Protein Kinase Cι Drives a NOTCH3-dependent Stem-like Phenotype in Mutant KRAS Lung Adenocarcinoma

Abstract

We report that the protein kinase Cι (PKCι) oncogene controls expression of NOTCH3, a key driver of stemness, in KRAS-mediated lung adenocarcinoma (LADC). PKCι activates NOTCH3 expression by phosphorylating the ELF3 transcription factor and driving ELF3 occupancy on the NOTCH3 promoter. PKCι-ELF3-NOTCH3 signaling controls the tumor-initiating cell phenotype by regulating asymmetric cell division, a process necessary for tumor initiation and maintenance. Primary LADC tumors exhibit PKCι-ELF3-NOTCH3 signaling, and combined pharmacologic blockade of PKCι and NOTCH synergistically inhibits tumorigenic behavior in vitro and LADC growth in vivo demonstrating the therapeutic potential of PKCι-ELF3-NOTCH3 signal inhibition to more effectively treat KRAS LADC.

Keywords: ELF3; KRAS-driven lung adenocarcinoma; NOTCH signaling; protein kinase Cι; therapeutic intervention.

Figure 1

LADC oncospheres exhibit tumor initiating cell properties. (A) Micrographs of parental cells (top, parental), oncospheres in low adherence (middle, oncosph) and re-differentiated oncosphere cells returned to adherent culture (bottom, Re. oncosph). Scale bars 50 μm. (B) Soft agar growth expressed as fold parental ± SEM, n=6, *p≤0.05 compared to parental, **p≤0.05 compared to oncospheres. (C) Single cells were assessed for clonal expansion. Plotted as % expanded +/−SEM; Cells expanded/total cells scored: A549: Parental (11/91), oncosph. (71/97), re. oncosph. (10/87); H358: parental (7/81), oncosph. (75/95), re. oncosph. (9/86); H23: parental (9/81), oncosph. (61/87), re. oncosph. (12/86). *p<0.0001 compared to parental; **p<0.0001 compared to oncospheres using Fisher’s exact test. Scale bars 50 μm. (D) Extreme Limited Dilution Analysis (ELDA) of A549 parental (black) and oncosphere (red) cells for orthotopic tumor engraftment. Results plotted as log fraction not responding (no engraftment) vs. cells inoculated. (E) Bioluminescent images of representative tumor bearing-mice at weeks 1 and 5 post-injection (a–d) and (F) corresponding images of lung tissue upon dissection (e, f). H&E staining reveals similar histology of oncosphere- and parental cell-derived tumors (g, h). Scale bars 50 μm. See also Figure S1.

Figure 2

Role of PKCι in the LADC TIC phenotype. (A) QPCR of parental and oncosphere cells for PKCι and MMP10. Results expressed as fold parental +/−SEM. n=3. *p<0.05 compared to parental. (B) PKCι knock down (KD) in A549 oncospheres using two RNAi constructs. Immunoblot (top), and QPCR (bottom). QPCR results expressed as fold parental cells +/−SEM, n=3. *p≤ 0.05 compared to parental. (C) Representative micrographs showing effect of PKCι KD on oncosphere growth. Scale bars 50 μm. (D) Oncosphere size expressed as mean diameter (μm) +/−SEM. Oncospheres assessed: A549: NT (63), KD1 (63), KD2 (58); H358: NT (44), KD1 (40), KD2 (39); H23: NT (41), KD1 (33), KD2 (37). *p≤0.05. (E) Single oncosphere cells were assessed for clonal expansion plotted as % expanded +/−SEM. Cells expanded/total cells: A549: NT (55/63), KD1 (7/63), KD2 (7/58); H358: NT (37/44), KD1 (5/40), KD2 (4/39); H23: NT (35/41), KD1 (4/33), KD2 (4/37). *p≤0.0001 compared to NT by Fisher’s exact test. (F) Soft agar growth expressed as colony number +/−SEM, n=6. *p≤ 0.05 compared to NT. (G) Immunoblot of A549 oncosphere cells stably transduced with NT or PKCι RNAi and stably transfected with either empty vector (EV) or vector encoding PKCι (+PKCι). (H) Representative micrographs of PKCι reconstitution on A549 oncosphere growth. Scale bars 50 μm. (I) Oncosphere size expressed as mean diameter (μm) +/−SEM. Number assessed: NT + EV (39), KD + EV (34), KD + PKCι (35). *p≤ 0.05 compared to NT+EV, **p≤0.05 compared to PKCι KD+EV. (J) Soft agar growth plotted as colony number +/−SEM, n=6. *p≤ 0.05 compared to NT+EV, **p≤0.05 compared to PKCι KD+EV. See also Figure S2.

