Understanding the biology of HER3 receptor as a therapeutic target in human cancer

Abstract

HER3 belongs to the human epidermal growth factor receptor (HER) family which also includes HER1/EGFR/erbB1, HER2/erbB2, and HER4/erbB4. As a unique member of the HER family, HER3 lacks or has little intrinsic tyrosine kinase activity. It frequently co-expresses and forms heterodimers with other receptor tyrosine kinases (RTKs) in cancer cells to activate oncogenic signaling, especially the PI-3K/Akt pathway and Src kinase. Elevated expression of HER3 has been observed in a wide variety of human cancers and associates with a worse survival in cancer patients with solid tumors. Studies on the underlying mechanism implicate HER3 expression as a major cause of treatment failure in cancer therapy. Activation of HER3 signaling has also been shown to promote cancer metastasis. These data strongly support the notion that therapeutic inactivation of HER3 and/or its downstream signaling is required to overcome treatment resistance and improve the outcomes of cancer patients.

Graphical abstract

HER3 belongs to the human epidermal growth factor receptor (HER) family. HER3 lacks or has little intrinsic tyrosine kinase activity, but it frequently co-expresses and forms heterodimers with other receptor tyrosine kinases (RTKs) to activate oncogenic signaling in cancer cells. Therapeutic inactivation of HER3 and/or its downstream signaling is required to overcome treatment resistance and improve the outcomes of cancer patients.

1. Introduction

Human epidermal growth factor receptor (HER) family includes the epidermal growth factor receptor (EGFR), HER2 (also known as erbB2/neu), HER3 (erbB3), and HER4 (erbB4). It is arguably the most important family of receptor tyrosine kinase (RTK) in normal development and tumorigenesis1, 2. These receptors are widely expressed in epithelial, mesenchymal, and neuronal cells³. Abnormal expression of HER family members is involved in carcinogenesis and progression of diverse types of human cancer4, 5. While EGFR, HER3, and HER4 have ligands, HER2 has no known ligand. When a ligand binds to the extracellular region of EGFR, HER3, or HER4 (domains I and III), the dimerization arm in domain II is exposed leading to receptor-receptor interaction⁶. Dimerization is an essential step for the receptor function and activation of the cytoplasmic signaling, including PI-3K/Akt, MEK/MAPK, Jak/Stat pathways, Src kinase, etc.5, 7. EGFR, HER3, and HER4 normally exist as inactive molecularly folded monomers to prevent dimerization8, 9, whereas HER2 is always in a constitutively active conformation with its dimerization arm opening even without ligand binding⁸. Accumulating evidence indicates that HER3 frequently co-expresses and interacts with other RTKs to form a heterodimeric complex, which subsequently activates oncogenic signaling, especially the PI-3K/Akt pathway and Src kinase to promote cancer cell survival, proliferation, and progression10, 11, 12. Studies on the underlying mechanisms demonstrate that HER3 signaling plays a major role causing treatment failure in cancer therapy13, 14. Recent reports reveal that enhanced HER3 signaling facilitates tumor cell motility and intravasation in breast cancer lung metastasis¹⁵; and a HER3-lncRNA (long noncoding RNA) axis regulates bone metastasis in breast cancer16, 17. Increased expression and activation of HER3 has also been observed in brain metastasis of breast cancer resistance to PI-3K inhibition18, 19. Collectively, these data support the importance of developing effective therapeutics to inhibit HER3 signaling for cancer treatment. A number of anti-HER3 monoclonal antibodies are actively under preclinical studies and clinical evaluation in cancer patients. There is currently no HER3-targeted therapy approved by the FDA for cancer treatment. This review summarizes our understanding of the unique biology of HER3 in cancer progression and discusses the latest advances in identifying therapeutic antibodies against HER3 for cancer treatment.

