Introduction

Chemotherapy regimens derived from results of clinical trials are valuable for determining optimal treatment options for large populations of patients. However, an individual patient’s response to chemotherapy can be very different from the predicted response of the average population, and the reasons for this variation are largely unknown. Several clinical and pathological factors have been identified as having prognostic value of treatment outcome and survival of cancer patients, e.g., tumor size, lymph node and distant metastasis, tumor grade, and, more recently, specific molecular biomarkers. These prognostic factors help to classify the standard risk of subpopulations of patients with the same tumor entity, but are still unable to predict the response of specific individuals to therapy. Therefore, there is an urgent need for reliable molecular tests to predict the individual patient’s risk of death from the disease irrespective of the treatment (prognostic markers) and sensitivity or resistance to chemotherapy (predictive markers). Such tests are necessary to develop individualized treatment schedules in the future. The field of chemotherapy is currently undergoing a paradigm shift from classical cytotoxic chemotherapy towards targeted therapy with the aim to eradicate tumor cells more efficiently with fewer side effects on normal tissue. Proteins encoded by genes carrying tumor-specific mutations serve as preferential targets for the development of novel drugs in cancer therapy.

It is now well accepted that mutations in three main types of genes contribute to carcinogenesis: oncogenes, tumor suppressor genes, and stability genes. Oncogene activation (gain-of-function mutation) results from point mutations, chromosomal translocation, or gene amplification. Without such mutations, tumors cannot grow (tumor addiction). On the other hand, mutations in tumor suppressor genes are loss-of-function mutations, e.g., missense mutations, chromosomal deletions or insertions, or epigenetic silencing, that allow the tumor to grow unchecked by normal cellular control mechanisms. Stability genes or caretakers including DNA repair genes controlling genomic stability and genes responsible for organizing mitotic recombination and chromosomal segregation. Inactivating mutations in these genes are dangerous because they increase the mutational rate in other genes. Out of the large number of potential targets for targeted chemotherapy, we focus on the epidermal growth factor receptor family in the present overview. Growth factor receptors have a tremendous relevance in cancer biology. Therefore, the therapeutic intervention to silence the function of epidermal growth factors and their related signaling pathways represents a highly attractive approach to improve treatment success of solid tumors.

Epidermal growth factor receptors in cancer biology

There are four human epidermal growth factor receptors (EGFR/ERBB1/HER1, HER2/ERBB2/c-neu, HER3, and HER4). After ligand binding, they activate downstream signaling routes, which regulate proliferation, differentiation, apoptosis, metastasis, and angiogenesis. Their 3D structures are represented in Fig. 1. EGFR and HER2 are over-expressed in many solid tumors, which is associated with unresponsiveness to chemo- and radiotherapy as well as short survival times of patients (see below) [1]. The heterodimer structure of EGFR/HER2 is depicted in Fig. 2. Thus far, 10 ligands have been identified, i.e., the epidermal growth factor family (EGF, transforming growth factor-α, β-cellulin, epiregulin, HB-EGF, AR) and the neuregulin family (heregulin, neuregulins) [2, 3]. Upon binding of a ligand to an EGFR monomer, homo-dimerization takes place with a second EGFR molecule or with another HER member. Similarly, HER2 can dimerize with HER3 or HER4 and HER3 with HER4. Ligand binding and dimerization leads to intracellular phosphorylation of HER receptors and thereby activation of the downstream signaling pathways. The existence of 10 ligands of different homo- and heterodimers consisting of four receptors create a considerable flexibility and complexity for signal transduction [2–4]. This complexity is even further increased by varying the duration and strength of receptor signaling, receptor internalization, and recycling as well as rates of phosphorylation and dephosphorylation [5].