Figure 3

Role of NOTCH3 in the LADC TIC phenotype. (A) QPCR for NOTCH3. Results expressed as fold parental +/−SEM, n=5. *p≤0.05 compared with parental; **p≤0.05 compared with NT oncospheres. (B) NOTCH3 KD in A549 oncospheres using two RNAi constructs. Immunoblot (top) and QPCR (bottom). QPCR results expressed as fold parental +/−SEM, n=3, *p<0.05. (C) QPCR for NOTCH1 and NOTCH2. Results expressed as fold parental +/−SEM, n=3. *p<0.05. (D) Growth of A549 oncospheres expressed as mean cell number +/−SEM, n=3, *p≤0.05. Scale bars 50 μm. (E) Effect of NOTCH3 KD on A549 oncosphere growth. Scale bars 50 μm. (F) A549 oncosphere size expressed as mean diameter (μm) +/−SEM. Number assessed: NT (40), KD1 (35), KD2 (33); *p≤0.05. (G) Single oncosphere cells were assessed for clonal expansion plotted as % expanded. Cells expanded/total cells scored: NT (35/40), KD1 (5/35), KD2 (4/33); *p<0.0001 by Fisher’s exact test. (H) A549 TIC viability plotted as fold NT +/− SEM, n=3, *p<0.05. (I) Soft agar growth plotted as colony number +/−SEM, n=6 and *p<0.05. (J) Representative micrographs showing NOTCH3 KD and reconstitution of A549 oncosphere growth. Scale bars 50 μm. (K) QPCR for NOTCH3. Results expressed as fold NT +/−SEM, n=5. *p<0.05 compared to NT; **p<0.05 compared to NOTCH3 KD. (L) Oncosphere size expressed as mean diameter (μm) +/−SEM. Number assessed: NT + EV (25), KD + EV (25), KD + PKCι (25). *p≤ 0.05 compared to NT+EV, **p≤0.05 compared to PKCι KD+EV. (M) Single oncosphere cells were assessed for clonal expansion plotted as % expanded +/−SEM. Cells expanded/total cells scored: NT (22/25), NOTCH3 KD (4/25), NOTCH3 KD+NOTCH3 (19/25). *p≤0.0001 compared to NT and **p<0.0001 compared to NOTCH3 KD by Fisher’s exact test. (N) Oncosphere viability (MTT reduction) plotted as %NT +/−SEM, n=3, *p<0.05 compared to NT; **p<0.05 compared to NOTCH3 KD. For (K, L, M and N) open bars, NT; black bars, Notch3 KD; gray bars, Notch3 KD+ Notch 3. See also Figure S3.

Figure 4

Effect of PKCι on ELF3 occupancy at the NOTCH3 promoter and NOTCH3 expression. (A) QPCR for ELF3 presented as fold parental +/−SEM, n=3. *p≤ 0.05. (B) Immunoblot (top) and QPCR (bottom) of ELF3 KD. QPCR expressed as fold NT +/−SEM, n=3. *p≤ 0.05 compared to NT. (C) Representative micrographs of ELF3 KD on oncosphere growth Scale bars 50 μm. (D) Oncosphere size expressed as mean diameter (μm) +/−SEM. Oncospheres assessed: A549: NT (37), KD (37); H358: NT (34), KD (34); H23: NT (34), KD (38). *p≤0.05 compared to NT. (E) Single oncosphere cells were assessed for clonal expansion plotted as %expanded. Cells expanded/total cells scored: A549: NT (33/37), KD (5/37); H358: NT (26/34), KD (4/34); H23: NT (26/34), KD (5/38). *p≤0.001 based on Fisher’s exact test. (F) TIC viability plotted as fold NT +/−SEM, n=3, *p≤0.05 compared to NT. (G) Soft agar growth plotted as colony number +/−SEM, n=6. *p<0.05 compared to NT. (H) ChIP analysis of ELF3 occupancy of the NOTCH3 promoter in A549 TICs. Inset depicts the NOTCH3 promoter; ChIP primer-probes used are indicated as A and B. Consensus ELF3 binding sites indicated by vertical slashes. Data presented as % Input +/−SEM, n=3. *p<0.05; data representative of two independent experiments. immunoglobulin G, (IgG). (I) QPCR of NOTCH3 expressed as fold NT +/−SEM, n=3. *p≤ 0.05 compared to NT. (J) NOTCH3 promoter reporter (inset) activity plotted as fold NT +/−SEM, n=5, *p<0.05 compared to NT. (K) NOTCH3 promoter occupancy. Data presented as % Input +/−SEM, n = 3; *p<0.05. (L) QPCR for NOTCH3 expressed as fold NT +/−SEM, n=3. *p<0.05 compared to NT, **p<0.05 compared to PKCι KD. See also Figure S4.

