Overcoming cetuximab resistance in HNSCC by Hsp90 inhibition to enhance EGFR degradation Authors: Ranjit K Mehta, Mingjia Tan, Krithika Suresh, Theodore S Lawrence, Mukesh K Nyati Published: 2026 ABSTRACT AbstractCetuximab, a monoclonal antibody targeting the epidermal growth factor receptor (EGFR), exhibits limited clinical benefit in head and neck squamous cell carcinoma (HNSCC) despite high receptor expression. We screened a panel of HNSCC cell lines and confirmed that neither total nor phospho-EGFR expression levels correlate with cetuximab sensitivity, identifying KRAS mutations as a primary determinant of de novo resistance. While cetuximab successfully inhibits EGFR phosphorylation and dimerization, it fails to induce significant receptor degradation in cetuximab-resistant lines. Here, we identify a novel “monomer trap” mechanism wherein Hsp90 stabilizes dephosphorylated, cetuximab-bound monomers to prevent their turnover and facilitate recycling to the plasma membrane. Using chemical crosslinking and biochemical assays, we show that disrupting the EGFR:Hsp90 interaction with a sub-cytotoxic minimal essential dose (MED) of an Hsp90 inhibitor selectively redirects monomeric EGFR toward proteasomal and lysosomal degradation. This strategy resulted in 70 to 90% EGFR depletion and synergistic cell killing independent of cell cycle effects, effectively blocking compensatory signaling that drives cetuximab resistance, including in KRAS-mutant models. In vivo, combination therapy significantly suppressed growth in aggressive KRAS-mutant HNSCC xenografts without overt systemic toxicity. Histopathological analysis confirmed that therapeutic efficacy was driven by a significant reduction in EGFR protein levels. Our findings establish Hsp90-mediated stabilization of monomeric EGFR as a novel resistance mechanism and provide a translational rationale for employing low-dose Hsp90 inhibition in combination with cetuximab to improve clinical outcomes for HNSCC patients currently lacking effective targeted options.Keywords: EGFR, cetuximab, HNSCC, Hsp90 inhibition, dimerization, protein degradation, drug resistance, KRAS Head and neck squamous cell carcinoma (HNSCC) is characterized by significant genetic heterogeneity, often marked by overexpression of wild-type epidermal growth factor receptor (EGFR) (1, 2, 3). This overexpression drives oncogenic processes, including cell proliferation and metastasis, and is associated with disease progression (4, 5, 6). While EGFR tyrosine kinase inhibitors (TKIs) are effective in cancers with EGFR mutations, such as non-small cell lung cancer (NSCLC), their efficacy in HNSCC, which predominantly expresses wild-type EGFR, remains limited (7). Chimeric monoclonal antibodies, such as cetuximab, provide a distinct approach to targeting wild-type EGFR. Derived from the murine monoclonal antibody C225, cetuximab binds to the extracellular domain of EGFR, inhibiting epidermal growth factor (EGF) binding and receptor dimerization, thereby reducing EGFR tyrosine kinase activity (8). This receptor blockade enables cetuximab to demonstrate efficacy in tumors expressing wild-type EGFR, including HNSCC (9). Despite these potential advantages, cetuximab monotherapy yields limited clinical benefit in HNSCC, with response rates of only 10 to 15% in recurrent or metastatic cases (10). Over the past 20 years, clinical research has failed to identify a reliable positive prognostic marker for cetuximab efficacy (11). While HPV-positive patients may exhibit improved outcomes, EGFR expression levels do not correlate with cetuximab response (12). Mutations in KRAS, though relatively rare in HNSCC, are associated with de novo resistance (13). KRAS mutations persistently activate downstream EGFR signaling pathways, such as AKT, despite EGFR inhibition, contributing to resistance in preclinical and clinical settings (14, 15, 16). Similarly, TKIs like gefitinib exhibit limited efficacy in HNSCC driven by wild-type EGFR, despite inhibiting EGFR kinase activity. Clinical data indicate that combining bortezomib with cetuximab and radiotherapy may stabilize EGFR, potentially promoting tumor progression, highlighting the critical role of EGFR protein stability in cell survival (17). In contrast, preclinical studies demonstrate that Hsp90 inhibition disrupts the molecular chaperone function stabilizing wild-type EGFR, promoting its proteasomal and lysosomal degradation and attenuating oncogenic signaling across EGFR-driven cancers (18). While global Hsp90 inhibition has faced clinical challenges due to dose-limiting toxicities, targeting the specific stabilization of oncogenic clients offers a more precise therapeutic window. These findings underscore the potential of targeting EGFR stability to enhance therapeutic efficacy in HNSCC and beyond. We previously reported that wild-type EGFR stability is regulated by Hsp90 through a direct interaction with the kinase domain (18, 19). This study investigates the hypothesis that Hsp90-mediated stabilization of cetuximab-induced inactive monomers contributes to drug resistance in HNSCC. We propose that cetuximab-mediated monomerization creates a “trap” where Hsp90 prevents the turnover of dephosphorylated receptors, thereby sustaining a signaling-competent pool at the plasma membrane. By exploring this interaction, we aim to identify a targetable mechanism using minimal essential dose (MED) of Hsp90 inhibitors such as AT-13387. This study evaluates the synergistic potential of cetuximab and AT-13387 to promote EGFR degradation and overcome resistance in HNSCC models, demonstrating that physical elimination of the receptor protein is a superior therapeutic strategy compared to signaling inhibition alone. Results Cetuximab sensitivity across HNSCC cell lines We screened a large panel of HNSCC cell lines for cetuximab sensitivity, including KRAS-mutant UMSCC-74B and isogenic clones of UMSCC-1 and UMSCC-38 expressing either wild-type (WT) or mutant (G12D) KRAS. To evaluate the generalizability of the observed phenomenon across the heterogeneous HNSCC landscape, we utilized a total of 18 HNSCC cell lines (Table 1). Table 1 provides details on each cell line's site of origin, HPV status, and KRAS status. Cetuximab sensitivity was measured using clonogenic survival assays for the 14 lines capable of forming colonies. For the four lines that did not form colonies (UMSCC-2, UMSCC-5, UMSCC-104, and UMSCC-105), the CellTiter-Glo assay was employed. Table 1Cetuximab sensitivity in HNSCC Cell lines with Varying HPV and KRAS statusCell linesSite of originHPV statusKRAS statusAssay typeCetuximab sensitivityIC50 (μg/ml ± SD)UMSCC-38Tonsil−VEWild-typeClonogenicSensitive0.04 ± 0.03UMSCC-38-G12DTonsil−VEMutant (G12D)ClonogenicResistant0.46 ± 0.03UMSCC-104∗Vocal Cord+VEWild-typeCellTiter-GloSensitive10.1 ± 