Fig. 1
figure 1

3D structure of human HER1/EGFR, HER2, HER3, and HER4. Extracellular and cytoplasmic domains of the proteins were retrieved from protein data bank (PDB) database and combined structures were formed with PyMol software

Full size image
Fig. 2
figure 2

Heterodimer of human HER1/HER2. Extracellular and cytoplasmic domains of the proteins were retrieved from PDB database and combined structures were formed with PyMol software

Full size image

Dimerization stimulates intrinsic tyrosine kinase activity of EGFR, which regulates specific signal transduction cascades, e.g., Raf/Mek/Erk, PI3K/PDK1/Akt, PLCγ/PKC, MAPK, and JNK signaling routes. Constitutive EGFR activation as consequence of point mutations or gene amplification causes deregulated cellular processes such as proliferation, invasion, angiogenesis, cell motility, cell adhesion, inhibition of apoptosis, and DNA synthesis. The kinase activity is also associated with autophosphorylation of five tyrosine residues in the C-terminal EGFR domain. Mutations affecting EGFR expression foster carcinogenesis.

The extraordinary relevance of EGFR in tumor biology makes it an exquisite molecular target for tumor therapy. Apart from therapeutic antibodies, several small molecules have been developed as EGFR inhibitors [6]. For example, gefitinib (Iressa®; Astra Zeneca, DE, USA) and erlotinib (Tarceva®; OSI-774, Genentech Inc., CA, USA) are first-generation inhibitors used for the treatment of non-small cell lung cancer and other tumor types [7]. Both quinazolinamines exhibit their inhibitory activity by competing with ATP for the ATP binding pocket of EGFR.

Despite considerable successes with these EGFR tyrosine kinase inhibitors in cancer therapy, resistance against these chemical compounds develop due to the selection of point-mutated EGFR variants [8]. Therefore, there is an urgent need for the identification of novel EGFR tyrosine kinase inhibitors. In recent years, medicinal plants came into the center of interest as resources for novel treatment strategies to target EGFR family members.

Role of epidermal growth factor receptors for drug resistance and patient prognosis

EGFR-expressing cell lines

The connection between EGFR and classical cytotoxic drug resistance has been known for more than two decades. Murine sarcoma 180 (S180) cells selected for resistance towards doxorubicin overexpress EGFR compared to drug-sensitive wild-type S180 cells [9, 10]. As subsequently shown, EGFR expression also plays a role for drug resistance of tumor cells not previously selected treated with cytostatic drugs. Since kidney carcinomas are frequently unresponsive to chemotherapy, they represent a suitable model to study inherent drug resistance. EGFR expression of human primary cell cultures of renal cell carcinomas of 18 patients subjected to hierarchical cluster analyses showed that the expression of c-ErbB1 and c-ErbB2 was higher in resistant cell cultures compared to sensitive cell cultures [11]. EGFR is involved in drug resistance by affection of apoptosis, DNA repair, or the induction of resistance gene expression [12]. These in vitro results were translated to clinical tumors. Tumors with EGFR expression were significantly more frequent resistant to doxorubicin than EGFR negative or weakly expressing cancers [13, 14].

Glioblastoma multiforme (GBM) is the most aggressive form of adult human brain tumor [15]. Malignant gliomas often show resistance to adjuvant radio- and chemotherapy due to the accumulation of genetic alterations that cause oncogene activation, e.g., EGFR [16]. In most GBMs, amplification and rearrangement of the EGFR gene resulted in mutant receptors, called ΔEGFR (kinase-deficient mutant EGFR) that enhanced tumorigenicity in vivo, and caused cisplatin resistance [17].

In addition to cisplatin, EGFR also reduced the activity of microtubule poisons, i.e., vincristine and paclitaxel [18]. Combination treatment of human ΔEGFR-expressing GBM cells with EGFR-directed tyrosine kinase inhibitor and cisplatin synergistically induced apoptosis in vitro and in vivo [17, 18].

