Overcoming Resistance to Cetuximab with Honokiol, A Small-Molecule Polyphenol

The HER family of receptor tyrosine kinases (RTK) includes EGFR, HER2, HER3, and HER4. Ligand binding of these receptors stimulates the kinase domain and allows for hetero- or homodimerization within the family (1). Activation of the HER family of RTKs leads to proliferation of several downstream signaling pathways. Most notably, the HER family activates proliferation via the MAPK and survival by protein kinase B (AKT) pathways (2, 3). Previous research demonstrated that the HER family also impacts metastasis and invasion via other downstream signaling networks (2).

EGFR overexpression is identified in many human cancers including head and neck squamous cell carcinoma (HNSCC; ref. 4), non–small cell lung cancer (NSCLC; ref. 5), colorectal, breast (6), pancreatic (7), and brain cancers (8). The deregulation of EGFR signaling in these cancers leads to poor prognosis in patients. Given the identified role of EGFR signaling in human cancers, several targeted therapies against this receptor have been developed. In particular, a mAb called cetuximab has significant promise against EGFR. Cetuximab binds to the extracellular domain of EGFR and inhibits ligand binding, dimerization, and promotes receptor internalization (9, 10). Cetuximab is approved by the FDA in the treatment of metastatic colorectal cancer and HNSCC (11, 12). In addition, studies show that cetuximab may have clinical benefit in NSCLC (13, 14).

Although cetuximab significantly impacts the management of HNSCC and colorectal cancer, most patients experience intrinsic resistance or develop acquired resistance to cetuximab (15–17). To understand mechanisms of resistance, our laboratory previously established models of acquired resistance to cetuximab (18). This was achieved by continually treating cetuximab-sensitive cell lines with increasing concentrations of cetuximab until single-cell–resistant clones were identified. Further evaluation of the resistant clones revealed that cetuximab-resistant lines exhibit increased activation of HER family members and downstream signaling markers such as MAPK and AKT (18, 19). We also reported that another RTK member, AXL, was the key mediator of cetuximab resistance (20). Furthermore, cetuximab-resistant cells depend heavily on HER2/HER3 signaling and therefore overexpression of HER2/HER3 significantly impacts the survival and proliferation of these cells (21). Resistance to cetuximab has also been thought to arise from copy number alterations or domain-specific mutations in EGFR or due to the key downstream effectors KRAS, BRAF, PIK3CA, and PTEN (22). In the clinical arena, RAS mutations and HER2/MET amplifications were the most frequently detected resistance mechanism to cetuximab in a cohort of 22 metastatic colorectal cancer patients (23). Although acquired resistance to tyrosine kinase inhibitor (TKI) therapy has been attributed to mutations in the cytoplasmic tail of the EGFR, no such correlation exists for cetuximab response (24), and we did not see any EGFR mutations in these cetuximab-resistant cells (18).

Given the clinical ramifications of acquired resistance to cetuximab, overcoming acquired resistance is a critical step in successfully treating metastatic colorectal cancer, HNSCC, and potentially NSCLC. We previously tested several small-molecule inhibitors including gefitinib and erlotinib. We reported that cetuximab-resistant tumors may still rely on EGFR for growth and survival and can be successfully targeted via alternative EGFR TKIs (18, 25). Researchers began investigating natural compounds as a means to overcome resistance in part because there is lower toxicity associated with these compounds when compared with established targeted therapies (26). Honokiol is one of the natural compounds that recently emerged as a promising therapy to overcome drug resistance (27). Honokiol is a small-molecule polyphenol attained from the bark of the Chinese Magnolia tree and is used in traditional Chinese medicine (28). It has been shown to have antiangiogenic (29), antiinflammatory (30), and antitumor properties without appreciable toxicity (31). Recent research revealed Honokiol's anticancer effects such as apoptosis and growth inhibition in leukemia (32), melanoma (33, 34), lung (35), pancreatic (36), and colorectal cancers (37). Honokiol is also established as a potent inhibitor of EGFR, AKT, and MAPK activation, which is thought to contribute to the decrease in proliferation of HNSCC (38). In addition, honokiol effectively suppressed growth in adriamycin- and tamoxifen-resistant breast cancer cell lines (31, 38). Thus, previous research suggests that honokiol is a powerful anticancer agent even in drug-resistant cell lines and inhibits activation of the HER family receptors in many human cancers.