2.1UMSCC-22BPharynx−VEWild-typeClonogenicSensitive25.2 ± 12.4UMSCC-12Larynx−VEWild-typeClonogenicSensitive42.8 ± 2.7UMSCC-1Floor of Mouth−VEWild-typeClonogenicSensitive51.3 ± 2.89UMSCC-29Oral Cavity−VEWild-typeClonogenicSensitive58.3 ± 7.8UMSCC-11BEpiglottis−VEWild-typeClonogenicSensitive66.4 ± 2.2UMSCC-47Tongue+VEWild-typeClonogenicModerately Sensitive123.5 ± 7.1UMSCC-10BMaxillary Sinus−VEWild-typeClonogenicModerately Sensitive130.5 ± 1.36UMSCC-2∗Tongue+VEWild-typeCellTiter-GloModerately Sensitive150 ± 10UMSCC-17BLarynx−VEWild-typeClonogenicResistant191.9 ± 67.8UMSCC-1-G12DFloor of Mouth−VEMutant (G12D)ClonogenicResistant278.3 ± 2.89UMSCC-33Maxillary Sinus−VEWild-typeClonogenicResistant310 ± 13.6UMSCC-5∗Oral Cavity−VEWild-typeCellTiter-GloResistant>300UMSCC-74BFloor of Mouth−VEMutant (G12D)ClonogenicResistant>300UMSCC-92Larynx−VEWild-typeClonogenicResistant>300UMSCC-105∗Larynx−VEWild-typeCellTiter-GloResistant>300Cetuximab sensitivity in HNSCC cell lines. Cell lines are ordered by sensitivity based on IC50 values. Ectopic expression of mutant KRAS (G12D) via lentiviral transduction in sensitive parental lines (UMSCC-1 and UMSCC-38) resulted in an approximately 10-fold upward shift in IC50 values, consistent with the de novo resistance observed in natural KRAS-mutant lines (e.g., UMSCC-74B). Sensitivity was determined using either clonogenic survival or CellTiter-Glo assays (indicated by an asterisk). Data represents the mean IC50 (μg/ml) ± SD from three independent experiments. Cetuximab sensitivity in HNSCC Cell lines with Varying HPV and KRAS status Cetuximab sensitivity in HNSCC cell lines. Cell lines are ordered by sensitivity based on IC50 values. Ectopic expression of mutant KRAS (G12D) via lentiviral transduction in sensitive parental lines (UMSCC-1 and UMSCC-38) resulted in an approximately 10-fold upward shift in IC50 values, consistent with the de novo resistance observed in natural KRAS-mutant lines (e.g., UMSCC-74B). Sensitivity was determined using either clonogenic survival or CellTiter-Glo assays (indicated by an asterisk). Data represents the mean IC50 (μg/ml) ± SD from three independent experiments. We categorized the cell lines into three groups based on IC50 values: sensitive, moderately sensitive, and resistant in Table 1. Notably, the isogenic clone UMSCC-38-G12D, derived from the highly responsive UMSCC-38 line via G12D-KRAS transduction, exhibited over ten-fold increase in IC50, reflecting a marked decrease in sensitivity while remaining less resistant than several other lines. While specific lines were selected for certain biochemical endpoints based on technical requirements (such as colony-forming efficiency or basal protein expression), we obtained overall similar results across all tested models, establishing the robustness of the mechanism. These findings underscore KRAS mutations as key determinants of cetuximab resistance in HNSCC, while neither HPV status nor the site of cancer origin showed significant correlation with cetuximab sensitivity. No correlation between cetuximab response and EGFR expression in HNSCC cell lines To determine whether cetuximab response correlates with EGFR expression levels, we assessed the basal expression of total EGFR and phospho-EGFR (pY845 and pY1068) in 15 HNSCC cell lines using immunoblotting. The cell lines included 14 colony-forming lines and four non-colony-forming lines, as defined in Table 1. Representative immunoblots revealed variable expression levels of total EGFR and phospho-EGFR across the cell lines (Fig. 1A). No significant correlation was observed between cetuximab response (IC50) and either total EGFR expression (Pearson r = 0.07, p = 0.80; Spearman r = −0.03, p = 0.91) or phospho-EGFR expression (Pearson r = −0.43, p = 0.11; Spearman r = −0.36, p = 0.19) (Fig. 1B). Basal Expression of Phospho-EGFR and Total EGFR in HNSCC Cell Lines and Correlation with Cetuximab Response.A, immunoblot analysis of basal levels of pY845 EGFR, pY1068 EGFR, and total EGFR across a panel of 15 HNSCC cell lines. GAPDH served as a loading control. Cell lines are arranged by increasing resistance to cetuximab (left to right). B, correlation analysis between mean expression levels of total EGFR (left) or pEGFR (right) and cetuximab IC50 values. Pearson and Spearman correlation coefficients and corresponding p-values are shown. Shaded areas represent the 95% confidence interval. Cetuximab inhibits EGFR dimerization and differentially affects EGFR expression in responsive and resistant HNSCC cell lines Given the absence of correlation between basal EGFR expression and cetuximab response, cetuximab’s effect on EGFR dimerization and pharmacodynamic changes in EGFR levels were examined in responsive and resistant HNSCC cell lines to identify mechanisms associated with differential sensitivity. Responsive (UMSCC-1, UMSCC-38) and resistant (UMSCC-17B, UMSCC-92) cell lines were treated with cetuximab (100 μg/ml), and EGFR dimer formation was assessed following EGF stimulation. Immunoblot analysis indicated inhibition of EGFR dimerization in both responsive and resistant cell lines (Fig. 2A). Cetuximab Effects on EGFR Dimerization and Levels in HNSCC Cell Lines.A, evaluation of EGFR dimerization by chemical crosslinking with DSS (500 μM). Cells were treated with EGF (30 ng/ml) and/or cetuximab (100 μg/ml) as indicated. Blots show monomeric (170 kDa) and dimeric (340 kDa) EGFR. B, immunoblot analysis of pY845-EGFR and total EGFR levels in sensitive (UMSCC-1, −38) and resistant (UMSCC-17B, −33, −74B) cell lines following 24 h treatment with cetuximab (100 μg/ml). Relative pEGFR levels normalized to total EGFR, and total EGFR levels normalized to GAPDH, are indicated below the blots. Quantification of pEGFR relative to the total EGFR pool reveals that while cetuximab induced receptor downregulation in sensitive lines, approximately 70% of the residual receptor remained phosphorylated. In contrast, total EGFR levels were only marginally affected in resistant lines; however, cetuximab successfully dephosphorylated the receptor, confirming potent target engagement and antibody efficacy despite therapeutic resistance. Additionally, steady-state levels of phospho-EGFR (pY845) and total EGFR were measured in responsive (UMSCC-1, UMSCC-38) and resistant (UMSCC-33, UMSCC-74B) cell lines after cetuximab treatment (100 μg/ml, 24 h). In responsive cell lines, phospho-EGFR levels decreased by 20 to 80% (UMSCC-1: 80%, UMSCC-38: 20%), and total EGFR levels decreased by 50 to 70% (UMSCC-1: 50%, UMSCC-38: 70%) compared to untreated controls, based on densitometry measurements. In resistant cell lines, phospho-EGFR levels decreased by 60 to 80% (UMSCC-17B: 90%, UMSCC-33: 60%, UMSCC-74B: 80%), while total EGFR levels showed reductions of 20% or less (UMSCC-17B: 0%, UMSCC-33: 10%, and UMSCC-74B: 20%). As