Figure 2. PKCι is required for maintenance of a stem-like phenotype in oncosphere cells
A) Effect of RNAi-mediated PKCι knockdown (immunoblot) on anchorage-independent growth of H1703, H1299 and ChagoK1 oncospheres. Results are expressed relative to NT RNAi control cells +/− SEM, n = 5; **p<0.001. B) Oncosphere size expressed as mean diameter in μm +/−SEM. n=98 (H1703), 138 (H1299) and 44 (ChagoK1). C) QPCR for HHAT, GLI1, ADRBK1 and CDK19. Results are expressed as fold NT parental cells +/−SEM, n=3; *p<0.05 compared to NT parental. Results are representative of three independent experiments. D) Effect of SMO inhibitor LDE225 on H1703, H1299 and ChagoK1 oncosphere (onco.) and parental cell (par.) proliferation. Results expressed as % DMSO control +/−SEM; n=6. Results are representative of three independent experiments. See also Figure S2 and Tables S1, S2 and S3.
Figure 3
Figure 3. Oncosphere growth requires a PKCι-dependent Hh signaling axis
A and B) RNAi-mediated knockdown of HHAT and GLI1 in H1299 oncospheres, respectively, expressed as fold of NT RNAi control +/− SEM n=3. *p<0.0005 compared to NT control. C) Effect of HHAT or GLI1 RNAi on clonal expansion expressed as oncosphere diameter in μm +/− SEM, n>15 per RNAi sample; *p<1.0×10−06. Data are representative of three independent experiments. D) Soft agar growth expressed relative to NT RNAi control cells +/− SEM, n = 5; * p<0.003. Data are representative of three independent experiments. E) NT, PKCι, HHAT and GLI1 oncospheres were treated with the indicated amounts of Hh-Ag1.5, a selective Hh agonist, and cell proliferation assessed by MTT at 5 days. Results expressed as % of NT DMSO control +/− SEM, n = 6. F) Detection of palmitoylated SHH in oncospheres. Parental (P) and oncosphere (O) NT, PKCι and HHAT RNAi cultures expressing FLAG-SHH were metabolically labeled with w-alkynyl-palmitate (C16) and palmitoylated SHH (palm-SHH) and total SHH were detected as described in Experimental Procedures. G) Oncosphere growth as lung orthotopic tumors. NT parental and NT, PKCι, HHAT, and GLI1 RNAi oncosphere tumor growth was monitored by bioluminescence detected by IVIS imaging. Data are presented as total flux in photons per second ± SEM; n = 10 per group except for PKCι where n=9; *p<0.05 significantly different than NT RNAi oncosphere tumors. See also Figure S3.
Figure 4
Figure 4. PKCι regulates HHAT expression through control of SOX2 occupancy of the HHAT promoter
A) RNAi-mediated knockdown of SOX2 in H1299 oncospheres. Results expressed as % NT control +/−SEM; n=3, *p<0.05. B) Clonal expansion expressed as oncosphere diameter in μm +/− SEM; n> 22 per RNAi sample, p< 3.0 ×10−14 and are representative of three independent experiments. C) Anchorage-independent growth relative to NT RNAi control cells +/− SEM, n = 5; *p<5.0×10−9. Results are representative of three independent experiments. D) Chromatin immunoprecipitation (ChIP) analysis to assess SOX2 occupancy of the HHAT promoter. Schematic depicts the HHAT promoter region; the position of ChIP probes used are indicated (A and B); Consensus SOX2 binding sites are indicated by vertical slashes. Data presented as % of input; n=3 +/−SEM. *p<0.00002; **p<0.00003. Data are representative of two independent experiments. E) and F) Effect of SOX2 RNAi on HHAT expression in oncosphere cells. Data are expressed relative to NT RNAi control cells +/− SEM; *p<0.0007 and are representative of three independent experiments. G) PRKCI and SOX2 amplification (upper panel) and overexpression (lower panel) in primary LSCC