2. Unique biology of HER3-initiated signaling in human cancer

HER3 is a unique member of the HER family as it has been considered as an inactive receptor20, 21, although a recent study suggests that HER3 contains weak kinase activity²². Sequence comparison of tyrosine kinase domains among the HER receptors reveals that certain residues, including Cys-721, His-740, and Asn-815, in HER3 have non-conservative substitutions. These changes significantly reduce the kinase activity of HER3²³. Thus, to fully transduce signaling, HER3 has to form dimers with other receptors and be phosphorylated by its interactive partners, with HER2 being the most important one²⁴. Of the four HER receptors, HER3 is best suited to induce activation of the PI-3K/Akt pathway, which is a well-known survival signaling pathway in normal development and tumorigenesis²⁵. This is likely due to the C-terminal tail of HER3 having multiple tyrosine residues, whose phosphorylation is able to bind to the p85 subunit of PI-3K²⁴. It is thought that, among all the homo- and hetero-dimerization complexes potentially formed by HER receptors, the HER2/HER3 heterodimer is the most biologically active and potent to activate the PI-3K/Akt signaling cascade26, 27.

Overexpression of HER3 is frequently observed in a wide variety of human malignancies, including colorectal carcinoma, head and neck squamous cell carcinoma, melanoma, and breast, gastric, ovarian, prostate, and bladder cancers28, 29, 30. Moreover, it has been shown that HER3 is a more potent partner than other HER receptors for the oncogenic activity of HER2 in HER2-overexpressing tumors29, 31, 32, 33. Especially in erbB2-amplified breast cancers, preferential phosphorylation of HER3, but not EGFR, is found²⁹. Indeed, most metastatic breast cancers have expression of either EGFR or HER2, and rarely express both³⁴. In contrast, HER2 and HER3 commonly co-express in breast cancer tissues³⁵ and breast cancer cell lines³⁶. Elevated expression of the endogenous mouse HER3 and its association with the transgene encoded erbB2 promote mammary tumorigenesis in erbB2/neu-transgenic mice37, 38. Despite its lack of20, 21 or weak kinase activity²², HER3 serves as a critical co-receptor of HER2 and its expression is essential for HER2-mediated breast cancer cell survival and proliferation10, 11. These data have been supported by a recent meta-analysis of 12 clinical studies of human cancers, including colorectal cancer, gastric cancer, breast cancer, melanoma, ovarian cancer, head and neck cancer, pancreatic cancer, and cervical cancer³⁹. It concludes that expression of HER3 is associated with worse survival in solid tumors, and the impact of HER3 on clinical outcome is greater in those tumors where HER2 is also overexpressed³⁹.

Overexpression of HER3 has been reported in 50%–70% of human breast cancers40, 41, 42 and appears to be associated with prognostic factors, such as distant metastasis, tumor size, risk of local recurrence, and etc.43, 44, although the prognostic value of HER3 in breast cancer is not well documented and the currently available data are inconsistent42, 43, 44, 45, 46. Some studies show that elevated expression of HER3 significantly correlates with reduced overall survival and disease-free survival35, 47, 48, whereas others report HER3 expression as a favorable prognostic factor of overall survival in breast cancer patients49, 50. Several theories have been proposed to explain the controversial findings, such as the potential influence of HER3 ligand—heregulin (HRG) and subcellular distribution of HER3¹⁴. The fact that we do not have a unified methodology to detect HER3 expression in clinical samples may also account for the inconsistent data, as each laboratory uses different antibodies and probes to detect the expression of HER3 protein and mRNA. In addition, breast cancer is a heterogenous disease with several intrinsic subtypes, including luminal, HER2-enriched, and triple negative breast cancer (TNBC)⁵¹. It is possible that HER3 exhibits distinct influences on patient survival in different subtypes of breast cancer. Thus, detailed evaluation of HER3 expression and its interactive partners in a specific subtype is warranted to define the prognostic value of HER3 signaling in such subtype of breast cancer patients.

3. Mechanism of HER3-mediated cancer progression

Accumulating evidence emphasizes the critical role of HER2/HER3 heterodimer-mediated PI-3K/Akt signaling in cancer development39, 52. Basic research on the underlying mechanisms indicates that HER2 contributes to breast carcinogenesis potentially via two major mechanisms—increased therapeutic resistance and enhanced metastatic potential53, 54. Thus, it is conceivable to hypothesize that HER3 signaling-mediated cancer progression is likely through its capability to induce therapeutic resistance and promote tumor metastasis.