expected, there were no obvious relationships between the degree of signaling inhibition and receptor loss in resistant cell lines (Fig. 2B). Effect of cetuximab and AT-13387 on EGFR-Hsp90 interaction We previously reported that EGFR is an Hsp90 client protein stabilized by direct interaction Hsp90 (18). To investigate the dynamics of this interaction during therapy, we assessed the effect of cetuximab and AT-13387, alone or in combination, in responsive (UMSCC-1, UMSCC-38) and resistant (UMSCC-74B) cell lines. Co-immunoprecipitation with anti-Hsp90 was performed following treatment. Under basal conditions, the interaction between EGFR and Hsp90 was minimal, requiring high-exposure settings for detection (Fig. 3). In contrast, cetuximab treatment triggered a dramatic accumulation of EGFR on Hsp90, resulting in a robust complex detectable even at low exposures (∼6–8-fold induction over control). We next examined if Hsp90 inhibition could disrupt this antibody-induced association. Densitometric analysis of non-saturated immunoblots demonstrated that AT-13387 destabilized the complex, though the extent of dissociation varied by cell line. The combination reduced the cetuximab-induced EGFR-Hsp90 association by 70% in the highly sensitive UMSCC-38 line, compared to partial reductions of 50% in UMSCC-74B and 30% in UMSCC-1. These data suggest that cetuximab drives EGFR into an Hsp90-chaperoned complex, which can be physically destabilized by concurrent Hsp90 inhibition. Cetuximab induces EGFR-Hsp90 interaction, and Hsp90 inhibition disrupts the complex.A, Co-immunoprecipitation (IP) analysis of the EGFR-Hsp90 complex. Lysates from UMSCC-1, UMSCC-38, and UMSCC-74B cells were harvested following 2 h treatment with vehicle control, cetuximab (50 μg/ml), AT-13387 (1 μM), or the combination. IgG served as a negative control for the IP. Immunoblots represent levels of co-precipitated EGFR (standard and high exposures) and total Hsp90 in the IP fraction (IP:Hsp90). Corresponding input lysates demonstrate total steady-state protein levels of EGFR, Hsp90, and β-actin. Molecular weight markers (kDa) are indicated to the right. B, densitometric quantification of Hsp90-bound EGFR. Data represent the fold-change in EGFR abundance within the Hsp90 IP fraction normalized to total immunoprecipitated Hsp90. These data statistically confirm the dramatic induction of the Hsp90-EGFR interaction (the "monomer trap") following cetuximab treatment (10–15-fold increase over control) and its subsequent disruption when combined with AT-13387. Values represent the mean ± SD of three independent biological replicates. Statistical significance was determined by one-way ANOVA (∗∗p < 0.01, ∗∗∗p < 0.001). To evaluate the stability of monomeric versus dimeric receptors, de novo protein synthesis was inhibited with cycloheximide (50 μg/ml), followed by treatment with AT-13387, cetuximab, or their combination. DSS crosslinking was subsequently employed to stabilize the preexisting EGFR pool (Fig. S2). AT-13387 monotherapy had minimal impact on the EGFR dimer. As expected, cetuximab alone inhibited EGFR dimerization, generating a pool of receptors comprised primarily of monomers. Densitometric quantification confirms that combined treatment did not further suppress the dimeric fraction but resulted in a significant reduction in monomeric EGFR levels within the 3 h treatment window. In contrast, cetuximab monotherapy showed minimal impact on total EGFR protein levels. These results demonstrate that the combination of cetuximab and AT-13387 destabilizes EGFR by specifically accelerating the depletion of the monomeric fraction. Effect of cetuximab and AT-13387 on cell survival and EGFR degradation in HNSCC cell lines Following the observed reduction in EGFR-Hsp90 interaction with cetuximab and AT-13387, the combination of these agents was assessed for its effect on cell survival and EGFR degradation in sensitive and resistant HNSCC cell lines to explore strategies for overcoming resistance. Clonogenic survival assays were conducted on sensitive (UMSCC-1, UMSCC-38), moderately sensitive (UMSCC-29), and resistant (UMSCC-17B, UMSCC-33, UMSCC-74B) cell lines treated with cetuximab (50 μg/ml), except UMSCC-38 (3 ng/ml) alone or in combination with AT-13387 (10 nM). In resistant cell lines, cetuximab or AT-13387 alone induced 10 to 20% cell death, while their combination induced 70 to 80% cell death (Fig. 4). There was evidence of synergy (negative interaction estimate, p < 0.05) for cell lines UMSCC-1, UMSCC-29, and UMSCC-74B. For UMSCC-17B and UMSCC-33, the combination had a significantly greater surviving fraction than the individual monotherapies but not beyond additivity. UMSCC-38, which is very responsive to cetuximab, showed a positive interaction estimate indicating that the combination effect was less than expected additivity. We confirmed synergistic cell killing in multiple lines, including UMSCC-1 and UMSCC-74B (p < 0.05), as detailed in the quantitative synergy analysis (Table 2). Significance of surviving fractions, as well as formal synergy analyses were summarized in Table 2 and Figure 4B. Combined Hsp90 inhibition and cetuximab synergistically reduce clonogenic survival in resistant HNSCC models.A, clonogenic survival assays were conducted across a panel of six HNSCC cell lines. Cells were treated with cetuximab (50 μg/ml), AT-13387 (10 nM), or the combination. Bars represent the surviving fraction relative to untreated controls. Values are shown as the mean ± SD from three independent experiments. Statistical significance: ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001 (one-way ANOVA). B, Formal synergy quantification. A two-factor linear model was utilized to determine the interaction effect between treatments on cell survival fraction. The interaction (synergy) estimate, 95% confidence interval (CI), and p-value are provided. Negative interaction estimates denote synergy (cell killing exceeding simple additivity), while positive estimates indicate sub-additivity. Significant synergistic interactions were observed in UMSCC-1, UMSCC-29, and the KRAS-mutant UMSCC-74B line (p < 0.05), with 95% CIs entirely below zero. While resistant lines UMSCC-17B and UMSCC-33 showed trends toward synergy (negative estimates), the highly sensitive line UMSCC-38 exhibited significant sub-additivity (p = 0.0011). These results suggest that the synergistic potential of combining Hsp90 inhibition with cetuximab is most pronounced in the setting of established cetuximab resistance. Table 2Quantitative analysis of Cetuximab and AT-13387 combination efficacy and synergyCell LinesMean surviving fraction (±SD)p-values (t test)Cetuximab (Cet)AT-13387 (AT)Cet + ATCet + AT vs CetCet + AT vs ATUMSCC-10.57 ± 0.070.63 ± 0.140.14 ± 0.030.00270.0001UMSCC-290.67 ± 0.050.69 ± 0.030.2 ± 0.030.00030.0004UMSCC-380.43 ± 0.050.54 ± 0.070.26 ± 0.050.01400.0076UMSCC-17B0.9 ± 0.060.55 ± 0.130.32 ± 0.070.00690.0409UMSCC-330.9 ± 0.070.52 ± 0.040.29 ± 0.070.00120.0343UMSCC-74B0.94 ± 0.040.55 ± 0.070.29 ± 0.030.00020.0128Comparative survival analysis. Cells were treated with cetuximab (50 μg/ml, except UMSCC-38 at 3 ng/ml), AT-13387 (10 nM), or the combination. Results represent mean surviving fractions (±SD) normalized to untreated controls. p-values were determined by two-sample t-tests comparing the combination to each respective monotherapy. Statistical interaction estimates providing proof of synergy are shown in the main Figure 4B. Quantitative analysis of Cetuximab and AT-13387 combination efficacy and synergy Comparative survival analysis. Cells were treated with cetuximab (50 μg/ml, except UMSCC-38 at 3 ng/ml), AT-13387 (10 nM), or the combination. Results represent mean surviving fractions (±SD) normalized to untreated controls. p-values were determined by two-sample t-tests comparing the combination to each respective monotherapy. Statistical interaction estimates providing proof of synergy are shown in the main Figure 4B. To establish that the synergistic effects of Hsp90 inhibition are not unique to AT-13387, we evaluated a second structurally distinct inhibitor, ganetespib (Fig. S1). Combination treatment with cetuximab and ganetespib resulted in similar patterns of enhanced cytotoxicity and inhibition of downstream signaling across the cell line panel. This expanded analysis confirms that ganetespib reduces total EGFR protein levels and inhibits receptor phosphorylation, mirroring the pharmacodynamic profile of AT-13387. These findings indicate that the observed synergy and receptor depletion represent a class effect of Hsp90 inhibition in the context of cetuximab resistance. To investigate whether the observed mechanism is generalizable to other cancer types driven by wild-type EGFR, we utilized an A549 lung cancer model engineered with a HiBiT tag to monitor surface EGFR protein abundance. Treatment with the combination of cetuximab and AT-13387 significantly reduced surface EGFR protein levels compared to either monotherapy (Fig. S4A). Furthermore, clonogenic survival assays demonstrated that the addition of 10 nM AT-13387 significantly sensitized A549-HiBiT cells to cetuximab across a range of concentrations (Fig. S4B). To isolate the combinatorial effect, the surviving fraction for the combination group was normalized to the baseline survival of cells treated with 10 nM AT-13387 monotherapy. These findings indicate that the physical depletion of the EGFR protein pool is a viable therapeutic strategy beyond HNSCC and demonstrate the robustness of the monomer-trap mechanism. Effect of cetuximab and AT-13387 on EGFR dimerization and steady-state EGFR levels in HNSCC cell lines To elucidate the mechanisms underlying the enhanced cell death observed with cetuximab and AT-13387, the effect of this combination on EGFR dimerization and steady-state EGFR levels was examined in sensitive and resistant HNSCC cell lines. The concentration dependence of the effect of cetuximab on EGFR dimerization was assessed, and 3 μg/ml was selected, at which 40 to 60% of EGFR dimers were inhibited. UMSCC-1 (sensitive) and UMSCC-17B (resistant) cell lines were then treated with cetuximab (3 μg/ml), AT-13387 (100 nM), or their combination, and EGFR dimerization was analyzed following EGF stimulation. The combination reduced EGFR dimerization by 70 to 80% compared to AT-13387 alone in both cell lines, while minimal additional reduction was seen compared with cetuximab monotherapy (Fig. 5A). Similar patterns of EGFR dimerization inhibition were observed in UMSCC-38 and UMSCC-74B cell lines, confirming the consistency of this effect across both cetuximab-responsive and resistant models (Fig. S2). Effect of Cetuximab and AT-13387 on EGFR Dimerization, Steady-State EGFR Levels, and Apoptosis.A, EGFR dimerization (monomer vs. dimer) in four HNSCC cell lines (sensitive and resistant) following treatment with cetuximab (3 μg/ml) and/or AT13387 (100 nM) in the presence of EGF. Quantification of EGFR in (B) dimer and (C) monomer forms from three independent biological replicates. D, immunoblot analysis of pY845-EGFR, total EGFR, and downstream signaling effectors (pERK, ERK, pAKT, AKT, and cleaved PARP) after 24 h treatment with cetuximab (50 μg/ml) and/or AT-13387 (30 nM). Densitometric quantification of (E) relative total EGFR levels and (F) cleaved PARP protein levels from the immunoblots shown in panel D. Data are presented as mean ± S.D. of three biological replicates; individual data points are shown to indicate reproducibility. (ns = not significant; asterisks indicate significance levels as defined in Fig. 4). Steady-state levels of total EGFR and phospho-EGFR were assessed in these cell lines after treatment with cetuximab (50 μg/ml), AT-13387 (30 nM), or their combination for 24 h. Total EGFR levels were reduced by 70 to 90% with the combination, while phospho-EGFR levels showed minimal additional reduction, as cetuximab alone dephosphorylated the receptor without reducing total protein levels (Fig. 5B). We further investigated the downstream signaling pathways and observed that phospho-ERK levels were reduced by 20 to 30% and phospho-AKT levels were reduced by 30% to 40% following combination treatment (Fig. 5C). Consistent with the initiation of apoptosis, we also observed induction of cleaved PARP at 24 h, primarily driven by AT-13387, with a modest enhancement in the combination arm in UMSCC-74B. This profound reduction in EGFR protein levels correlates with enhanced cell death in both sensitive and resistant cell lines. Effect of cetuximab on cell cycle progression in sensitive and resistant HNSCC cell lines To explore the mechanistic basis of cetuximab’s differential effects, we investigated its impact on cell cycle progression in a sensitive cell line (UMSCC-38) and a resistant cell line (UMSCC-17B). Cells were treated with a high concentration of cetuximab (300 μg/ml), using the EGFR tyrosine kinase inhibitor erlotinib as a positive control due to its established role in inducing G1 arrest. Following treatment, cell cycle distribution was analyzed by flow cytometry. Compared to untreated controls, cetuximab treatment resulted in a modest increase of 5 to 10% in the proportion of cells in the G1 phase in both sensitive and resistant cell lines. As expected, erlotinib treatment induced 40 to 50% G1 arrest (Fig. S3). These findings suggest that cetuximab’s effect on the cell cycle is limited and does not account for the observed differences in sensitivity between cell lines, indicating that other mechanisms, such as EGFR degradation, may play a more significant role in mediating cetuximab’s cytotoxic effects. The