The fundamental studies of targeting the transformed cell and reversing phenotype have influenced the practice of medicine. The notion that cancer therapy can be manipulated in such a way that the transformed or cancerous cells would be targeted and normal cells spared is one of the most startling discoveries of modern medicine. The repercussions of this finding were that different cancers could not be treated identically and even cancers originating in the same primary organ could be further divided based on their specific aetiology. This molecularly defined characterization is the basis for therapeutic intervention. In other words, the discoveries not only had an impact on cancer medicine, but also identified the field of personalized therapy. Greene’s laboratory pioneered the idea that monoclonal antibodies (mAbs), which recognize cell surface epitopes of oncogenic receptor tyrosine kinases, could be used to reverse the malignant phenotype of transformed cells. Ultimately, this finding was modified for use in humans and is now prescribed as trastuzumab (Herceptin®, Roche Products Ltd, Hertfordshire, UK) for erbB2/human epidermal growth factor receptor (HER)2-positive breast cancer. The focus of this review shall be to summarize the body of work from Greene’s laboratory that initiated targeted therapy as well as to highlight other key contributions.
History and time line
The evolution of targeted therapy originated with Greene’s studies1 beginning in 1978 on the transforming middle-T antigen of the oncogenic polyomavirus. Those studies clearly showed that the middle-T kinase encoded a cell surface epitope that associated with histocompatibility antigens and could be recognized by the immune system.1
The epidermal growth factor (EGF) – epidermal growth factor receptor (EGFR), erbB1 or HER1 – was contemporaneously identified as a receptor tyrosine kinase.2 Shortly after, a transforming high-molecular- weight DNA was identified from brain tumours of the offspring of BDIX pregnant rats exposed to ethylnitrosourea and named neu by convention as it was identified from neuroblastomas.3,4 Subsequent cloning of neu revealed a high degree of similarity with EGFR; hence, the human gene was named erbB2 or HER2.4–9
Using an antiserum developed in the laboratory, Greene showed that the product of the genomic oncogenic neu gene is a cell-surface protein with an apparent molecular weight of 185 000 Da. Greene’s laboratory raised mAbs that recognized the neu protein product and found these mAbs to be specific for the protein of 185 kDa (p185).10,11 These studies established that the p185 protein was expressed only on the surface of neu-transformed cells and defined this as an oncogene-encoded tumour antigen. Dr Greene’s laboratory then determined that mAb treatment of neu-transformed NIH-3T3 cells sufficiently reversed the transformed phenotype in vitro and promoted survival in tumour studies of athymic nude mice.12
The unexpected effect of a specific monoclonal that binds the ectodomain of p185 was remarkable and both the kinase activity of the receptor and the phenotype of the cells transformed by the neu oncogene were dramatically altered. Kinase activity diminished and the cells assumed a more normal phenotype in anchorage-independent and -dependent assays.11 A unique experiment was undertaken using the original tumour and syngeneic animals. Dramatic inhibition of neu-transformed rat neuroblastoma in a syngeneic BDIX rat host revealed that this approach was therapeutic and potentially relevant to human tumours. The possible relevance to human disease was further revealed as the anti-rat neu antibodies developed were also shown to bind some human tumours (X. Real and M. Greene, Harvard Medical School, 1985).
These novel findings and relevance to human disease were supported by studies from Semba et al.8 and Yamamoto et al.,9 who found that erbB2, the human homologue, was amplified in adenocarcinomas, and King et al.,13 who found a 10% amplification of the erbB gene in breast cancer cell lines. Subsequently, Maguire, Slamon and Nusse’s laboratory determined that the HER2/neu gene was amplified in breast cancer and the amplification correlated with shorter time to relapse and lower survival rates.14,15
Based on these results, Genentech (San Francisco, CA, USA) began to develop antibodies using the immunization protocols our laboratory had pioneered and created similar rodent antibodies to the ones we had made. By 1989, Genentech began humanizing its antibodies to produce a humanized mAb species capable of recognizing human HER2/neu with the idea of inhibiting cell growth in HER2-positive breast cancer cells.16 In fact, one of the original most efficacious hybridomas (7.16.4) binds to the same epitope as the humanized form that Genentech created that ultimately reached the clinic.
Dimeric forms of the oncoprotein are the drivers of transformation, and Dr Greene’s laboratory defined the formation of homodimers of two p185 subunits and heterodimer formations between EGFR and HER2 and how to disable these complexes.17–19 In 1991, the first antiHER2 antibody was tried in humans. The antibody was further improved through refined humanization using a gene conversion strategy to create a mAb that specifically recognized the extracellular domain of HER2.20
Clinical trials of antiHER2/erbB2/neu antibodies began with phase I studies to determine the pharmacokinetic profile; however, Dr Larry Norton’s group performed the definitive phase II studies.21 Clinical observations of side-effects and basic developmental studies revealed that HER2/neu is critical to cardiac and central nervous system development;22 therefore, these data revealed a potential side-effect of treatment.