tumors. Red bars, tumors with amplification (upper panel) or overexpression (lower panel); blue bars, tumors with decreased expression; gray bars, tumors with no change in gene copy number (upper panel) or expression (lower panel). H) Analysis of PRKCI, SOX2, HHAT and GLI1 expression in primary LSCC tumors. Primary LSCC tumors were force ranked on PRKCI expression and grouped into top and bottom tertiles corresponding to high and low PRKCI expression, respectively. Box plots denote the expression of PRKCI, SOX2, HHAT and GLI1 in tumors expressing low PRKCI (low) and high PRKCI (high). Bars represent the median, boxes denote the 25 and 75% intervals; whiskers represent the 90% confidence intervals. See also Figure S4, and Tables S4 and S5.
Figure 5
Figure 5. Primary LSCC cells require PKCι, SOX2 and HHAT for oncosphere formation and proliferation
A) Phase contrast photomicrographs of oncospheres from three surgically-resected primary LSCC tumors. B) QPCR analysis of PKCι, SOX2, HHAT and GLI1 expressed as % NT RNAi control +/− SEM, n=3, *p<0.02. C) Proliferation of oncospheres by MTT assay expressed as % NT RNAi control +/− SEM; n=3, *p<4.0×10−07. D) Clonal expansion of oncospheres. Results expressed as % NT RNAi control +/− SEM; n=3, *p<0.0002. E) Photomicrographs of NT, PKCι, SOX2 and HHAT KD oncospheres. Arrows show membrane blebbing indicative of cell death. Results are representative of three independent experiments.
Figure 6
Figure 6. PKCι directly phosphorylates SOX2 at a unique site, T118, which is required for SOX2 function
A) Recombinant human SOX2 was incubated in kinase reaction buffer containing 32P-ATP in the absence or presence of recombinant PKCι. Phosphorylated SOX2 was detected by autoradiography and total SOX2 was detected by immunoblot analysis. B) Schematic of SOX2 protein structure; HMG=high mobility group domain, TAD=transactivation domain, NLS=nuclear localization sequence. Mass Spectrometric analysis revealed a single PKCι-mediated phosphorylation site on SOX2 at T118 which conforms to a consensus atypical PKC phosphorylation site motif (insert). C) Immunoblot analysis of H1299 oncopsheres stably transduced with NT or SOX2 RNAi, and then stably transfected with either empty vector (V) or vector encoding WT-SOX2 (WT), T118A-SOX2 (T118A), or T118D-SOX2 (T118D) mutants. D) QPCR analysis for HHAT and GLI1 expressed relative to NT RNAi control cells +/− SEM; n=3, *p<0.03. E) Clonal expansion expressed relative to NT RNAi control cells +/− SEM; n>20 per cell type, *p<1.0×10−08. F) Soft agar growth expressed relative to NT RNAi control cells +/− SEM; n=5, *p< 4.0×10−05 G) Cellular localization and promoter occupancy of T118-SOX2 phosphorylation mutants. Immunoblot analysis of cytoplasmic and nuclear fractions for FLAG, Lamins A/C and MEK1 (immunoblots). Lamins A/C served as a marker of nuclei and MEK1 as a marker of cytoplasm. HHAT promoter occupancy (bar graph, lower panel) expressed as fold of IgG control +/− SEM; n=3; *p< 0.001. Results in D-G are representative of three independent experiments. See also Figure S5.

Comment in

  • "Atypical" Regulation of Hedgehog-dependent Cancers
    SX Atwood et al. Cancer Cell 25 (2), 133-4. PMID 24525228.
    Growing evidence indicates targeting PKCι may be effective in treating Hedgehog-dependent cancers. In this issue of Cancer Cell, Justilien and colleagues present the surp …

Similar articles

See all similar articles

Cited by 117 PubMed Central articles

See all "Cited by" articles

Publication types

MeSH terms

Associated data

Feedback