combination showed minimal additional effects on cell cycle distribution compared to cetuximab or AT-13387 alone (Fig. 4B). Cetuximab and AT-13387 induce proteasomal and lysosomal-mediated EGFR degradation in HNSCC and lung cancer cells Given the enhanced EGFR degradation observed with cetuximab and AT-13387, the mechanism of EGFR degradation induced by this combination was investigated in HNSCC and lung cancer cell lines to determine the involvement of proteasomal and lysosomal pathways. UMSCC-1 (sensitive), UMSCC-17B (resistant), and A549-HiBiT (lung cancer) cells were treated with cetuximab (50 μg/ml) and AT-13387 (10 nM) for 12 h, followed by the addition of either the proteasomal inhibitor MG-132 (2 μM) or the lysosomal inhibitor 3-methyladenine (3-MA, 5 mM) or chloroquine for the last 4 h. Inhibition of proteasomal activity with MG-132 partially rescued EGFR levels, with degradation reduced to 10 to 40%. Treatment with lysosomal inhibitors 3-MA or chloroquine also partially rescued EGFR levels, with degradation reduced to 40 to 50%, indicating that both proteasomal and lysosomal pathways are involved (Fig. 6). Cetuximab and AT-13387 Induce Proteasomal and Lysosomal-Mediated EGFR Degradation in HNSCC and Lung Cancer Cells.A, UMSCC-1, UMSCC-17B, and A549-HiBiT cells were treated with the combination of cetuximab (50 μg/ml) and AT-13387 (10 nM) in the presence or absence of the proteasomal inhibitor MG-132 (2 μM) or lysosomal inhibitors chloroquine (CQ, 50 μM) and 3-methyladenine (3-MA, 5 mM). B, bar graphs showing the quantification of relative EGFR levels, demonstrating the partial rescue of EGFR degradation by inhibition of either degradation pathway. Combination of cetuximab and AT-13387 significantly enhanced antitumor efficacy in cetuximab-resistant UMSCC-74B xenografts To evaluate the therapeutic efficacy of combined cetuximab and AT-13387 in vivo, UMSCC-74B cells were implanted into immunocompromised mice to establish xenograft tumors. Combination treatment resulted in a significant effect on the tumor growth rate (Fig. 7A). Time to tumor doubling was not significantly different across treatment groups; however, there was a significant difference across groups in time to second doubling (p = 0.0001), with the combination group having significantly longer delays compared with either cetuximab (p = 0.0006) or AT-13387 (p = 0.0002). For both doubling and second doubling (Fig. 7, B and E), the interaction effects were estimated to be negative, which indicates a trend towards synergistic growth delay; however, neither reached statistical significance (doubling p = 0.46; second-doubling p = 0.34). We observed a minimal effect on tumor doubling time with either cetuximab or low-dose AT-13387 alone. The AT-13387 treatment as monotherapy or combination with cetuximab was well-tolerated (Fig. 7C). Individual biological replicates at the study endpoint showed a marked reduction in tumor volume in the combination group (Fig. 7D). Histopathological analysis of excised tumors by H&E staining revealed a marked reduction in viable tumor cell content in the combination-treated group. Consistent with these findings, immunohistochemical analysis demonstrated significantly reduced EGFR expression in tumors from mice receiving combined cetuximab and AT-13387 (Fig. 7, G and F). To evaluate the induction of apoptosis in vivo, we performed immunohistochemical analysis for cleaved PARP on the tumor specimens. We observed no significant differences in staining intensity between treatment groups (data not shown), suggesting that the therapeutic delay in tumor doubling is not primarily driven by apoptosis-mediated cell death at the study endpoint. Because the therapeutic delay in tumor doubling was not primarily driven by apoptosis at the endpoint, we assessed whether the combination treatment reduces cellular proliferation. Immunohistochemical analysis for Ki-67 revealed a dramatic reduction in the proliferative index in the combination-treated group (25.5% ± 6) compared to vehicle (80.5% ± 7.1) or cetuximab monotherapy (81.5% ± 8.7; p < 0.001), identifying decreased proliferation as a major driver of tumor growth inhibition (Fig. 7F). Dual targeting of EGFR and Hsp90 enhances antitumor efficacy in UMSCC-74B xenografts.A, tumor growth kinetics in UMSCC-74B xenografts (n = 6 per group). Mice received vehicle control, cetuximab, AT-13387, or the combination. Data represent the fold-change in tumor volume relative to Day 0. B, Kaplan-Meier plot showing time to second tumor doubling (TD2). TD2 was defined as the time from baseline until tumor volume reached four-fold the baseline measurement. C, longitudinal monitoring of mouse body weight. Weights are expressed as a percentage of the weight recorded at Day 0. D, distribution of relative tumor volumes at Day 14. Individual biological replicates are shown superimposed on box plots representing the median and interquartile range. E, summary of median TD2, 95 percent confidence intervals (CI), and global p-values. NR indicates the median was not reached during the study period. F, representative histological analysis of harvested tumors. Top panels show H&E staining; middle panels show IHC staining for total EGFR. IHC H-scores (mean ± SD) are provided below the respective images. Bottom panels shows representative Ki-67 immunohistochemistry, which demonstrate the effect of combination treatments on cellular proliferation. The corresponding mean percentage of positive cells are provided below each image. Scale bars represent 100 μm (top) and 50 μm (middle and bottom panels). G, quantification of the percentage of viable tumor area. Bars represent the mean ± SD for each treatment group. Together, these results demonstrate that dual targeting of EGFR and low-dose Hsp90 enhances antitumor efficacy in vivo without overt systemic toxicity. Discussion Cetuximab, the only approved targeted therapy for locoregionally advanced and recurrent/metastatic HNSCC, exhibits limited efficacy, with only approximately 15% of patients responding when combined with cisplatin or radiotherapy (20). To begin to improve on this modest response, we screened over 18 HNSCC cell lines, identifying six sensitive, four moderately sensitive, and eight resistant lines. Neither EGFR expression levels nor HPV status correlated with cetuximab responsiveness, consistent with clinical observations that EGFR expression does not predict response. However, the presence of G12D-KRAS mutations in resistant lines, such as UMSCC-74B, was associated with reduced efficacy, aligning with findings in other cancers where RAS pathway alterations confer resistance to EGFR-targeted therapies (21). A key finding from our study is the role of EGFR stabilization via Hsp90 interaction as a novel resistance mechanism in HNSCC. To our knowledge, this is the first