Given the previous data published on honokiol, we sought to determine whether honokiol was a potential agent that could be utilized to overcome acquired resistance to cetuximab and to define the mechanism of this effect. We found that the combination treatment of cetuximab and honokiol inhibited proliferation more robustly than either honokiol or cetuximab alone in cetuximab-resistant NSCLC clones. The decrease in proliferation of these cells could potentially be explained by the decreased activation of HER3, MAPK, AKT, and Dynamin-related protein 1 (DRP1). Furthermore, the inhibition of pDRP1 led to decreased production of reactive oxygen species (ROS), which may contribute to the efficacy of honokiol treatment. In vivo, the combination treatment exhibited a robust inhibition of HNSCC patient derived xenograft (PDX) tumor growth. Collectively, our findings suggest that honokiol may be an effective therapy in overcoming acquired resistance to cetuximab in NSCLC and HNSCC.

Drs. J. Minna and A. Gazdar (University of Texas Southwestern Medical School, Dallas, TX) provided the human NSCLC cell line H226. The H226 cells were maintained in 10% FBS in RPMI1640 (Mediatech, Inc) with 1% penicillin and streptomycin. The development of cetuximab-resistant cells has been described previously (18). The validity of these cell lines was consistently verified on the basis of cell morphology and genomic short tandem repeat (STR) profile of each cell line. Four to 6 months prior to the initiation of our investigation, H226 cells were tested for authenticity in agreement with ATCC guidelines.

Cetuximab (IMC-225, Erbitux) was purchased from the University of Wisconsin Pharmacy. Honokiol was generously provided from Dr. Jack Arbiser at Emory University (Atlanta, GA). Intralipid was purchased from Santa Cruz Biotechnology, Inc.

All antibodies were purchased from commercial sources as indicated below: EGFR, HER2, HER3, and HRP-conjugated goat anti-rabbit IgG and goat anti-mouse IgG were obtained from Santa Cruz Biotechnology, Inc. pMAPK (T202/Y204), pAKT (S473), AKT, prpS6 (S235/236), rpS6, pDRP1 (S616), DRP1, pC-RAF (S338), C-RAF, were obtained from Cell Signaling Technology. α-tubulin was purchased from Calbiochem.

This assay was performed as described previously (39).

The fractional product method was used for analysis of interactions in cetuximab-resistant cells (40, 41). Briefly, this method utilizes the relative cell density following treatment with each individual agent and calculates the expected (E) effect of combination therapy as a product of the individual responses. The observed (O) effect is the relative cell density following dual treatment. A ratio of the observed to expected (O:E) values was calculated and used to estimate the synergy, additivity, or antagonism. A value less than 1 indicated synergism, greater than 1 demonstrated antagonism, and 1 represented additivity.

Whole-cell protein lysate was attained using Tween-20 lysis buffer [50 mmol/L HEPES, pH 7.4, 150 mmol/L NaCl, 0.1% Tween-20, 10% glycerol, 2.5 mmol/L EGTA, 1 mmol/L EDTA, 1 mmol/L DTT, 1 mmol/L Na₃VO₄, 1 mmol/L phenylmethylsulfonylfluoride, 1 mmol/L b-glycerophosphate (BGP), and 10 mg/mL leupeptin and aprotinin]. Immunoblot analysis was performed as described previously (42).

