NOTCH pathway inactivation reprograms stem-like oral cancer cells to JAK-STAT dependent state and provides the opportunity of synthetic lethality

Abstract

Background: We have recently provided the evidence of interconvertible cellular states, driving non-genetic heterogeneity among stem-like oral cancer cells (oral-SLCCs). Here, NOTCH pathway-activity status is explored as one of the possible mechanisms behind this stochastic plasticity.

Methods: Oral-SLCCs were enriched in 3D-spheroids. Constitutively-active and inactive status of NOTCH pathway was achieved by genetic or pharmacological approaches. RNA sequencing and real-time PCR was performed for gene expression studies. in vitro cytotoxicity assessments were performed by AlamarBlue assay and in vivo effects were studied by xenograft growth in zebrafish embryo.

Results: We have observed stochastic plasticity in oral-SLCCs, spontaneously maintaining both NOTCH-active and inactive states. While cisplatin refraction was associated with post-treatment adaptation to the active-state of NOTCH pathway, oral-SLCCs with inactive NOTCH pathway status showed aggressive tumor growth and poor prognosis. RNAseq analysis clearly suggested the upregulation of JAK-STAT pathway in NOTCH pathway-inactive subset. The 3D-spheroids with lower NOTCH-activity status displayed significantly higher sensitivity to JAK-selective drugs, Ruxolitinib or Tofacitinib or siRNA mediated downregulation of tested partners STAT3/4. Oral-SLCCs were programmed to adapt the inactive status of NOTCH pathway by exposing to γ-secretase inhibitors, LY411575 or RO4929097, followed by targeting with JAK-inhibitors, Ruxolitinib or Tofacitinib. This approach resulted in a very significant inhibition in viability of 3D-spheroids as well as xenograft initiation in Zebrafish embryos.

Conclusion: Study revealed for the first time that NOTCH pathway-inactive state exhibit activation of JAK-STAT pathways, as synthetic lethal pair. Therefore, co-inhibition of these pathway may serve as novel therapeutic strategy against aggressive oral cancer.

Keywords: Cellular plasticity; JAK-STAT; NOTCH; Oral cancer; Stemness; Synthetic lethal pair.

Graphical abstract

Fig. 1

Maintenance of stemness in both NOTCH pathway active and inactive oral-SLCCs. A–C. Changes in average sphere number (>60 μm) in LY and RO treatment compared to DMSO (0.01%) control in SCC029b, SCC032 and SCC070 and the lower panel of the bar graphs shows the representative bright field images from each treatment condition in the cell lines, scale bar=60 μm D,E. Relative normalized mRNA expression changes of NOTCH1, HES1, ALDH1A1, SOX2, SOX9, CK10 and CK14 in Control (0.01% DMSO) and LY, RO treated spheres of GBC02 and SCC070 spheres by qRT-PCR. F,G. Relative normalized mRNA expression changes of NOTCH1, HES1, ALDH1A1, SOX2, SOX9, CK10 and CK14 in pLenti, hICN, hHes1 and BHes1 spheres of GBC02 and SCC070 by qRT-PCR H. The dot plots represents the sphere diameter (μm) distribution in pLenti, hICN, hHes1 and BHes1 cells of GBC02 and SCC070 cell lines I. The bar graphs represent changes in average spheres numbers in pLenti, hICN, hHes1 and BHes1 cells of GBC02 and SCC070 cell lines J. Representative in-vivo zebrafish images, where the GFP regions shows the tumor formed from the injected GFP positive OSCC cells. K. The line graph shows the survivability of the zebrafish embryos injected with pLenti, hICN, hHes1 and BHes1. Number of dots represent the average values obtained from each biological samples with experiments performed with three technical repeats. Error bar is representing the standard deviation between the biological repeats. * indicates p value < 0.05; ** indicates p value < 0.01; *** indicates p value < 0.001; N.S= Non-significant.

Fig. 2

Plasticity between NOTCH pathway -active and -inactive status of oral-SLCCs. A. The bar graphs represent the categorized <70 μm, 70–100 μm and >100 μm of spheres into 1°, 2° and 3° (primary, secondary and tertiary) spheres from hHes1 and BHes1 clones of SCC070 cells. B. Relative normalized mRNA expression changes of ALDH1A1, SOX2, HES1, cMYC, CK10 and CK14 in <70 μm and >100 μm spheres from SCC070 by qRT-PCR, scale bar= 60 μm C. The bar graphs represent average sphere numbers in 2° (secondary) sphere culture from <70 μm and >100 μm, 1° (primary) spheres of SCC070 cells, scale bar= 60 μm D. The relative normalized mRNA expression of NOTCH1 and HES1, - in ALDH-low and -high cell populations from SCC070 and GBC02. Number of dots represent the average values obtained from each biological samples with experiments performed with three technical repeats. Error bar is representing the standard deviation between the biological repeats. * indicates p value < 0.05; ** indicates p value < 0.01; *** indicates p value < 0.001; N.S= Non-significant.