report identifying the Hsp90-mediated prevention of EGFR degradation as the specific mechanism of action (MOA) for de novo cetuximab resistance. While cetuximab effectively inhibited EGFR dimerization across all cell lines, resistant lines showed minimal reduction in total EGFR levels, indicating that signaling inhibition alone is insufficient to overcome resistance. We observed that Hsp90 binds to the cetuximab-bound EGFR monomer, preventing its degradation and sustaining EGFR function in resistant cells. This conclusion is supported by the steric mechanism of the antibody, which blocks the dimerization interface of Domain III. Because cetuximab dramatically reduces the dimer fraction (Figs. 2A and 5A), the EGFR that co-precipitates with Hsp90 in the presence of the antibody is monomeric by definition. While we cannot rule out Hsp90 binding to transient dimers with absolute certainty, the physical blockade of the dimerization interface makes a monomer-specific interaction the most biochemically plausible model. This mechanism contrasts with previous attempts to enhance cetuximab's efficacy using other targeted therapies, such as bortezomib, which failed possibly due to antagonistic interactions that inhibit EGFR turnover (17). The identification of Hsp90-mediated stabilization provides a new perspective on cetuximab resistance, emphasizing the importance of targeting protein stability to improve therapeutic outcomes in HNSCC. A critical distinction highlighted by our data is the difference between EGFR kinase inhibition and physical protein degradation. While cetuximab successfully inhibits EGFR phosphorylation and downstream canonical signaling (p-ERK and p-AKT), total EGFR protein levels remain maintained in resistant lines. It is well established that EGFR possesses kinase-independent functions, including the maintenance of nutrient transporters essential for cellular metabolism and survival. Our findings demonstrate that the Hsp90-stabilized monomer pool maintains these non-canonical functions, thereby promoting cell survival and driving de novo resistance. To address this resistance mechanism, we developed a strategy combining cetuximab with a low- or sub-cytotoxic concentration of AT-13387, an Hsp90 inhibitor. We found that a minimal essential dose (MED) is sufficient to dissociate the Hsp90-EGFR interaction, allowing the freed monomer to be sorted for degradation. In resistant HNSCC lines, this combination induced 70 to 80% cell death compared to 10 to 20% with either agent alone, an effect directly correlated with increased EGFR degradation, even in KRAS-mutant cells. Validation in the A549-HiBiT lung cancer model further demonstrated the strategy's applicability across EGFR-driven cancers (Fig. S4). Unlike cetuximab alone, which primarily inhibits EGFR dimerization and phosphorylation, the combination's efficacy did not rely on altering cell cycle distribution, suggesting that degradation-driven cytotoxicity is the primary mechanism of action. Notably, synchronous induction of apoptosis does not appear to play a major role in the reduction of tumor volume in vivo. This finding is particularly significant because the EGFR protein possesses multiple non-canonical functions in cellular survival that persist even when its kinase activity is inhibited, explaining why many tumors fail to respond to receptor blockade alone (22, 23). Further investigation demonstrated that the combination of cetuximab and AT-13387 induces EGFR degradation via both proteasomal and lysosomal pathways. This was evidenced by partial rescue of EGFR levels with MG-132, which reduced degradation to 40 to 80%, and 3-MA or chloroquine, which limited degradation to 40 to 60%, in HNSCC and lung cancer cell lines. This dual mechanism facilitates robust EGFR suppression, potentially mitigating compensatory upregulation, a common resistance mechanism. The involvement of the lysosomal pathway suggests a role for autophagy in EGFR degradation, which may enhance the efficacy of the combination therapy. In contrast, although cetuximab alone substantially reduced EGFR phosphorylation, it had minimal impact on cell cycle distribution, suggesting that EGFR activity inhibition alone is not the primary mediator of cetuximab efficacy. These mechanistic insights underscore the importance of targeting EGFR degradation pathways to achieve sustained EGFR suppression in resistant cancers (23). The limited clinical efficacy of cetuximab in HNSCC, despite its ability to inhibit EGFR signaling, underscores the challenge of translating preclinical findings to patients, where complex resistance mechanisms prevail. Preclinical models, such as xenografts in immunocompromised mice, often fail to capture resistance mechanisms like Hsp90-mediated EGFR stabilization due to their simplified tumor microenvironments (24). Our finding that cetuximab combined with AT-13387 overcomes resistance in KRAS-mutant HNSCC cell lines by enhancing EGFR degradation addresses a critical gap, offering a strategy to target tumors with reduced EGFR signaling dependency. However, the absence of reliable biomarkers for cetuximab response, beyond KRAS/NRAS wild-type status, complicates identifying patients who would benefit from such combinations (25). Importantly, the enhanced antitumor effects were recapitulated in vivo using aggressive KRAS-mutant HNSCC xenografts grown in immune-compromised NSG (NOD-SCID) mice. Because antitumor immunity and antibody-dependent cellular cytotoxicity (ADCC) are largely absent in this model, the significant inhibition of tumor growth observed was driven by the cell-intrinsic reduction in the EGFR protein. The absence of significant body-weight loss and overt toxicity supports the translational feasibility of this combinatorial approach. The dual degradation pathway induced by cetuximab and AT-13387 offers a promising therapeutic strategy, particularly for HNSCC patients with limited treatment options. This combination opens a new therapeutic possibility for patients with wild-type EGFR and mutant-KRAS expressing tumors, where cetuximab is known to fail to produce a desired response. Adding a low-dose Hsp90 inhibitor to regimens combining cetuximab with radiotherapy or chemotherapy could accelerate EGFR degradation while minimizing dose-limiting toxicities, potentially improving clinical outcomes. The strategy's efficacy in KRAS-mutant cells further suggests its potential to benefit a broader patient population. Future studies should evaluate this combination in settings where cetuximab is used with radiotherapy or chemotherapy, assessing its ability to enhance efficacy without increasing toxicity. Experimental procedures Reagents AT-13387 was provided by the Cancer Therapy Evaluation Program (CTEP), National Institutes of Health (NIH), USA, and supplied by Astex