Using an antibody-based array from R&D Systems, the activity of several RTKs were screened. This phospho-RTK array selectively screens for proteins/receptors associated with the membrane. Antibodies against 42 RTKs were prespotted in duplicate on nitrocellulose membranes. Cetuximab-resistant HC4 cells were treated with vehicle, cetuximab (100 nmol/L), and/or honokiol (20 μmol/L) for 24 hours, and cell lysates were collected and incubated with the prepared membrane. Phosphorylated tyrosines on activated RTKs were detected using a pan-anti-phosphotyrosine antibody. Scanned images were quantified by Image Quant software and relative expression of specific phosphorylated RTKs was determined.

Cells were plated at a density of 8 × 10⁵ cells per 100-mm² plate and allowed 24 hours to adhere. The cells were treated with vehicle, 100 nmol cetuximab, 20 μmol/L honokiol, or 100 nmol cetuximab and 20 μmol/L honokiol in combination for 24 hours. Bromodeoxyuridine (BrdUrd) cell distribution analysis was executed as described previously (43).

Approximately 30,000 cells/well were plated on a four-well glass chamber slide (Millipore) for 24 hours. After treatment, cells were briefly rinsed in PBS and fixed at room temperature with 4% methanol-free formaldehyde for 15 minutes. Cells were washed with PBS again and permeabilized with PBS containing 0.2% TritonX-100 for 15 minutes. Cells were blocked in 5% normal goat serum diluted in PBS containing 0.3% Triton X-100 for 1 hour at room temperature, and primary pDRP1 antibody was incubated overnight at 4°C. The next day, cells were washed with PBS three times, and incubated with Alexa Fluor-488 rabbit secondary antibody for 30 minutes at room temperature (Life Technologies). Cells were washed in PBS again and DAPI was applied for 5 minutes. Cells were then mounted with ProLong gold mounting solution (Life Technologies). An Olympus BX41 microscope and Olympus DP25 camera was used for fluorescent microscopy and photography. Quantified average cell fluorescence of approximately 50 cells per treatment arm using ImageJ analysis.

Approximately 30,000 cells/well were plated on a four-well glass chamber slide (Millipore). Cells were washed in PBS 24 hours later and treated with vehicle, cetuximab (100 nmol/L), and/or honokiol (20 μmol/L). Again, 24 hours later, cells were washed with PBS again. Cells were treated with 3μmol/L MitoSox in Hank balanced salt solution for 30 minutes. Cells were washed in PBS and then fixed at room temperature with 4% methanol-free formaldehyde for 15 minutes. Cells were washed with PBS and treated with DAPI (Molecular Probes) for 5 minutes. Slides were mounted with Prolong gold mounting solution (Life Technologies). An Olympus BX41 microscope and Olympus DP25 camera was used for fluorescent microscopy and photography. Quantified average cell fluorescence of approximately 20 cells per treatment arm using ImageJ analysis.

Athymic nude mice (4- to 6-week-old females) were acquired from Envigo. All institutional guidelines of the University of Wisconsin were in accordance with all animal procedures and maintenance. Cetuximab-resistant UW-SCC1 PDX tumors were inoculated in the dorsal flank of the mouse at respective day 0. Tumors were allowed to reach 200 mm³ before mice were randomized into treatment or control groups and given their respective treatments (Vehicle, cetuximab, honokiol, or combination). Cetuximab (0.2 mg/mouse) was given twice a week and/or 3 mg/mouse/day of honokiol dissolved in 20% intralipid given three times a week by intraperitoneal injection. Tumor volume was assessed by digital caliper measurements and calculated by the formula (π)/6 (large diameter) (small diameter)². The details of our PDX model were described previously (44). Twenty-four hours after the last cetuximab and/or honokiol treatment, tumors were harvested. Details of tumor collection and protein isolation were described previously (43). Protein quantitation and immunoblot analysis were completed following procedure as stated above.

Statistical analyses were performed using Prism 7 software (GraphPad Software, Inc.). Paired t test analysis was performed for in vivo study and P ≤ 0.01 was considered significant at the 99% confidence level as shown by the asterisks (**).