Furthermore, the combination treatment of a c-Met kinase inhibitor and either an EGFR kinase inhibitor or cisplatin enhanced cytotoxicity of mutant EGFR-expressing GBM cells [19]. Taken together, EGFR expression is associated with drug resistance in vitro making it an exquisite target for novel drugs inhibiting EGFR function. EGFR-mediated resistance is also known for cytotoxic natural products. Two interesting bioactive compounds derived from the Chinese coniferous tree Cephalotaxus hainanensis are cephalotaxine (CET) and its ester homoharringtonine (HHT) [20]. Although HHT possessed the highest growth inhibitory activity towards human leukemic cells [21], human ΔEGFR-expressing GBM cells were 14-fold more resistant to HHT than control cells [22]. These findings indicated a causative role of mutation-activated EGFR for cellular resistance towards CET and HHT. Similar results have been obtained for artesunate (ART), which is a semisynthetic derivative of artemisinin, the active principle of Artemisia annua L. This herb was traditionally used in Chinese medicine for the treatment of fever and chills. Nowadays, artemisinin is used as anti-malarial drug [23]. Artemisinin and its derivative ART also reveal profound anti-cancer activity [24–26]. In sum, drug resistance mediated by EGFR is not restricted to established anticancer drugs but also occurs towards other cytotoxic compounds of natural origin. Hence, EGFR-mediated resistance may represent a general type of cellular defense mechanisms towards a broad range of toxic xenobiotics.

EGFR in clinical tumors

EGFR is expressed in different human tumors, e.g., in cancers of the lung, head and neck, colon, pancreas, breast, ovary, bladder and kidney, and gliomas. EGFR expression correlated is of prognostic significance for cancer patients. Patients with EGFR-overexpressing tumors reveal worse prognosis [27]. To illustrate this, we exemplarily focus on lung cancer in more detail.

Lung cancer is the leading cause of cancer mortality worldwide, and cure rates are less than 15 % [28]. Lung cancers are classified into two histological types: small cell carcinoma and non-small cell carcinoma. The majority of bronchogenic carcinomas can be classified into four histological types: small cell carcinoma, adenocarcinomas, squamous cell lung carcinomas, and large cell carcinomas. The histological features, clinical course, and response to therapy indicate that small cell lung carcinoma (SCLC) is a separate entity. The behavior of the other three histological subtypes is similar and therefore is referred to as non-small cell lung carcinoma (NSCLC) [29]. NSCLC represent the majority of lung cancers and are usually associated with poor prognosis. While SCLC is drug sensitive, NSCLC are the opposite. Clinical oncology still regards resistance to chemotherapy NSCLC patients as a major problem. As numerous mechanisms are operative in drug-resistant tumors [11, 30, 31], the relative quantitative contributions of each of these resistance mechanisms have to be determined. Hence, understanding the complex genetic network in clinically relevant drug resistance needs more holistic approaches to the entire battery of genes conferring drug resistance.

In 81 human primary squamous cell lung carcinomas, EGFR expression was described as prognostic factor [13]. EGFR expression level was reduced in NSCLC patients with long-term survival [32]. In addition, down-regulation of HER2 expression played an important role in resistant NSCLC [33]. Interestingly, carcinomas of smokers expressed EGFR more frequently than carcinomas of nonsmokers do [34]. The importance of EGFR signaling in lung cancer and its beneficial effects on patient survival led to clinical usage of the EGFR inhibitor erlotinib for the treatment of NSCLC [33].

Taken together, these data clearly speak for an important role of EGFR in drug resistance in vitro as well as in the clinical setting. This is why it appears an exquisite target for novel drugs specifically inhibiting EGFR function and signaling.

HER2, HER3, and HER4 in clinical tumors

HER2 overexpression plays a major role in breast cancer, but it can be also found in other tumor types. HER2 positivity in breast cancers varies from 10 to 40 % [35–38]. The overexpression of HER2 mRNA and protein is a poor prognostic factor [39, 40] and correlated with poor responsiveness to chemotherapy [36]. While EGFR and HER2 have been intensively studied during the past years, less data are available for HER3 and HER4. The prognostic significance of HER3 has been discussed in a contradictory manner. Some authors reported associations of HER3 expression to poor prognosis in breast cancer patients, while others described HER3 as a favorable prognostic factor [41]. HER4 mediates anti-proliferative and differentiation effects [35]. Hence, it is plausible that this receptor represents a favorable prognostic marker in breast cancer patients [41–43].