Pharmaceuticals. Ganetespib was purchased from Selleckchem. Cetuximab (ERBITUX, C225) was obtained from the University of Michigan Pharmacy (Ann Arbor, MI). Erlotinib and cycloheximide were purchased from Sigma-Aldrich (St Louis, MO). Recombinant human EGF was purchased from R&D Systems. The crosslinker disuccinimidyl suberate (DSS) was purchased from Thermo Fisher Scientific (Waltham, MA). Antibodies targeting phospho-EGFR (pY845, cat# 2231; pY1068, cat# 3777), total EGFR (cat# 4267), Hsp70 (cat# 4872), GAPDH (cat# 2118), phospho-ERK1/2 (Thr202/Tyr204, cat# 4376), total ERK1/2 (cat# 9107), phospho-AKT (Ser473, cat# 4060), and total AKT (cat# 4691) were acquired from Cell Signaling Technology (Danvers, MA). The EGFR antibody used for immunoprecipitation (sc-03) and the β-Actin antibody (sc-47778) were purchased from Santa Cruz Biotechnology. The Hsp90 antibody (cat# 610419) was purchased from BD Transduction Laboratories. The human-specific anti-Ki-67 monoclonal antibody (clone B56, cat# 550609) was purchased from BD Pharmingen (San Jose, CA). All primary antibodies used in this study were purchased from commercial sources and validated for specificity by the manufacturers. The proteasomal inhibitor MG-132 was sourced from Calbiochem. Lysosomal inhibitors 3-methyladenine (3-MA) and chloroquine (CQ) were purchased from Sigma-Aldrich. Cell viability was measured using the CellTiter-Glo Luminescent Cell Viability Assay (Promega, Madison, WI). Cell culture A panel of human head and neck squamous cell carcinoma (HNSCC) cell lines (UMSCC-1, −2, −5, −10B, −11B, −12, −17B, −22B, −29, −33, −38, −47, −74B, −92, −104, and −105) was provided by Dr Thomas Carey (University of Michigan, Ann Arbor, MI). Isogenic cell line pairs (UMSCC-1-G12D-KRAS and UMSCC-38-G12D-KRAS) were generated via lentiviral transduction to express the mutant G12D-KRAS gene as previously described (14). All HNSCC lines were cultured in RPMI 1640 medium (Gibco) supplemented with 10% fetal bovine serum (FBS; HyClone, Logan, UT), 100 units/ml penicillin, and 100 μg/ml streptomycin. UMSCC-22B and UMSCC-92 were maintained in Dulbecco’s Modified Eagle Medium (DMEM; Gibco) with the same supplements. The A549-HiBiT lung cancer cell line, engineered with a HiBiT tag to monitor EGFR protein dynamics, was a gift from Promega. This cell line was recently validated by our group for real-time EGFR protein detection and was utilized as an orthogonal model to assess the efficacy of the cetuximab and AT-13387 combination in inducing EGFR degradation (26). All cell lines were maintained at 37 °C in a humidified atmosphere with 5% CO2 and were routinely tested for Mycoplasma contamination using a PCR-based detection kit (Sigma-Aldrich) to ensure experimental integrity. Clonogenic cell survival assay Clonogenic cell survival assays were conducted following established protocols (27, 28). Briefly, 500–3000 cells were seeded in triplicate in 60-mm dishes containing media with 3% FBS. The following day, cells were treated with cetuximab, AT-13387, or a combination of both. After 8 to 12 days, colonies were fixed with a 1:7 (v/v) acetic acid/methanol solution, stained with 0.5% (w/v) crystal violet, and counted using a stereomicroscope. Drug cytotoxicity was quantified by normalizing the survival of drug-treated cells to that of untreated control cells. The cell survival was calculated as the ratio of the mean inactivation dose under control conditions to the mean inactivation dose following drug exposure. All experiments were conducted in triplicate, and data are presented as the mean ± standard deviation (SD). Cell viability assay Short-term cell viability was assessed using the CellTiter-Glo Luminescent Cell Viability Assay (Promega) according to the manufacturer’s instructions. This assay was primarily utilized for non-colony forming lines (UMSCC-2, −5, −104, and −105) and for comparative viability analysis in UMSCC-1 and −38 (Table 1). Cells were seeded in 96-well opaque-walled plates at a density of 1000 to 3000 cells per well. After 24 h of attachment, cells were treated with cetuximab at concentrations ranging from 0 to 300 μg/ml. Following 72 h of drug exposure, CellTiter-Glo reagent was added, and luminescence was measured using a GloMax Discover Microplate Reader (Promega). Data are presented as the mean ± standard deviation (SD) from quadruplicate replicates. Immunoblotting Cells were seeded in 60-mm dishes at a density of 3 × 10ˆ5 cells per dish and incubated overnight or until reaching 70% confluence. Cells were then treated with cetuximab (50–100 μg/ml), AT-13387 (10 nM), or a combination of both. After 24 h, cells were harvested, washed twice with ice-cold phosphate-buffered saline (PBS), and lysed in lysis buffer supplemented with protease and phosphatase inhibitors for 30 min. Lysates were sonicated, and particulate material was removed by centrifugation at 13,000 rpm for 10 min at 4 degrees C. The soluble protein fraction was denatured at 95 degrees C for 5 min, separated on a 4% to 12% Bis-Tris Precast Gel (Invitrogen) alongside molecular weight markers, and transferred to a polyvinylidene difluoride (PVDF) membrane. Membranes were blocked for 1 h at room temperature in a blocking buffer (5% bovine serum albumin [BSA] and 1% normal goat serum in Tris-buffered saline [TBS; 137 mM NaCl, 20 mM Tris-HCl, pH 7.6, 0.1% (v/v) Tween 20]). Membranes were incubated overnight at 4 degrees C with primary antibodies diluted in blocking buffer, followed by a 1-h incubation with horseradish peroxidase-conjugated secondary antibodies. After three washes in TBS with 0.1% Tween 20 (TBST), bound antibodies were visualized using enhanced chemiluminescence plus reagent. Relative protein levels were quantified using iBright Imaging Systems or by scanning X-ray films and analyzing with Fiji ImageJ (Version 2.14.0/1.54f) software. All experiments were performed in triplicate, and protein levels were normalized to untreated controls. Immunoprecipitation Cells were trypsinized, washed twice with 1x PBS, and lysed in lysis buffer (1% Triton X-100, 0.1% SDS, 150 mM NaCl, 10 mM sodium phosphate [pH 7.2], 1 mM phenylmethylsulfonyl fluoride, 2 μg/ml aprotinin, 0.2 mM sodium orthovanadate, 50 mM sodium fluoride, 2 mM EDTA, and 20 mM ammonium molybdate) for 30 min on ice. Immunoprecipitation of EGFR and Hsp90 was performed in triplicate as previously described (29). Cell cycle analysis Cell cycle analysis was conducted following established methods (29). Briefly, 1 × 106 cells were seeded in 100-mm dishes with medium containing 3% FBS and incubated overnight. The following day, cells were treated with 300 μg/ml cetuximab, 10 nM AT-13387, or 10 μM erlotinib (positive control) for 24 h. Cells were harvested, washed with ice-cold PBS, and fixed in 70% ethanol at 4 degrees C for 30 min. Fixed cells were washed