We previously established a model of cetuximab resistance using the NSCLC H226 cell line. Cetuximab-sensitive H226 clones were continually treated with increasing doses of cetuximab and we selected clones based on continued proliferation despite treatment (18). Cetuximab-resistant clones (HC1, HC4, HC8) were found to have increased activation of EGFR/HER2/HER3 and subsequent increased activation of the MAPK/AKT pathways. Given these results, we hypothesized that honokiol could have potent antiproliferative affects in cetuximab-resistant NSCLC lines. To test this hypothesis, we performed a comparative cell proliferation assay looking at cetuximab-resistant clones (HC1, HC4, HC8) versus the parental, cetuximab-sensitive, control (HP) following 72 hours of treatment with increasing concentrations of honokiol (Fig. 1A). This revealed a statistically significant, dose-dependent growth inhibition of all resistant clones. Treatment with 25 μmol/L of honokiol resulted in approximately 30%–50% decrease in proliferation, whereas proliferation of the parental line remained relatively stagnant despite increased concentrations of honokiol. The IC₅₀ value of honokiol is 32, 24, 24, and 23 μmol/L in HP, HC1, HC4, and HC8 cells, respectively. To determine honokiol's effect on the HER family members and downstream signaling, we analyzed the expression and activity of the HER family networks in cetuximab-resistant clones (HC1, HC4, HC8) versus the cetuximab-sensitive parental line (HP) after treatment with increasing concentrations of honokiol (Fig. 1B). Phosphorylation of all three receptors decreased after 25 μmol/L of honokiol. In addition, phosphorylation of downstream cascades in resistant clones (HC1, HC4, HC8) specifically C-RAF, MAPK, AKT, and ribosomal protein S6 (rpS6) was inhibited after treatment with honokiol. HER3 total expression also decreased following 25 μmol/L of honokiol; however, total expression of HER2 and EGFR remained steady following honokiol treatment. Collectively, these findings suggest that honokiol effectively decreases proliferation of cetuximab-resistant clones which may be explained by the decreased activation of HER family receptors and downstream signaling networks.

To determine whether honokiol could resensitize cetuximab-resistant clones to cetuximab treatment, we treated cetuximab-resistant clones (HC1, HC4, HC8) with vehicle, cetuximab (100 nmol/L), honokiol (20 μmol/L) or the combination (Fig. 2A). In all cetuximab-resistant clones, cell proliferation decreased by approximately 40% after dual treatment as compared with the vehicle control. Furthermore, synergistic interactions between cetuximab and honokiol were found in all cetuximab-resistant clones we tested as analyzed by the fractional product method (Fig 2A). The honokiol treatment of the cetuximab-sensitive parental control (HP) did not elicit any additional effects to the cetuximab treatment. To determine the effect of combination treatment on the activation of RTKs, an RTK array of a cetuximab-resistant clone, HC4, was preformed following vehicle, cetuximab, honokiol, or combination treatment (Fig. 2B). The combinatorial treatment showed an approximately 20% decrease in activation of HER2 and HER3. This was greater inhibition than either treatment alone. Interestingly, activation levels of insulin receptor (IR), insulin growth factor 1 receptor (IGF1R), and c-Ret were increased after honokiol treatment (Fig. 2B). This activation, however, appeared to be downregulated in the combination arm. We performed an immunoblot analysis after the same treatment regime as stated above in HP, HC1, HC4, and HC8 to examine the expression and activity of downstream signaling (Fig. 2C). Activation of HER2, HER3, and downstream signaling markers robustly decreased after combination treatment, more so than either treatment strategy alone. The most significant inhibition of phosphorylation was found in MAPK, AKT, rpS6, and C-RAF. Across all cetuximab-resistant lines (HC1, HC4, HC8), total expression of the HER family and downstream markers remained constant across treatments. Furthermore, combination treatment induced a G₁-phase cell-cycle arrest in cetuximab-sensitive and -resistant lines (Fig. 3). Dual treatment led to the greatest inhibition of the S-phase followed by honokiol and then cetuximab in all cetuximab-resistant lines. Overall, combination treatment of honokiol and cetuximab in resistant clones led to robust inhibition of proliferation and activation of the HER family network and induced a G₁-phase cell-cycle arrest.