In 1998, herceptin received Food and Drug Administration (FDA) fast-track designation for the treatment of metastatic breast cancer and was approved for use with paclitaxel for HER2-positive breast cancer as a first-line treatment and as a single agent for second- and third-line therapy. Herceptin is recognized as the first therapeutic antibody targeted at an oncogene to achieve drug approval.
In summary, this time line represents a synopsis of the initial discovery of a mAb therapy from the basic scientific advancement through clinical development and concluding with an overview of clinical trials and subsequent FDA approval. It is remarkable to consider this as a summary of over 20 years of work and certainly emphasizes the importance of the targeted therapy discovery.
Original work was an extension of studies linking tumour antigens to transforming gene products
The origin of targeted therapy can be largely attributed to the concept of exploiting allelic and adaptive changes in transformed cells that manifest themselves when a cancerous cell expressing antigens interacts with, or is recognized by, the host immune system. Studies carried out after the polyoma experiments that linked the middle-T kinase with immune recognition were carried out in collaboration with the Weinberg laboratory at Massachusetts Institute of Technology, MA, USA. The initial collaboration further established that the activated immune system, in particular T-cells, could inhibit growth of tumour cells that were transformed by what turned out to be Ki-RAS.23 Thus, efforts to develop targeted therapy were stimulated from the aforementioned polyoma studies of middle-T antigens and the Ki-RAS efforts.1,24
Although others attempted to capitalize on these discoveries23 with crude antisera, Greene’s laboratory focused on mAbs that could specifically recognize oncogene-transformed cells. Transfection of fibroblasts with high-molecular-weight DNA from rat neuroblastomas was used to select cells that reflected the neoplastic transformation, and Greene utilized these cells to generate mAbs. The potential of the concept was realized when the mAbs were shown to specifically recognize a 185-kDa cell surface protein that was exclusive to the transformed cells. Not only did the mAbs identify the tumour antigen, but they ultimately could reverse the malignant phenotype.10
Early work and reversal of the phenotype
The erbB or HER family of receptor tyrosine kinases (RTKs) comprises erbB1 (EGFR/HER1), erbB2 (p185/neu/HER2), erbB3 (HER3) and erbB4 (HER4), and when overexpressed or mutated, members of this kinase family form homomeric or heteromeric kinase assemblies that drive human malignancies. ErbB2/HER2/neu amplification is found in approximately 25% of breast cancer patients and is associated with poor prognosis and decreased survival.25 Targeted therapy of this class of breast cancers evolved from studies showing that mAbs to the ectodomain of erbB2/neu would reverse the malignant phenotype.
Work performed by Dr Jeffrey Drebin, while a student in Dr Greene’s laboratory, revealed that disabling the kinase complex with mAbs specific for the ectodomain could reverse aspects of the malignant phenotype. This reversion of the malignant phenotype was true in vitro using neu-transformed fibroblasts as well as when the neuroblastoma cells from which neu was originally isolated were subjected to similar therapies. Studies in vivo in athymic BALB/c nude mice confirmed the findings as well as revealed an antibody-dependent, cell-mediated cytotoxicity (ADCC) that relied on natural killer cells and macrophages; however, this ADCC was not necessary to limit tumour growth. Importantly, the antibody cross-linked the receptor and resulted in a degradation of receptor protein.11,12 Another key aspect of these studies is that, when the antibody was removed, the malignant features returned. This was the first example of disabling a protein complex to reverse a transformed phenotype. A critical study that distinguished targeting efforts from any others was the use of a totally syngeneic in vivo system.12 BDIX rats were implanted with the original neu-caused neuroblastoma (it had been derived from BDIX animals and then treated with the monoclonal that recognized rat p185). As these rats expressed the normal proto-oncogenic p185 and were not harmed by the administered antibody, this suggested that human therapy was indeed possible. Greene’s laboratory identified a potential therapeutic targeted application of antierbB2 monoclonals that could extend to humans.
Homodimers and heterodimers
The erbB RTKs are single-pass membrane glycoproteins whose intracellular C-terminal domain possesses tyrosine kinase activity and whose N-terminal extracellular domain contains the ligand-binding and dimerization domains. Known ligands for EGFR include EGF, transforming growth factor-alpha, heparin-binding EGF (HB-EGF), amphiregulin, epiregulin and betacellulin. No HER2 ligand has been reported. HER3 and HER4 are activated by the four neuregulins (NRG1–4) and, in addition, HER4 is also activated by HB-EGF, betacellulin and epiregulin (more in-depth information can be found in Fuller et al.26).