twice with PBS by centrifugation at 850g for 5 min, treated with 2 mg/ml RNase A for 1 h at 37 degrees C, and stained with propidium iodide (final concentration 1.8 mg/ml) at 4 degrees C in the dark for 2 h. Cell acquisition was performed using a ZE5 Cell Analyzer (Bio-Rad), and data were analyzed with FlowJo software (version 10.10.0). A minimum of 10,000 cells per sample were analyzed in triplicate at the Flow Cytometry Core of the University of Michigan Rogel Cancer Center. Tumor growth studies To generate tumor xenografts, 1 × 106 UMSCC-74B cells were transplanted into the flanks of NSG mice (strain# 572, Charles River). All animal procedures were approved by the University of Michigan Institutional Animal Care and Use Committee (Protocol No. PRO00012197). When the tumors reached a volume of approximately 100 to 150 mm3, mice were randomized into four groups (n = 5 per group): vehicle control, cetuximab, AT-13387, and the combination of cetuximab with AT-13387. Mice received repeated intraperitoneal dosing (cetuximab: 50 mg/kg, intraperitoneally, once weekly; AT-13387: 30 mg/kg, intraperitoneally, twice weekly). Body weight and other signs of toxicity were monitored daily, and tumor volumes were recorded using calipers. Tumor volumes were calculated as follows: volume (mm3) = (L x W2)/2. All animals were euthanized when tumors reached more than 1000 mm3 volume. Time to second tumor doubling (TD2) was calculated as the time in days from the start of treatment to the instance of second tumor doubling. TD2 was summarized using the Kaplan-Meier method, and survival curves were compared using the log-rank test. Immunostaining Specimens for immunohistochemistry (IHC) were prepared with assistance from the Tissue and Histology Core of the University of Michigan. Paraffin-embedded tissue sections were deparaffinized in xylene and rehydrated using serial dilution of ethanol. Antigen retrieval was performed by heating sections in citrate buffer (pH 6.0) for 20 min under microwave. Slides were washed in PBS, blocked for 1 h, and incubated with primary EGFR antibody at 4 degrees C overnight. For proliferation analysis, sections were incubated with the human-specific anti-Ki-67 primary antibody at a dilution of 1:250 overnight at 4°C. The specificity of this antibody for human tumor cells over murine stromal cells ensures that the proliferative index accurately reflects the tumor cell response to therapy. Following PBS washes, sections were incubated with appropriate HRP-conjugated secondary antibodies for 1 h, developed with diaminobenzidine (DAB), and mounted with coverslips. Images were acquired using a DS-Fi1 (Nikon) camera mounted on an Olympus LX71 microscope. H-scores were determined by multiplying the staining intensity (0, none; 1, weak; 2, moderate; 3, strong) by the percentage of positive cells, resulting in a semi-quantitative score ranging from 0 to 300. Values represent the mean ± SD from five random fields per sample. Ethics statement Here, we confirm that all the animal studies conducted here were approved by the University Committee on Use and Care of Animals of the University of Michigan (protocol No. PRO00012197). Statistical analysis Spearman and Pearson correlations were used to evaluate the association between cetuximab IC50 values and pEGFR/EGFR protein expression across HNSCC cell lines. For clonogenic assays, two-sample t-tests were used to compare surviving fractions between the combination group (cetuximab + AT-13387) and each monotherapy group. To assess treatment synergy, a two-factor linear model was fit with main effects for cetuximab, AT-13387, and their interaction (30). In this model, the outcome variable is the surviving fraction; consequently, a negative interaction coefficient represents synergistic cell killing where the combination effect exceeds simple additivity. We have clarified this nomenclature to ensure consistency with standard pharmacological vernacular. The interaction coefficient was examined to assess whether the observed effect of the combination exceeded the expected additive effect of the two monotherapies, with a significant negative interaction effect indicating synergistic cell killing. Cell death percentages were calculated as one minus the surviving fraction, normalized to untreated controls (set as 0% cell death). For in vivo tumor growth, tumor doubling and second doubling (TD2) times were defined as the time from baseline until the tumor volume reached at least two- or four-fold the baseline measurement, respectively. Event times were estimated using linear interpolation, and observations were censored at the last available measurement if the event was not observed. Time-to-doubling and TD2 were compared between treatment groups using Kaplan-Meier curves and global and pairwise log-rank tests. Synergy was evaluated by examining the interaction effect of the treatments in a Cox proportional hazards model, with a negative interaction coefficient indicating synergistic delay in tumor growth. Differences in IHC H-scores and tumor volume kinetics were analyzed using one-way ANOVA followed by Tukey's post hoc test. Statistical significance was defined as a two-sided p-value < 0.05. All statistical analyses were performed using R version 4.5.1. Data availability All data are contained within the manuscript. Supporting information This article contains supporting information. Conflict of interest The authors declare that they have no conflicts of interest with the contents of this article. Acknowledgments We thank the Flow Cytometry Core Facility at the University of Michigan for their assistance with cell cycle analysis. We also acknowledge Dr Thomas Carey for providing the HNSCC cell lines, Promega for the A549-HiBiT cells, and Astex Pharmaceuticals, via the Cancer Therapy Evaluation Program, for providing AT-13387. Author contributions T. S. L., M. K. N., R. K. M., and M. T. writing – review & editing; T. S. L. and M. K. N. supervision; T. S. L. funding acquisition; M. K. N. and R. K. M. validation; M. K. N. funding acquisition; M. K. N., K. S. formal analysis; M. K. N., K. S. data curation; M. K. N. conceptualization; R. K. M. writing–original draft; R. K. M. and M. T. visualization, R. K. M., M. T., and K. S. Methodology; K. S. and M. T. investigation. M. T.: Writing – review & editing. Funding and additional information This work was supported by the National Institutes of Health (NIH) through grant R01CA248310 (awarded to M. K. N. and T. S. L.) and the Department of Defense (DoD) award W81XWH-22-1-0552 (awarded to M. K. N.). Reviewed by members of the JBC Editorial Board. Edited by Alex Toker Supporting information Supplementry Figures References Associated Data Supplementary Materials Supplementry Figures Data Availability Statement All data are contained within the manuscript.