Activation of the RAS/RAF/MAPK pathway alters mitochondrial function through the activation of DRP1 (45). Activated DRP1 promotes mitochondrial fission which increases mitochondrial function as the mitochondria have increased surface area after fission. In addition, increased mitochondrial function is shown to contribute to the tumor growth of oncogenic MAPK-dependent cancers (45). Inhibition of DRP1 could affect the rate of mitochondrial fission which could decrease mitochondrial output, limiting the energy available for proliferation. Recent data suggests that honokiol downregulates mitochondrial function in melanoma (34). Given the decrease in proliferation and activation of C-RAF and MAPK after honokiol treatment, we hypothesized that honokiol may inhibit phosphorylation of DRP1 in NSCLC cetuximab-resistant clones. We compared the phosphorylation of DRP1 after cetuximab, honokiol, and combination treatment in cetuximab-sensitive and -resistant NSCLC clones (Fig. 4A). Interestingly, we found that the parental cetuximab-sensitive line had low endogenous levels of pDRP1 while all cetuximab-resistant lines had higher endogenous levels of pDRP1 (Fig. 4A). Results from this study indicated that the combination of honokiol and cetuximab treatment led to greater inhibition of pDRP1 when compared with either treatment alone. These results were corroborated by immunofluorescent analysis of pDRP1 (Fig. 4B). Quantification of average cell fluorescence revealed that HC4 clones had almost 2 times the amount of pDRP1 fluorescence after vehicle treatment. In addition, it was shown that pDRP1 fluorescence had approximately 65% decrease after combination treatment. These results suggest that cetuximab-resistant clones have increased activation of DRP1 endogenously. To examine the changes in mitochondrial function after combination treatment, we used MitoSox immunofluorescence (red) to stain ROS in cetuximab-resistant (HC4) and sensitive clone (HP). Interestingly, the cetuximab-resistant clone (HC4) had increased endogenous levels of ROS compared with the cetuximab-sensitive clone (HP). Quantification of average cell fluorescence showed that HC4 clones contained nearly 2.5× the amount of ROS staining after vehicle treatment. The combination treatment led to the most robust decrease (∼70%) in ROS staining in our cetuximab-resistant clone (Fig. 4C). Remarkably, both pDRP1 and ROS staining increased after cetuximab treatment in our cetuximab-resistant clone. Combination treatment of honokiol plus cetuximab effectively inhibited the phosphorylation of DRP1 and thus suppressed the production of ROS.

Given the promising in vitro data and the established effects of pDRP1 on tumor growth (46), we tested the efficacy of a combination treatment of cetuximab and honokiol in vivo (Fig. 5A). Cetuximab-resistant UW-SCC1 PDX tumors (44) were inoculated in each dorsal flank of the mice on day 0 and were allowed to reach 200 mm³ before animals were randomly assigned to treatment groups (vehicle, cetuximab, honokiol, or combination). Thirty-two mice were analyzed: 8 per treatment group. Each group was treated with cetuximab (0.2 mg/mouse twice weekly), honokiol (3 mg/mouse three times/week), or combination that was administered through intraperitoneal injection for 30 days. This experiment demonstrated that tumor growth was significantly inhibited when compared to vehicle or either treatment alone. Average volume of tumors after 30 days of vehicle treatment was approximately 1,000 mm³, after cetuximab treatment approximately 900 mm³, after honokiol treatment approximately 600 mm³, and after combination treatment approximately 250 mm³. Importantly, the size of the tumors after 30 days of combination treatment was approximately the same as the volume when treatment began. Further evaluation of tumor samples revealed a similar downregulation of HER family phosphorylation and activation of downstream signaling markers as seen in vitro (Fig. 5B). Specifically, activation of HER3, C-RAF, MAPK, AKT, and rpS6 was effectively inhibited following the combinatorial treatment. pDRP1 expression was also inhibited in tumor lysates after combination treatment. Collectively, these data suggest that combination treatment of cetuximab and honokiol inhibits phosphorylation of the HER family, downstream signaling cascades, and DRP1. This may offer an explanation for the effective growth delay observed in the HNSCC PDX tumors.