Since the development of monoclonal antibody C225 to treat EGFR-positive cancers [44], many other therapeutic antibodies and small molecule tyrosine kinase inhibitors against EGFR and HER2 have been developed [45]. In contrast, HER4 does not serve as target for drug development because of its positive prognostic significance. While EGFR- or HER2-overexpressing cancers are adverse prognostic factors if standard cytotoxic chemotherapy is applied, the contrary occurs upon application of EGFR or HER2 inhibitors. Tumors with high EGFR or HER expression are preferentially killed by such targeted antibodies and small molecule inhibitors [37]. This is an instructive example how the specific therapeutic targeting of proteins with worse prognosis can be exploited to improve treatment success rates. Unfortunately, tumors can also develop resistance against EGFR- or HER2-directed antibodies and small molecules, and the search for novel drugs to fight cancer continues.

In this context, the tremendous chemodiversity of phytochemicals comes into play. Novel compounds from natural sources may serve as lead compounds for a new generation of drugs eradicating resistant tumors.

Inhibition of epidermal growth factor signaling by phytochemicals

Natural products as resource for cancer treatment

As pointed out by a survey of the National Cancer Institute, USA, the majority of established cancer drugs are natural products, derivatives of natural products, or drugs mimicking the mode of action of natural products [46]. Searching in nature for novel scaffolds is a promising way to find new chemical tools to bypass and overcome such drug resistance. Novel natural product inhibitors may serve as lead compounds for drug development. A plethora of data in the literature shows that natural products can serve as inhibitors for EGFR-associated signaling molecules such as the RAS/RAF/MEK/ERK and PI3K/AKT/mTOR pathways. This indicates that the identification of novel inhibitors from natural resources is not beyond the scope of expectations.

Inhibitors of EGFR signaling

Phytochemicals from different chemical classes such as flavonoids, terpenoids, and alkaloids have been shown to exert their cytotoxic activity towards cancer cells by affecting EGFR signaling (Table 1). Some specific compounds were intensively investigated such as genistein, curcumin, quercetin, resveratrol, and (−) epigallocatechin-3-gallate. The frequent observation that natural products act in a multifactorial manner [47] applies here as well. As shown in Table 1, phytochemicals inhibited both phosphorylation and expression of EGFR (by ubiquitination and degradation). Furthermore, natural compounds inhibited the phosphorylation of downstream kinases either as consequence of EGFR inhibition or by binding of compounds to corresponding kinase domains of signal transducers. In addition, translocation of kinases (e.g., ERK, MAPK) from the cytosol to the nucleus can be blocked by some compounds. As consequence of silencing EGFR signaling routes, various effects were observed in cancer cells, e.g., induction of cell cycle arrest and apoptosis, inhibition of cell mobility, and inhibition of invasion of metastasis.

Table 1 Phytochemicals with activity against EGFR
Full size table

It is important to note that several compounds have been shown to exert their effects not only in vitro but also in vivo, e.g., curcumin, (−) epigallocatechin-3-gallate, 11,11′-dideoxy-verticillin, quercetin, deguelin, proanthocyanidins, luteolin, artesunate, platycodin D, berberine, capsaicin, and delfinidin. Further evaluation of the compounds mentioned in Table 1 in terms of EGFR inhibition was performed with in silico molecular docking analyses on human EGFR tyrosine kinase domain. Molecular docking analyses in silico on human EGFR tyrosine kinase domain revealed silibinin to interact with comparable binding energies as the known inhibitor, lapatinib with similar docking poses (Fig. 3). Anti-tumor activity in vivo represents a precondition to consider compounds for clinical application. It is also interesting to study the interaction of natural products with anticancer drugs. Phytochemicals caused increased or even synergistic inhibition of tumor growth in combination with established drugs. This has been shown for the combinations of curcumin plus gefitinib/erlotinib, honokiol plus cetuximab, and (−) epigallocatechin-3-gallate plus 5-fluorouracil erlotinib/gefitinib (Table 1). Furthermore, natural products can reduce the side effects of standard anticancer therapy on normal organs as shown by the combination of curcumin and gefitinib, which led to reduced gastrointestinal side effects compared to gefitinib alone in xenograft tumor-bearing mice (Table 1).

