An integer linear programming approach for finding deregulated subgraphs in regulatory networks.

Deregulation of cell signaling pathways plays a crucial role in the development of tumors. The identification of such pathways requires effective analysis tools that facilitate the interpretation of expression differences. Here, we present a novel and highly efficient method for identifying deregulated subnetworks in a regulatory network. Given a score for each node that measures the degree of deregulation of the corresponding gene or protein, the algorithm computes the heaviest connected subnetwork of a specified size reachable from a designated root node. This root node can be interpreted as a molecular key player responsible for the observed deregulation. To demonstrate the potential of our approach, we analyzed three gene expression data sets. In one scenario, we compared expression profiles of non-malignant primary mammary epithelial cells derived from BRCA1 mutation carriers and of epithelial cells without BRCA1 mutation. Our results suggest that oxidative stress plays an important role in epithelial cells of BRCA1 mutation carriers and that the activation of stress proteins may result in avoidance of apoptosis leading to an increased overall survival of cells with genetic alterations. In summary, our approach opens new avenues for the elucidation of pathogenic mechanisms and for the detection of molecular key players.

Proteomic and transcriptomic characterisation of FIA10, a novel murine leukemic cell line that metastasizes into the brain.

Brain metastasis leads to increased mortality and is a major site of relapse for several cancers, yet the molecular mechanisms of brain metastasis are not well understood. In this study, we established and characterized a new leukemic cell line, FIA10, that metastasizes into the central nervous system (CNS) following injection into the tail vein of syngeneic mice. Mice injected with FIA10 cells developed neurological symptoms such as loss of balance, tremor, ataxic gait and seizures, leading to death within 3 months. Histopathology coupled with PCR analysis clearly showed infiltration of leukemic FIA10 cells into the brain parenchyma of diseased mice, with little involvement of bone marrow, peripheral blood and other organs. To define pathways that contribute to CNS metastasis, global transcriptome and proteome analysis was performed on FIA10 cells and compared with that of the parental stem cell line FDCP-Mix and the related FIA18 cells, which give rise to myeloid leukemia without CNS involvement. 188 expressed genes (RNA level) and 189 proteins were upregulated (log2 ratio FIA10/FIA18 ≥ 1) and 120 mRNAs and 177 proteins were downregulated (log2 ratio FIA10/FIA18 ≤ 1) in FIA10 cells compared with FIA18 cells. Major upregulated pathways in FIA10 cells revealed by biofunctional analyses involved immune response components, adhesion molecules and enzymes implicated in extracellular matrix remodeling, opening and crossing the blood-brain barrier (BBB), molecules supporting migration within the brain parenchyma, alterations in metabolism necessary for growth within the brain microenvironment, and regulators for these functions. Downregulated RNA and protein included several tumor suppressors and DNA repair enzymes. In line with the function of FIA10 cells to specifically infiltrate the brain, FIA10 cells have acquired a phenotype that permits crossing the BBB and adapting to the brain microenvironment thereby escaping immune surveillance. These data and our model system FIA10 will be valuable resources to study the occurrence of brain metastases and may help in the development of potential therapies against brain invasion.

Notch1, Notch2, and Epstein-Barr virus-encoded nuclear antigen 2 signaling differentially affects proliferation and survival of Epstein-Barr virus-infected B cells.

The canonical mode of transcriptional activation by both the Epstein-Barr viral protein, Epstein-Barr virus-encoded nuclear antigen 2 (EBNA2), and an activated Notch receptor (Notch-IC) requires their recruitment to RBPJ, suggesting that EBNA2 uses the Notch pathway to achieve B-cell immortalization. To gain further insight into the biologic equivalence between Notch-IC and EBNA2, we performed a genome-wide expression analysis, revealing that Notch-IC and EBNA2 exhibit profound differences in the regulation of target genes. Whereas Notch-IC is more potent in regulating genes associated with differentiation and development, EBNA2 is more potent in inducing viral and cellular genes involved in proliferation, survival, and chemotaxis. Because both EBNA2 and Notch-IC induced the expression of cell cycle-associated genes, we analyzed whether Notch1-IC or Notch2-IC can replace EBNA2 in B-cell immortalization. Although Notch-IC could drive quiescent B cells into the cell cycle, B-cell immortalization was not maintained, partially due to an increased apoptosis rate in Notch-IC-expressing cells. Expression analysis revealed that both EBNA2 and Notch-IC induced the expression of proapoptotic genes, but only in EBNA2-expressing cells were antiapoptotic genes strongly up-regulated. These findings suggest that Notch signaling in B cells and B-cell lymphomas is only compatible with proliferation if pathways leading to antiapototic signals are active.