Cetuximab, an mAb against the EGFR, is approved by the FDA for treatment of metastatic colorectal cancer and HNSCC (12, 47). In addition, recent studies demonstrate that it is also a promising therapy for NSCLC (13, 14). Although some patients benefit from cetuximab treatment, many patients are intrinsically resistant or develop acquired resistance to the drug. To study potential mechanisms of acquired resistance, our laboratory previously developed a model of acquired resistance in NSCLC and HNSCC. Further evaluation revealed that cetuximab-resistant clones overexpressed HER family members as well as MAPK and AKT (18, 21, 43). A recent study conducted by Leeman-Neill and colleagues concluded that honokiol, a small-molecule polyphenol, decreased phosphorylation of MAPK and AKT in HNSCC. In addition, honokiol is effective in overcoming drug resistance in myeloma, melanoma, and breast cancer (31, 33, 34). It also elicited robust anticancer activity in leukemia (32), melanoma (33, 34), lung (35), pancreatic (36), and colorectal cancers (37) without appreciable toxicity. Previous research also demonstrated that honokiol overcame drug resistance in multiple cancers and downregulated proteins that are abundant in our acquired resistant models. This provided justification for our current study evaluating the efficacy of honokiol in overcoming acquired resistance to cetuximab in NSCLC.

In this study, we found that combination treatment of cetuximab and honokiol had a robust antiproliferative affect in cetuximab resistant clones in vitro and in vivo (Figs. 2A and 5A). This can largely be attributed to the down regulation of cell proliferation and survival pathways that are hyperactive in cetuximab-resistant clones. Dual treatment with honokiol and cetuximab showed significant decreases in the levels of pHER2, pHER3, pMAPK, pAKT, and pC-RAF (Fig. 2C). This inhibition was greater than either treatment alone in the cetuximab-resistant setting. Concurrently, it has been shown that although targeting AKT using MK-2206 resulted in growth inhibition, the greatest growth inhibition was achieved after dual treatment of MK-2206 and cetuximab in NSCLC cetuximab–resistant lines (19). Recent studies observed an inhibition of pMAPK in HNSCC, melanoma, and myeloma after honokiol treatment (33, 34, 38). In addition, inhibition of pAKT was demonstrated in HNSCC, myeloma, and breast cancers after honokiol treatment (31, 33, 38). These results corroborate our findings. pMAPK and pAKT inhibition might offer a further explanation for the G₁ cell-cycle arrest after dual treatment found in our study (Fig. 3). Decreased phosphorylation of MAPK and AKT were also found in the in vivo setting (Fig. 5B). Inhibition of pAKT and pMAPK and G₁ phase cell-cycle arrest after combination treatment helps explain the delayed PDX tumor growth found in our experiment. Furthermore, honokiol has also been found to have antitumor properties in skin, lung, breast, ovarian, prostate, colorectal, and even malignant gliomas (27). These data suggest that across a wide variety of cancers, honokiol successfully inhibits both pMAPK and pAKT. This inhibition likely contributes to the decrease in cell proliferation and delayed tumor growth found in cetuximab-resistant cells and PDX tumors after combination treatment of cetuximab and honokiol.