3.1. HER3 and cancer treatment resistance

A recent report implicates HER3 activation as a major cause of treatment failure in cancer therapy¹³. It has been shown that HER3 signaling plays a crucial role in the development of various human cancers, including HER2-overexpressing breast cancer10, 11, castration-resistant prostate cancer⁵⁵, platinum-resistant/refractory ovarian cancer56, 57, and non-small cell lung cancer (NSCLC) resistance to EGFR tyrosine kinase inhibitor (TKI)58, 59. A number of studies reveal that compensatory upregulation of HER3 along with the sustained PI-3K/Akt signaling is implicated as an important mechanism resulting in resistance to EGFR-targeted therapy60, 61, 62, 63. In addition, elevated expression of the HER3 ligand (HRG) is a possible mechanism of resistance to anti-EGFR antibody (Ab)-cetuximab in the treatment of patients with colorectal cancer⁶⁴. Furthermore, HER3 may work in concert with other RTKs, such as hepatocyte growth factor receptor (HGFR or MET)⁶⁵. Amplification of MET oncogene may also result in resistance to EGFR-TKI (gefitinib). Phosphorylated HER3 was able to interact with the p85 subunit of PI-3K in a MET kinase-dependent manner in NSCLC, suggesting a role of HER3 in MET-induced resistance to gefitinib⁶⁵. In squamous cell carcinomas of head and neck cancer cell lines sensitive to the dual EGFR/HER2 inhibitor lapatinib, increased HRG and activated HER3 strongly correlated with lapatinib sensitivity⁶⁶. However, the potential mechanism by which HER3 may be a valuable biomarker for lapatinib sensitivity and gefitinib resistance remains unclear. It may be through distinct activation mechanisms that need to be further investigated.

Studies in our laboratory have been focusing on the biologic function of HER3 in the progression of erbB2-aberrant breast cancer. We show that elevated expression of HER3 in HER2-overexpressing breast cancer cells results in resistance to hormone therapy (tamoxifen), HER2-targeted therapy (trastuzumab and lapatinib), and chemotherapy (paclitaxel)67, 68, 69, 70, 71. Our data demonstrate the crucial role of HER3 signaling in HER2-mediated therapeutic resistance to tamoxifen, trastuzumab, and paclitaxel in breast cancer12, 14. One innovative finding comes from our studies on the underlying mechanism of HER3-mediated resistance to the anti-HER2 antibody trastuzumab (also known as Herceptin). It was reported that both HER3 and the insulin-like growth factor-I receptor (IGF-1R)-mediated signaling contributed to trastuzumab resistance72, 73, 74, whereas the relationship between HER3 and IGF-IR in trastuzumab resistance was less understood. Our studies uncovered that HER2 interacted with both HER3 and IGF-1R to form a heterotrimeric complex in the trastuzumab-resistant breast cancer cells we tested. In fact, it was the heterotrimer of HER2/HER3/IGF-1R, not the heterodimer of HER2/HER3 or IGF-1R/HER2, that played a causal role leading to trastuzumab resistance⁶⁷. Further studies on downstream signaling revealed that HER3 and IGF-1R triggered different signaling pathways contributing to trastuzumab resistance - HER3 activated both PI-3K/Akt signaling and Src kinase, whereas IGF-1R mainly elicited Src activation⁶⁷. Interestingly, our recent data show that HER3 and IGF-1R exhibit distinct effects on the sensitivity of HER2-overexpressing breast cancer cells to lapatinib, another HER2-targeted therapy⁷¹. While HER3 signaling also induces lapatinib resistance in the trastuzumab-resistant breast cancer cells, IGF-1R signaling did not alter lapatinib sensitivity⁷¹.