Fig. 3
figure 3

Molecular docking results for the compounds showing the strongest interaction with HER1/EGFR tyrosine kinase domain (PDB ID = 3W2O) and HER2 tyrosine kinase domain (PDB ID = 3PP0). Residues labeled bold at the tables were stated in the literature to reside at the site where the known EGFR inhibitors bind

Full size image

Inhibitors of HER2/HER3 signaling

Although the inhibition of other EGFR family members was much less investigated, several studies provided results for the inhibition of HER2 and HER3 and their related downstream signaling routes. As can be seen in Table 2, most evidence has been gathered for (−) epigallocatechin-3-gallate, one of the active ingredients of green tea (Camelia sinensis), the flavonoid genistein from soy (Glycine max), and curcumin from Curcuma longa.

Table 2 Phytochemicals with activity against HER2 and HER3
Full size table

Few publications investigated other compounds such as houttuyninum, 11,11′-dideoxy-verticcillin, ZH-4B, resveratrol, pterostilbene, dihydrocalcones, quercetin, and apigenin. The mechanisms of actions how these phytochemicals affect HER2 and HER3 are comparable with those observed for EGFR. They include inhibition of HER2/HER3 phosphorylation and expression as well as inhibition of downstream signal transducers, e.g., ERK1/2, AKT, STAT3, p38MAPK, cRAF, Elk-1, and PI3K (Table 2). Further evaluation of the compounds mentioned in Table 2 in terms of HER2 inhibition was performed with in silico molecular docking analyses on human HER2 tyrosine kinase domain. Molecular docking analyses in silico on human HER2 tyrosine kinase domain revealed curcumin to interact with comparable binding energies as the known inhibitor, lapatinib with similar docking poses (Fig. 3). The results for the in silico molecular docking analyses on EGFR and HER2 tyrosine kinase domains are represented in Table 3 and Table 4, respectively. The inhibition of HER2/HER3 and related signal transduction pathways led to growth inhibition and induction of apoptosis as well as to the inhibition of human tumor xenograft growth in vivo. Comparable to EGFR inhibition, HER2 and HER3 inhibition by curcumin also caused synergistic growth inhibition with 5-fluorouracil/oxaliplatin (Table 2).

Table 3 Molecular docking results of selected inhibitors (white rows) and phytochemicals (light gray rows) against EGFR
Full size table
Table 4 Molecular docking results of selected inhibitors (white rows) and phytochemicals (light brown rows) against HER2
Full size table

Conclusions and perspectives

The identification of tumor target molecules with prognostic relevance for patients opened avenues for the development of more specific treatment options. Important examples in current cancer biology and pharmacology are epidermal growth factor receptors and specific small molecules inhibiting their signaling in tumors. Nevertheless, resistance can also occur towards targeted therapies and novel drugs attacking these receptors are needed. Natural products have been identified as possible novel drug candidates specifically inhibiting EGFR in tumor cells. An important perspective for EGFR/HER2/HER3 inhibiting natural products is their use for personalized treatment options. The individual testing of the mutational status would allow selecting the right EGFR/HER2/HER3 inhibitor for the right patient. In this respect, natural products may represent valuable tools for the development of personalized therapy in the years to come.

The reliable prediction of resistance development is still a major unresolved issue. Deep sequencing and next-generation sequencing have great potentials in monitoring the development of drug resistance in individual tumors and thus offer a new dimension in personalized medicine. In addition to monitoring clinical course of tumor diseases upon drug treatment, whole genome sequencing techniques may be useful to measure the modulation of drug resistance by natural compounds. In addition to studies with large numbers of patient samples taken before and after treatment, longitudinal studies monitoring the same patients at the beginning of and during therapy may provide better insight into the individual mechanisms of resistance development in each individual tumor. This kind of research opens avenues for the prediction of individual response of a tumor patient to therapy. It would be of great value for patients to know whether or not a tumor will respond to the proposed therapy [31]. If a tumor is resistant, therapy will only cause toxic effects in normal tissues without effect on the tumor. Then, another more effective regimen could be applied or natural products alleviating the adverse side effects of chemotherapy could be applied.