Figure 5

Role of PKCι-mediated ELF3 phosphorylation on NOTCH3 expression and oncosphere growth. (A) Recombinant human ELF3 incubated in the absence or presence of recombinant PKCι. ELF3 was subjected to mass spectrometric analysis. Schematic of ELF3 domain structure indicating the PKCι phosphorylation site at S68, and the S68A and S68D mutants generated for subsequent experiments (inset). (B) ELF3 expression (top) and NOTCH3 promoter reporter activity (bottom) assayed in NT or ELF3 KD cells stably transduced with empty vector (EV), or ELF3 KD cells transduced with WT-ELF3 (WT), S68A-ELF3 (S68A), or S68D-ELF3 (S68D). Activity is plotted as fold NT +/−SEM, n=5, *p<0.05 compared to NT+EV, **p<0.05 compared to ELF3-KD+EV. (C) ChIP analysis of ELF3 occupancy at the NOTCH3 promoter. Data presented as % input +/−SEM, n=3; *p<0.05 compared to EV. Data are representative of two independent experiments. (D) QPCR of NOTCH3 expressed as fold EV +/−SEM, n=3. *p<0.05 compared to EV. (E) Representative micrographs of immunofluorescence localization of exogenous WT-ELF3, S68A-ELF3 and S68D-ELF3 mutants in the nucleus (dotted line) and cytoplasm in ELF3-KD cells. Scale bars 10 μm. (F) Cells (300–350/ELF3 construct) were assessed for intracellular localization of ELF3 (nucleus vs. cytoplasm) as described in Experimental Procedures. (G) Representative micrographs showing the effect of ELF-WT and ELF3 mutants on oncosphere growth. Scale bars 50 μm. (H) Oncosphere size expressed as mean diameter (μm) +/−SEM. Oncospheres assessed: A549: NT+EV (27), KD+EV (29), KD+WT ELF3 (28), KD+S68A ELF3 (30), KD+S68D (29); H358: NT+EV (31), KD+EV (27), KD+WT ELF3 (29), KD+S68A ELF3 (28), KD+S68D ELF3 (30). *p<0.05 compared to NT, **p<0.05 compared to EV. See also Figure S5.

Figure 6

Effect of the PKCι-ELF3-NOTCH3 axis on asymmetric cell division and tumor initiation. (A) Immunofluorescence of A549 oncosphere cells for CD133 (red) and DAPI (blue) during interphase (a–c) and mitosis (d–i) demonstrating symmetric (d–f) and asymmetric (g–i) cell divisions. Scale bars 10 μm. (B) PKCι, ELF3 and NOTCH3 KD inhibit mitotic index in LADC oncosphere cells. Data expressed as mitotic index (% mitotic cells) +/−SEM. n=3. *p<0.05 compared to NT. (C) Effect of PKCι, ELF3 and NOTCH3 KD on asymmetric cell division. Results expressed as % asymmetric cell divisions (>300 mitoses evaluated/cell line). *p≤0.0001 compared to NT using Fisher exact test. (D) Bioluminescent images of representative tumor bearing-mice 5 weeks after injection of 50,000 NT, PKCι KD, ELF3 KD or NOTCH3 KD A549 oncosphere cells. (E) Tumor size expressed as bioluminescence flux +/−SEM, n=10. *p<0.05 compared to NT control. (F) Representative bioluminescence images of lung tumors ex vivo. (G) Kaplan-Meier survival analysis of mice injected with the indicated A549 oncosphere cells (50,000 cells). N=10. *p<0.05 compared to NT. (H) QPCR for ELF3, NOTCH3 and CD133. Results are expressed as fold parental +/−SEM, n=10, *p<0.05 compared to NT parental, **p<0.05 compared to injected NT oncosphere cells.

Figure 7

Effect of Auranofin (ANF) and γ-secretase inhibitor (GSI) on oncosphere formation and tumor growth. (A) Combination index analysis of ANF and GSI. (B) Representative micrographs of ANF and GSI effects on A549 and H358 oncosphere growth. Scale bars 50 μm. (C) Effect of ANF, GSI and the combination on oncosphere size expressed as mean diameter (μm) +/−SEM. Oncospheres assessed: A549: DMSO (54), GSI (45) ANF (49), GSI/ANF (55); H358: DMSO (57), GSI (49), ANF (51), GSI/ANF (55). *p<0.05. (D) Cell viability expressed as MTT reduction. n=3. *p<0.05. (E) Soft agar growth expressed as colonies +/−SEM, n=6. *p<0.05 compared to DMSO. **p<0.05 compared to ANF or GSI alone. (F) Tumor volume plotted +/−SEM. N=20. *p<0.05 compared to DMSO; **p<0.05 compared to ANF or GSI alone. (G) Tumor burden assessed at time of sacrifice (30 days). Results expressed as mean +/−95% confidence intervals. n=10. *p<0.05 compared to DMSO. **p<0.05 compared to ANF or GSI alone. (H) Association analysis for expression of the indicated genes in primary LADC tumors (see Experimental Procedures). See also Figure S6.