3.2. HER3 and tumor metastasis

HER3 frequently co-expresses and interacts with HER2 to activate oncogenic signaling, especially the PI-3K/Akt pathway and Src kinase, and promote cancer cell survival, proliferation, and progression10, 11, 12. We have shown that elevated expression of HER3 confers resistance to several commonly used therapeutics against HER2-overexpressing breast cancer67, 68, 69, 70, 71. Drug-resistant tumors likely recur and metastasize to distant organs. Thus, it is generally believed that overexpression of HER3 and its downstream signaling can promote tumor metastasis. Activation of HER3 signaling facilitates tumor cell motility and intravasation in lung metastasis of human breast cancer¹⁵. Our analysis of clinical database reveals that increased HER3 expression leads to a worse overall survival (OS) in lymph node positive breast cancer patients. Especially in HER2-overexpressing breast cancer, the patients with higher expression of HER3 show poorer OS and distant metastasis-free survival (Liu laboratory unpublished data). In addition, the HER3 ligand, HRG can stimulate chemotaxis and invasion via HER2/HER3 heterodimers⁷⁵. Recent studies suggest that the HRG-HER3 signaling axis plays a crucial role in the brain metastasis of breast cancer18, 19. While overexpression of HER3 is found in the brain metastatic legions of breast cancer19, 76, activation of HER3 and its downstream signaling has also been observed in breast cancer brain metastasis likely via increased HRG production by the stromal cells in brain microenvironment18, 19, 77. Activation of the downstream signaling, such as the PI-3K/Akt and MEK/MAPK pathways can be critical for cell motility and chemotaxis75, 78, 79, 80, 81, 82. PI-3K is capable of regulating cytoskeleton through Rho family G proteins and Akt activation83, 84, 85. MAPKs can influence adhesion dynamics directly and control gene expression patterns essential for motility and invasion86, 87, 88. It is possible that HER3-dependent motility contributes to cancer metastasis independent of its effects on tumor growth⁸⁹. Studies on the underlying mechanisms involved in ovarian cancer spread to the omentum shows that elevated expression of HER3 in ovarian cancer cells and increased HRG in the omentum allows for cancer cell localization and growth in the omentum. These findings suggest that the HRG-HER3 signaling axis is also a dominant mechanism responsible for ovarian cancer metastasis via blood stream⁹⁰.

Interestingly, noncoding RNA (ncRNA), including the long ncRNA (lncRNA) MAYA also plays an important role in HER3-mediated tumor metastasis¹⁷. It has been reported that a ROR1-HER3-lncRNA axis regulates bone metastasis in breast cancer16, 17. In our efforts to identify key downstream mediators of HER3 signaling in breast cancer metastasis, we found that HER3 signaling specifically downregulates expression of the tumor suppressive miR-203 and miR-542-3p in HER2-overexpressing breast cancer cells⁹¹. Bioinformatics analyses reveal that miR-203 and miR-542-3p target several genes, including Survivin, ZEB1, ZEB2, Snail1, and/or Slug, which are critical for drug resistance, epithelial-mesenchymal transition (EMT), and tumor metastasis (Liu's laboratory unpublished data). These data support the notion that HER3 signaling regulates expression of lncRNAs and miRNAs to promote cancer metastasis. Studies in this innovative area will not only further our understanding of HER3 signaling in cancer biology, but may also provide a new avenue for identification of novel therapeutic approaches to abrogate HER3-mediated treatment resistance and tumor metastasis. Figure 1 shows a simple diagram depicting that activation of HER3 and its major downstream signaling induces expression of a cohort of critical molecules, including some EMT markers and ncRNAs, responsible drug resistance and cancer metastasis.

Figure 1.

A diagram showing the major signaling pathways of HER3 during cancer progression and the mono-specific HER3 blocking Abs currently in clinical trilas of cancer patients. The ligand, HRG bound HER3 recruits another RTK to form a heterodimer, which subsequently triggers activation of multiple signaling pathways, including PI-3K/Akt, MEK/MAPK, Jak/Stat pathways, and Src kinase. The downstream signaling will further induces expression of a cohort of crucial genes responsible for drug resistance and cancer metastasis. HER3 signaling is also able to regulate expression of some ncRNAs, including miRNAs and lncRNAs. Currently, there are several human or humanized anti-HER3 mono- and bi-specific Abs in clinical trilas testing their therapeutic activity to abrogate drug resistance and inhibit cancer metastasis.