Structural and biochemical studies have revealed that erbB1, erbB3 and erbB4 receptors, in the absence of a ligand, are usually maintained in an autoinhibitory state in which the dimerization arm within domain II is obscured by domain IV. Following ligand binding to domain I, the receptor undergoes a conformational change whereby domains I and II rotate, exposing the dimerization surface and allowing homo- or heterodimerization and subsequent receptor activation.27
Notably, although the erbB2 receptor does not have an identified ligand, its native structure facilitates a constitutively active species. ErbB2 appears to maintain an extended conformation and serves as the preferred heterodimerization partner for the other erbB receptors.28,29 Heterodimers are more kinase active and Greene’s laboratory identified that erbB2/HER2/neu-EGFR heteromers readily promote metastatic dissemination.17 Others have found that the combination erbB2-erbB3 also leads to potent transformation potential.30
Sequence analysis revealed that the proto-oncogenic form of neu differed from the oncogenic sequence by a single amino acid alteration of a valine to glutamic acid that introduces a negative charge into the transmembrane portion of the receptor.31 Dr Greene’s group resolved how the introduction of a single negative charge caused the protein to become transforming by showing that the oncogenic form of the protein exists predominantly as a homodimer while the proto-oncogenic isoform is primarily a monomer.19 Furthermore, tyrosine kinase activity was limited to the oncogenic, homodimeric isoform, establishing that dimerization itself is required for tyrosine kinase activity.
Overexpression can also lead to dimers. Greene’s laboratory found that transfecting either the non-oncogenic neu or EGFR singly to cells, to modest levels of overexpression, did not completely transform fibroblasts to the point of forming tumours in nude mice.17 However co-transfection of even minimally overexpressed forms of normal neu and normal EGFR created neu/EGFR dimers and resulted in transforming capability.18 Further, erbB3 and erbB4 also form heterodimers and neu is their favoured binding partner.28,29
Inhibiting dimers and monomers
The antitumour property of the erbB2/HER2/neu-specific mAbs is derived from the ability to disable homo- and heterodimeric receptor complexes. In addition to inhibiting tumour growth, use of the antineu mAb could also prevent the emergence of tumour development in transgenic mice.32 These studies revealed that cells expressing oncogenic neu must undergo further neoplastic transformation in order to become completely transformed. The use of antiHER2 mAb prior to full transformation would be of therapeutic benefit; therefore, HER2 antibodies were examined as an adjuvant to prevent tumour re-emergence in women following surgical removal of HER2-positive breast tumours.33 Moreover, inclusion of a second antibody that is reactive with a distinct epitope of HER2 enhanced the reversal of phenotype in vitro and completely eradicated some tumours in vivo in animal models.34
Humanized forms of antibodies that react with distinct domains of the ectodomain of p185 have been approved as dual agents for breast cancer therapy. Trastuzumab and pertuzumab (Perjeta, Roche Products Ltd, Hertfordshire, UK) as dual antibody therapies are now applied to human diseases such as erbB2-driven breast and stomach cancer,35–38 reflecting the diverse oncogenic capacity of HER2 and supporting targeted therapy as a landmark discovery in oncology research.
Reversal of phenotype and radiation sensitivity
Although the development and implementation of targeted therapy aimed at disabling erbB signalling with mAbs and small molecule tyrosine kinase inhibitors has produced significant advances, cancer cells frequently develop resistance to these treatments, even in combination with chemotherapy or radiation therapy. ErbB2/neu- and EGFR-transformed cells are inherently resistant to radiation-induced apoptosis.
Radiation treatment predominantly induces double-strand DNA breaks, which, if not repaired, are lethal to the cell. Some tumours can resist radiation treatment owing to their enhanced ability to repair double-strand breaks. Greene’s laboratory, as well as others, showed that, even in the presence of ionizing radiation, active erbB kinase signalling appears to enhance the repair process, such that transformed cells resist genotoxic signal-induced cell death.
The erbB RTK signalling pathways are critical to DNA repair as activation with either a ligand or a constitutively active mutant form enhances DSB repair kinetics,39,40 and EGFR-enhanced DSB repair is mediated primarily through the error-prone non-homologous end-joining pathway.41,42
Dr Greene’s laboratory was pioneering in this field, as were studies performed by Dr Donald O’Rourke, who was a fellow in the group. Specifically, disabling erbB2 stops downstream signalling43,44 and causes a more radiosensitive phenotype.45 These landmark findings are supported by work from several other laboratories using mAbs, peptides, small molecule inhibitors and siRNA to examine the DNA damage response role of HER2 as well as EGFR in vitro and in vivo.46–50 Clearly, the erbB family of receptors is critical in mediating the cellular response to ionizing radiation. Finally, these findings are in the process of being translated to the clinic. A study that combined trastuzumab with radiation therapy as a treatment for breast cancer was evaluated in a phase II trial and the treatment regimen was well tolerated.51
Over the last 30 or more years, Greene’s laboratory group has pioneered the field of targeted therapy in oncology research. These studies have altered the practice of cancer medicine and have translated to clinical benefit. Targeted therapy is one of the most promising avenues of therapeutic intervention.