Establishment and characterization of cetuximab resistant head and neck squamous cell carcinoma cell lines: focus on the contribution of the AP-1 transcription factor

Abstract

Background: After an initial response to EGFR targeted therapy, secondary resistance almost invariably ensues, thereby limiting the clinical benefit of the drug. Hence, it has been recognized that the successful implementation of targeted therapy in the treatment of HNSCC cancer is very much dependent on predictive biomarkers for patient selection.

Methods: We generated an in vitro model of acquired cetuximab resistance by chronically exposing three HNSCC cell lines to increasing cetuximab doses. Gene expression profiles of sensitive parental cells and resistant daughter cells were compared using microarray analysis. Growth inhibitory experiments were performed with an HB-EGF antibody and the MMP inhibitor, both in combination with cetuximab. Characteristics of EMT were analyzed using migration and invasion assays, immunofluorescent vimentin staining and qRT-PCR for several genes involved in this process. The function of the transcription factor AP-1 was investigated using qRT-PCR for several genes upregulated or downregulated in cetuximab resistant cells. Furthermore, anchorage-independent growth was investigated using the soft agar assay.

Results: Gene expression profiling shows that cetuximab resistant cells upregulate several genes, including interleukin 8, the EGFR ligand HB-EGF and the metalloproteinase ADAM19. Cytotoxicity experiments with neutralizing HB-EGF antibody could not induce any growth inhibition, whereas an MMP inhibitor inhibited cell growth in cetuximab resistant cells. However, no synergetic effects combined with cetuximab could be observed. Cetuximab resistant cells showed traits of EMT, as witnessed by increased migratory potential, increased invasive potential, increased vimentine expression and increased expression of several genes involved in EMT. Furthermore, expression of upregulated genes could be repressed by the treatment with apigenin. The cetuximab resistant LICR-HN2 R10.3 cells tend to behave differently in cell culture, forming spheres. Therefore, soft agar assay was performed and showed more and larger colonies when challenged with cetuximab compared to PBS challenged cells.

Conclusions: In summary, our results indicate that increased expression of the ligand HB-EGF could contribute to resistance towards cetuximab in our cetuximab resistant HNSCC cells. Furthermore, several genes upregulated or downregulated in cetuximab resistant cells are under control of the AP-1 transcription factor. However, more studies are warranted to further unravel the role of AP-1 in cetuximab resistance.

Keywords: ADAM19; HB-EGF; Head and neck squamous cell carcinoma (HNSCC); cetuximab; interleukin 8; resistance; transcription factor AP-1.

Figure 1

Dose response curves of cetuximab for isogenic cetuximab resistant and sensitive HNSCC cell lines. A. Dose response curves for the sensitive mother cell line (suffix PBS) and the resistant daughter cell line (suffix Rx) after exposure to cetuximab for 168 hours. The graph represents three independent experiments. B. Dose response curves of the three cetuximab resistant cell lines after 6 weeks of culture in drug-free medium, followed by cetuximab treatment for 168 hours. This graph represent one experiment executed in six-fold.

Figure 2

Simplified model of our hypothesis for acquired cetuximab resistance in our HNSCC cells. A. Schematic representation of the localization of HB-EGF, ADAM19 and EGFR. After a trigger (e.g. interleukin 8) ADAM becomes active and cleaves the pro-HB-EGF ligand in a soluble HB-EGF (sHB-EGF) and a carboxy-terminal fragment cyto-HB-EGF. The latter translocates to the nucleus where it activates several genes involved in cell cycle progression, whereas sHB-EGF can transactivate EGFR. B. Relative mRNA expression of HB-EGF, ADAM19 and interleukin 8 assessed by qRT-PCR in cetuximab resistant and sensitive cells challenged with cetuximab or vehicle (PBS) solution for 13 hours.

Figure 3

Relative cell survival of LICR-HN5 R9.1 and SC263 R10.2 cells after treatment with a neutralizing antibody targeting HB-EGF or the matrix metalloproteinase inhibitor GM6001. A & B. Cell survival of LICR-HN5 R9.1 cells (A) and SC263 R10.2 cells (B) after several conditions of a neutralizing HB-EGF antibody in combination with cetuximab. C & D. Cell survival of LICR-HN5 R9.1 cells (C) and SC263 R10.2 cells (D) after several conditions of a broad-spectrum MMP inhibitor GM6001 in monotherapy and in combination with cetuximab. Cells treated with vehicle-solution (PBS or DSMO) were used as absolute control and assimilated to 100% survival. All experiments were performed in triplicate. E. Statistical analysis of LICR-HN5 R9.1 and SC263 R10.2 cells and the MMP inhibitor GM6001 for the experimental conditions using Mann-Whitney U test.

Figure 4

Characteristics of EMT: migration, invasion and vimentine expression. (A & C) Migration (A) and invasion (C) assay for cetuximab resistant LICR-HN5 R9.1 and SC263 R10.2 and cetuximab sensitive LICR-HN5 PBS and SC263 PBS cells. Migration experiments were performed in triplicate, whereas invasion assay were performed in quadruplicate. (B & D) Vimentin immunofluorescent staining of (B) LICR-HN5 and (D) SC263; left: resistant and right: sensitive isogenic cell lines. Blue = dapi, nucleus and red = vimentin.

Figure 5

Relative gene expression profiles of EMT-related genes for the two HNSCC cell lines LICR-HN5 and SC263 (sensitive and resistant forms) challenged with cetuximab “cetu” or PBS “pbs” for 13 or 72 hours (blue and green respectively). qRT-PCR reactions were performed in triplicate.

Figure 6

A. List of AP-1 target genes and the effect on their expression; +: positively regulated, -: negatively regulated and ±: positively or negatively regulated depending on the AP-1 dimer composition [66,69]. B. Gene expression profiles of several AP-1 target genes after treatment (13 h) with 15 nM cetuximab (control) and after treatment with 15 nM cetuximab plus ERK inhibitor 25 mM apigenin. qRT-PCR reactions were performed in triplicate.

Figure 7

(A) Macroscopic image of LICR-HN2-R cells in culture. (B & C) Soft agar assay of LICR-HN2 R10.3 challenged with 15 nM cetuximab (B), and LICR-HN2 R10.3 cells challenged with PBS (C). Soft agar assay was performed in 2 wells and 3 non-overlapping fields were taken per well at a 4x objective on an EVOS ®FL Digital Fluorescence Microscope.