Integrated single-cell and bulk gene expression and ATAC-seq reveals heterogeneity and early changes in pathways associated with resistance to cetuximab in HNSCC-sensitive cell lines

Abstract

Background: Identifying potential resistance mechanisms while tumour cells still respond to therapy is critical to delay acquired resistance.

Methods: We generated the first comprehensive multi-omics, bulk and single-cell data in sensitive head and neck squamous cell carcinoma (HNSCC) cells to identify immediate responses to cetuximab. Two pathways potentially associated with resistance were focus of the study: regulation of receptor tyrosine kinases by TFAP2A transcription factor, and epithelial-to-mesenchymal transition (EMT).

Results: Single-cell RNA-seq demonstrates heterogeneity, with cell-specific TFAP2A and VIM expression profiles in response to treatment and also with global changes to various signalling pathways. RNA-seq and ATAC-seq reveal global changes within 5 days of therapy, suggesting early onset of mechanisms of resistance; and corroborates cell line heterogeneity, with different TFAP2A targets or EMT markers affected by therapy. Lack of TFAP2A expression is associated with HNSCC decreased growth, with cetuximab and JQ1 increasing the inhibitory effect. Regarding the EMT process, short-term cetuximab therapy has the strongest effect on inhibiting migration. TFAP2A silencing does not affect cell migration, supporting an independent role for both mechanisms in resistance.

Conclusion: Overall, we show that immediate adaptive transcriptional and epigenetic changes induced by cetuximab are heterogeneous and cell type dependent; and independent mechanisms of resistance arise while tumour cells are still sensitive to therapy.

Fig. 1. Single-cell RNA-seq profiling of cetuximab-treated and -untreated HNSCC cell lines.

a SCC1, SCC6 and SCC25 cell lines were treated with cetuximab or PBS (untreated controls) for 5 consecutive days after which cells were collected for single-cell RNA-seq (scRNA-seq). b scRNA-seq analysis demonstrates that each cell line presents a specific gene expression profile. c In response to cetuximab, the SCC6-treated (red) and untreated (black) clones separate completely, while the SCC1 and SCC25 present some overlap in the distribution regarding the transcriptional profile. d Inter-cell heterogeneity is more evident for TFAP2A and VIM mRNA levels, with SCC1 presenting high levels of TFAP2A and no expression of VIM. The co-expression analysis shows that in SCC1 there is no change in the levels of TFAP2A or VIM in response to cetuximab; SCC6-treated cells are VIM + (orange and purple), while untreated are negative (green and blue) with different status for TFAP2A expression; and most of the SCC25 cells responding with increase in VIM, but with some untreated clones presenting the same expression profile for VIM and TFAP2A (purple) and with VIM- clones only detected in the untreated group. e, f, g Bar plots represent the number of treated and untreated cells per each gene signature.

Fig. 2. Single-cell heterogeneity and RNA velocity changes as a response to cetuximab.

a, b SCC1 presents increased heterogeneity as measured by EVA statistics in MSigDB Hallmark pathways. Changes in SCC6 and SCC25 heterogeneity are not as significant, but also suggest an adaptive early response to cetuximab. b, c RNA velocity analysis corroborates the heterogeneity analysis and shows that cetuximab triggers a dynamic process in all three cell lines and suggests that cells would move from a more homeostatic state (PBS, grey) to a state with increased transcriptional complexity (CTX, red) in response to therapy.

Fig. 3. TFAP2A targets and EMT markers expression in response to cetuximab.

a SCC1, SCC6 and SCC25 cell lines were treated with cetuximab or PBS (untreated controls) for 5 consecutive days, and cells were collected daily for bulk RNA-seq (RNA-seq). b Among the genes differentially expressed among all three cell lines as a response to cetuximab therapy, the gene set enrichment analysis shows significant presence (p ≤ 0.05) of genes that are TFAP2A targets or that participate in the EMT process. When analysed individually, the TFAP2A and EMT differential expressed genes are specific in each of the cell lines. c, f SCC1 and e, h SCC25 present changes as soon as 24 h (1 day) after cetuximab therapy, while in (d, g) SCC6 the changes are only detected at 96 h (4 days) after cells are treated.

Fig. 4. Chromatin structure changes during short time treatment with cetuximab.

a ATAC-seq was performed after SCC1, SCC6 and SCC25 were treated for 5 days with cetuximab, and also in the untreated (PBS) controls. b–d Differential binding analysis show that the promoters accessibility changes in response to 5 days of therapy are capable of separating the cetuximab from the PBS replicates in all three cell lines. e With the exception of SCC1, there are enrichment for TFAP2A and EMT promoters among the ATAC-seq peaks in SCC6 and SCC25. f The differential binding analysis show that SCC25 is the gene with the highest number of genes with chromatin changes in response to cetuximab, and also identified promoters that are changed in more than one cell line (underlined gene names).

Fig. 5. Effect of TFAP2A knockdown in HNSCC cell growth during treatment with cetuximab and JQ1 alone or the combination.

a Functional validation of the role of TFAP2A in HNSCC in vitro was evaluated by siRNA gene silencing in SCC1, SCC6 and SCC25. Cells were treated with cetuximab, JQ1, combination (combo) or vehicle (mock) for 5 days, and the impact of gene knockdown and therapy was determined by measuring proliferation rates. b–d Transfected groups (full lines, left) were compared with the groups with normal levels of TFAP2A (dashed lines, right - NTC). In all cell lines, TFAP2A knockdown induce lower proliferation rates (black lines) at different levels, depending on the cell. Cetuximab treatment (red lines) present a synergistic effect, but JQ1 (blue lines) efficacy is even greater in reducing cell growth. Little effect is noted with the combination (orange lines, COMBO) when compared with the effects of JQ1 alone.

Fig. 6. Cell migration in the absence of TFAP2A and the effect of cetuximab and JQ1 therapy.

a To further evaluate the interplay between TFAP2A and EMT, cells transfected with siRNA against TFAP2A and treated with cetuximab, JQ1, combination (combo) or vehicle (mock) were used for a migration assay. Migration was measured for a total of 24 h immediately after insert removal. b–d No significant changes in migration was noted when comparing the non-transfected (dashed lines, left) and transfected (full lines, right - NTP) SCC1, SCC6 and SCC25 cells and different treatment groups. e–g Although migration changes were not observed, there are changes in VIM expression as response to siRNA silencing and the different therapies in all three cell lines.