In addition to pMAPK and pAKT downregulation, combination treatment also inhibited pC-RAF and pDRP1 expression (Fig. 4A). Kashatus and colleagues showed that phosphorylation of DRP1 by pMAPK can lead to mitochondrial fission and increased mitochondrial function. A similar mechanism proposed by Tailor and colleagues also links pDRP1 to mitochondrial fission (48). Furthermore, honokiol can alter mitochondrial function in lung cancer (49) and can inhibit pDRP1 expression in melanoma (34). Our results supported these mechanisms as we observed decreased activation of C-RAF, MAPK, and DRP1 following honokiol and combination treatments. We speculate that mitochondrial fission and subsequent increased mitochondrial function contributes to the survival and proliferation of cetuximab-resistant clones and PDX tumors as the mitochondria are providing energy for these activities. The balance between mitochondrial fusion and fission has significant implications on mitochondrial function as changes to morphology affect the surface area and the electron transport chain which could alter the production of reactive oxygen species. Important mitochondrial function including proliferation, metabolic efficiency, apoptosis, and autophagy are therefore impacted by the balance between mitochondrial fission and fusion. Previous research also revealed that downregulation of DRP1 led to decreased cell proliferation in lung cancer and colorectal cancer (46, 48, 50). This could further provide an explanation for the decrease in proliferation found in cetuximab-resistant clones and PDX tumors after combination treatment.

Our study also found that decreased pDRP1 expression after honokiol and cetuximab treatment correlated with decreased ROS production (Fig. 4). The antioxidant properties of honokiol were previously established, although the biological ramifications of ROS production are still debated (27, 30). It is speculated that increased ROS production in cells could lead to apoptosis due to increased oxidative stress (49). Alternatively, ROS-driven tumors with increased MAPK activation have been identified in tobacco-induced oral and lung carcinoma and other cancers caused by common viruses such as Esptein–Barr virus and hepatitis C virus. These cancers have been found to use reactive oxygen in molecular signaling pathways that can contribute to the increased survival of these tumors (51). Results from our current study demonstrated ROS inhibition following combination treatment of honokiol and cetuximab (Fig. 4C). This inhibition was also correlated with decreased proliferation, survival, G₁ phase cell-cycle arrest, and delayed PDX tumor growth.

Collectively, the results reported indicate that combination treatment of cetuximab and honokiol successfully leads to decreased proliferation and survival of cetuximab-resistant clones and decreased tumor growth of intrinsically resistant HNSCC PDX tumors. This was likely achieved through the decreased phosphorylation of AKT, MAPK, S6rp, and DRP1. Our previous findings targeting other members of the HER family such as HER3 (42) or using a pan-HER antibody targeting all family members (21) as well as other key mediators such as AKT (19) can overcome resistance to cetuximab due to the compensatory mechanisms that allow for sustained proliferation and survival upon inhibition of both AKT or MAPK signaling pathways. Although work has shown that honokiol can enhance cetuximab, this is the first study to identify honokiol as a potential therapy to overcome acquired cetuximab resistance and is also the first study to examine the link between honokiol, mitochondrial function, and ROS in this setting. Given honokiol's wide-ranging affects in multiple cancers, low appreciable toxicity, and the ability to overcome drug resistance, it is a promising agent for the future of cancer therapy.

R.J. Kimple reports receiving commercial research grants from Threshold Pharmaceutical and Peloton Therapeutics. No potential conflicts of interest were disclosed by the other authors.

Conception and design: H.E. Pearson, M. Iida, J. Arbiser, D.L. Wheeler

Development of methodology: H.E. Pearson, M. Iida, R.A. Orbuch

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): H.E. Pearson, M. Iida, K.P. Nickel, R.J. Kimple, J. Arbiser

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): H.E. Pearson, M. Iida, R.A. Orbuch, R.J. Kimple, J. Arbiser, D.L. Wheeler

Writing, review, and/or revision of the manuscript: H.E. Pearson, M. Iida, R.A. Orbuch, N.K. McDaniel, R.J. Kimple, D.L. Wheeler

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): N.K. McDaniel

Study supervision: D.L. Wheeler

This work was supported in part by Department of Human Oncology Seed Grant and the University of Wisconsin Carbone Cancer Center Support Grant P30 CA014520. This study was funded by National Institutes of Health grant RO1 AR47901 (JLA).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.