4. Therapeutic antibody against HER3 for cancer treatment

Elevated expression of HER3 plays an essential role in human cancer progression and correlates with a worse overall survival in many solid tumors13, 25, 39, emphasizing the importance in developing novel effective strategic targeting of HER314, 52, 92. Inhibition of HER3 is believed to be required to overcome resistance and effectively treat cancer patients. Because of its lack of or low kinase activity21, 22, targeting HER3 with a blocking Ab is the only strategy under preclinical studies93, 94 and clinical evaluations in patients with advanced solid tumors (http://www.clinicaltrials.gov). Advances have been made to identify HER3-targeted therapy⁹⁵, and several anti-HER3 monoclonal Abs exhibit antitumor activity in vivo and show promise as novel cancer therapeutics96, 97. Recent studies have identified bispecific Abs dual-targeting EGFR/HER3⁵⁹ or HER2/HER3⁹⁸, that exert potent antitumor activities in both laboratory studies and clinic testing⁹⁵. The HER3 inhibitors based on a novel biologic scaffold termed surrobody have been developed and display anti-proliferative effects on cancer cells in vitro and in vivo⁹⁹.

MM-121 (also known as seribantumab, Merrimack Pharmaceuticals, Cambridge, MA), a human anti-HER3 monoclonal IgG2 Ab, blocks ligand-induced HER2/HER3 dimerization and subsequently inhibits downstream signaling. MM-121 exerts antitumor activity in preclinical studies of various human cancers93, 94. We have tested the hypothesis that MM-121 may be able to abrogate HER3 signaling-mediated resistance to trastuzumab and paclitaxel in HER2-overexpressing breast cancer cells via inactivation of HER3 and its downstream PI-3K/Akt signaling. We reported that MM-121 was able to overcome paclitaxel resistance and significantly enhanced paclitaxel-induced apoptosis in the otherwise resistant breast cancer cell lines¹⁰⁰. We also showed that MM-121 dramatically inhibited PI-3K/Akt signaling in HER2-overexpressing breast cancer cells refractory to trastuzumab, and significantly enhanced trastuzumab-induced growth inhibition¹⁰¹. MM-121 in combination with trastuzumab mainly induced cell cycle G1 arrest in vitro, whereas the combinations of MM-121 and trastuzumab potently inhibited tumor growth in vivo likely due to induction of both growth inhibition and apoptosis¹⁰¹. Our data strongly support the initiation of clinical trials to evaluate the efficacy of MM-121 in combination with trastuzumab or paclitaxel in HER2-overexpressing breast cancer patients who have developed resistance to the therapeutics. Interestingly, recent studies suggest that higher HRG mRNA expression and low HER2 levels predict a clinical benefit from the addition of seribantumab (MM-121) to standard of care therapies in patients with platinum-resistant/refractory ovarian cancer, hormone receptor-positive HER2-negative breast cancer, and EGFR wild-type NSCLC102, 103. MM-111 (Merrimack Pharmaceuticals, Cambridge, MA) is a bispecific antibody, dual-targeting HER2/HER3, inhibiting the PI-3K/Akt signaling⁹⁸. The safety and clinical activity of MM-111 is now being tested in several phase I clinical trials. Another HER3-targeted drug, U3–1287/AMG-888 (originally developed by Amgen Inc., Thousand Oaks, CA; later acquired by Daiichi Sankyo Co., Ltd., Tokyo, Japan and re-named as patritumab) is the first fully human anti-HER3 monoclonal Ab and currently under phase III clinical investigations in patients with advanced solid tumors¹⁰⁴. This Ab has been shown to inhibit proximal and distal HER signaling and induces rapid internalization of HER3¹⁰⁵. Patritumab induces growth inhibition in various cancer cell lines (breast, lung, colorectal) that are resistant to other HER inhibitors¹⁰⁵. It significantly decreases colony formation in pancreatic cancer cells and tumor growth in tumor xenograft models of pancreatic cancer, NSCLC, and colorectal cancer⁵⁵. Interestingly, patritumab is also able to overcome HRG-dependent resistance to EGFR inhibitors in NSCLC in vitro and in vivo, suggesting that patritumab may be useful in combination with EGFR TKIs, such as erlotinib to treat the NSCLC patients with high expression of HRG106, 107. Lumretuzumab (RG7116) is a humanized anti-HER3 IgG1 monoclonal Ab developed by Roche Diagnostics GmbH (Penzberg, Germany). It binds to the extracellular domain of HER3 with high affinity to prevent HRG binding¹⁰⁸. As a glycoengineered monoclonal Ab, lumretuzumab displayed an enhanced antibody-dependent cell-mediated cytotoxicity (ADCC) activity as compared with the non-glycoengineered parental antibody¹⁰⁹. Although lumretuzumab was well tolerated and showed evidence of clinical activity in a phase I trial¹¹⁰, two recent phase Ib studies suggest otherwise. The toxicity profile of lumretuzumab in combination with the EGFR-targeted therapies cetuximab and erlotinib was managable, but it exerted little clinical benefit in various cancers¹¹¹. The therapeutic window of lumretuzumab in combination with the anti-HER2 Ab pertuzumab and chemotherapeutic drug paclitaxel for HER3-positive metastatic breast cancer was too narrow to warrant further clinical development¹¹². Several anti-HER3 mono-specific Abs currently under clinical trials with a hope to abrogate HER3-mediated drug resistance and cancer metastasis are shown in Figure 1. Recently, a new anti-HER3 Ab (MP-RM-1) and its humanized version (EV20) exhibit potent antitumor effects in several cancer types in vitro and in vivo113, 114. Because of EV20׳s capability to inhibit both ligand-dependent and -independent activation of HER3113, 114, it is speculated that EV20 may have a broader effect on blocking HER3 signaling than the Abs (like MM-121) which can only block ligand-induced HER3 activation.