Fig. 3

Cisplatin tolerant state maintained by active NOTCH pathway status. A. Schematic diagram of cisplatin treatment of 1° spheres and, subsequent dissociation of primary spheres into single cells for 2° sphere formation. B. Representative 10X images of spheres in the respective control and cisplatin treatment conditions of SCC029B, SCC032, GBC02, and SCC070 cell lines C. The bar graphs shows relative viability of control and cisplatin treated spheres of SCC029B, SCC032, GBC02, and SCC070 D. Relative normalized mRNA expression changes of ALDH1A1, NOTCH1 and HES1 in control and cisplatin treated 1° spheres of SCC029B, SCC032, GBC02, and SCC070 and E. in 2° spheres generated from 1° control and cisplatin treated spheres of SCC070 by qRT-PCR F.i-ii. Line graph demonstrating the IC₅₀ value of cisplatin in LY and RO pretreated sphere culture in comparison to control of GBC02 and SCC070 cell lines, and G.i-ii. IC₅₀ value of cisplatin in spheres generated from NOTCH-active, hICN and NOTCH-inactive, BHes1 cells of GBC02 and SCC070. Number of dots represent the average values obtained from each biological samples with experiments performed with three technical repeats. Error bar is representing the standard deviation between the biological repeats. * indicates p value < 0.05; ** indicates p value < 0.01; *** indicates p value < 0.001; N.S= Non-significant.

Fig. 4

JAK-STAT pathway activation in NOTCH-inactive oral-SLCCs: A. Volcano plot of DEGs from RNAseq. B. The pathway enrichment analysis has shown significant upregulation of top 8 biological pathway significantly enriched in Notch low BHes1 spheres. C. Gene Set Enrichment Analysis (GSEA) of differentially expressed genes between hHes1 and BHes1 spheres has shown significant enrichment of inflammatory response genes, in NOTCH-inactive BHes1 cells D. Venn diagram overlap of NOTCH-active, NOTCH-inactive and JAK-STAT genes has shown significant co-expression JAK-STAT genes in NOTCH-inactive oral-SLCC E. Comparison of gene expressions of unique-upregulated oral-hybrid SLCC, which is more like progenitor in nature showed significant co-expression in NOTCH-inactive state compared to unique-upregulated genes in oral-SLCC, which is more like stem cells F. Western blots of JAK-STATs proteins in their original and phosphorylated state has shown upregulation of signaling activity in Notch inhibited spheres compared to the control G. The Kaplan-Meier survival plot of 5 upregulated gene signature from RNAseq of Notch inactive SLCCs using GEPIA2 showing the TCGA-HNSCC patient cohort having higher expression of this signature experienced poor prognosis.

Fig. 5

NOTCH-inactivated oral-SLCCs shows dependency on JAK-STAT pathway. A. i-ii. Representative 10X images of the spheres from hICN and BHes1 cells of SCC070 and GBC02, treated with siSTAT3, siSTAT4 and siControl was kept as control, scale bar= 275 μm B.i-ii. The bar graphs shows changes in viability of hICN and BHes1 cells of SCC070 and GBC02, treated with siSTAT3, siSTAT4 and siControl was kept as control C.i-iv. The dot plots represents the sphere diameter distribution and viability changes in Tofa and Ruxo. treated spheres of hICN and BHes1 cells from GBC02 (i-ii) and SCC070 (iii-iv) cell lines. Number of dots represent the average values obtained from each biological samples with experiments performed with three technical repeats. Error bar is representing the standard deviation between the biological repeats. * indicates p value < 0.05; ** indicates p value < 0.01; *** indicates p value < 0.001; N.S= Non-significant.

Fig. 6

NOTCH—HES and JAK-STAT pathways acts in synthetic lethal manner: A. i-iii. The bar graphs represent viability of spheres of GBC02, SCC029 and SCC070 cell lines, where Notch pathway was inhibited by LY and RO from the time of plating and JAK inhibitors, Tofa. and Ruxo. were added on 5th day of sphere formation and followed for another 48 h. B.i-iii. The bar graphs represent viability of spheres of GBC02, SCC029 and SCC070 cell lines, where Notch pathway and JAKs were inhibited in a combinatorial mode by LY, RO and by Tofa., Ruxo. On 5th day of sphere formation and followed for another 48 h C. Representative in-vivo zebrafish images, where the GFP regions showing the tumor formed from the injected, Ruxo. and Tofa. pretreated hICN and BHes1 cells of SCC070. D. The line graph shows the survivability of the zebrafish embryos injected with Ruxo and Tofa pretreated hICN and BHes1 cells of SCC070. Number of dots represent the average values obtained from each biological samples with experiments performed with three technical repeats. Error bar is representing the standard deviation between the biological repeats. *indicates p value < 0.05; ** indicates p value < 0.01; *** indicates p value < 0.001; N.S= Non-significant.