5. Perspectives

Research on HER receptors has been focusing on the dysregulation of EGFR and HER2 in human malignancies. The importance of HER3 as an obligate partner for receptor dimerization and in resistance to HER2- or EGFR-targeted therapy and other therapeutics has drawn a lot of attention to define HER3 as a molecular target for cancer treatment. Increased awareness of HER3 function in drug resistance and tumor metastasis has critical implications in the directions of future studies. First, the crucial downstream mediators of HER3 signaling in cancer progression remain elusive. Basic research deciphering the molecular basis of HER3-mediated drug resistance and tumor metastasis is essential to improve our understanding of the unique biology of HER3 in human cancer. Such studies will also facilitate the development of novel therapeutic approaches inhibiting the key downstream mediators against those cancers driven by HER3 signaling. Second, although several anti-HER3 Abs are actively under clinical evaluations in various human cancers, to date no HER3-targeted therapy has been approved by the FDA for cancer treatment. This is possibly due to the uniqueness of HER3 receptor, which may influence the antitumor activity of anti-HER3 Abs. Since HER3 has to form heterodimer or heterotrimer complexes with other RTKs in order to fully transduce signaling10, 11, 67, anti-HER3 monotherapy is unlikely to show significant efficacy against human cancer. We must consider effective combination strategies with a HER3-targeted therapy plus other targeted therapies or chemotherapeutic agents for cancer treatment. Third, it has been shown that HRG expression at tumor sites predicts efficacy of seribantumab (MM-121) in the treatment of human cancers102, 103, suggesting that identification of predictive biomarkers will stratify the usage of anti-HER3 Abs for effective cancer treatment. Indeed, a new anti-HER3 Ab, 9F7-F11, which does not compete with the ligand (HRG), shows higher efficacy than the Abs that compete with the ligand for binding to HER3¹¹⁵. In human tumor cell xenograft models, 9F7-F11 exerts an enhanced antitumor activity in the presence of HRG and thus represents a novel treatment strategy for HRG-addicted tumors¹¹⁵. We believe that HER3 is a focal point in HER receptors-mediated tumorigenesis and plays an essential role in cancer progression. Thus, HER3 constitutes a unique biomarker and molecular target for effective treatment of human cancer.