The Roche Cancer Genome Database (RCGDB).

Sequence variations are being studied for a better understanding of the mechanism and development of cancer as a mutation-driven disease. The systematic sequencing of genes in tumors and technological advances in high-throughput techniques combined with efficient data acquisition methods have resulted in an explosion of available cancer genome-related data. Despite the technological progress and increase of data, improvements in the application area, for example, drug target discovery, have failed to keep pace with increased research and development spending. One reason for this discrepancy is the ever increasing number of databases and the absence of a unified access to the mutation data. Currently, researchers typically have to browse several, often highly specialized databases to obtain the required information. A more complete understanding of relations and dependencies between mutations and cancer, however, requires the availability of an efficient integrative cancer genome information system. To facilitate this, we developed the Roche Cancer Genome Database (RCGDB), a freely available biological information system integrating different kinds of mutation data. The database is the first comprehensive integration of disparate cancer genome data like single nucleotide variants, single nucleotide polymorphisms, and chromosomal aberrations (CGH and FISH). RCGDB is freely accessible via a Google-like Web interface at http://rcgdb.bioinf.uni-sb.de/MutomeWeb/.

Short-hairpin-RNA-mediated silencing of fucosyltransferase 8 in Chinese-hamster ovary cells for the production of antibodies with enhanced antibody immune effector function.

Antibody-producing Chinese-hamster ovary cells (CHO-DG44) were converted into cells producing antibodies with strongly enhanced ADCC (antibody-dependent cellular cytotoxicity) by knocking down FuT8 (alpha-1,6-fucosyltransferase or fucosyltransferase 8) via constitutive expression of shRNA (short-hairpin RNA) against FuT8. After the introduction of a FuT8 shRNA expression plasmid under the control of a U6 promoter, CHO-DG44 clones with less than 5% residual FuT8 mRNA expression were isolated by selection for neomycin resistance, followed by low affinity nerve growth factor receptor enrichment and selection for LCA [Lens culinaris (culinary lentil) agglutinin] resistance. The CHO-DG44 clones identified produced highly afucosylated anti-[IGF-1R (insulin-like-growth-factor-1 receptor)] antibodies (up to 88%) that exhibited considerably enhanced ADCC compared with anti-IGF-1R wild-type antibodies produced by parental CHO cells. At the same time, antibody productivity was not significantly decreased. Analysis of stability showed that the clones obtained may be suitable for up-scaling, since low residual levels of FuT8 mRNA and production of afucosylated antibodies were maintained for at least 4 weeks.

Dissection of transcriptional programmes in response to serum and c-Myc in a human B-cell line.

Proliferation of higher eukaryotic cells is triggered by the proto-oncogene c-myc (myc), which is induced downstream of a large number of growth factor receptors. Myc, a basic helix-loop-helix leucine zipper transcription factor, transmits growth signals by up- and downregulation of target genes. The importance of Myc in growth control is well established. However, the number of growth control genes requiring Myc as an essential factor for regulation after mitogenic stimulation of cells is not yet clear. Here, we have studied the transcriptional programme of a human B-cell line, P493-6, in response to Myc and serum. P493-6 cells do not express the endogenous myc, nor is it induced by serum stimulation. Proliferation of the cells is dependent upon both the expression of a tetracycline-regulated myc gene and serum stimulation. Using DNA microarrays, expression profiling was performed following stimulation of cells with serum, with Myc, or with both. We observed serum regulation of >1000 genes. A number of these genes were synergistically or antagonistically regulated by Myc. Moreover, we identified >300 Myc-regulated genes that were almost unresponsive to serum. Gene ontology analysis revealed that a high proportion of Myc target genes are involved in ribosome biogenesis and tRNA metabolism. The data support our current notion that Myc is essential for the regulation of a large number of growth-related genes in B cells, and cannot be replaced by other serum-induced factors.

A role for c-Myc in the regulation of ribosomal RNA processing.

The proto-oncogene c-myc encodes a basic helix-loop-helix leucine zipper transcription factor (c-Myc) that has a profound role in growth control and cell cycle progression. Previous microarray studies identified various classes of c-Myc target genes, including genes involved in ribosome biogenesis. By screening the human B-cell line P493-6 and rat fibroblasts conditionally expressing c-Myc, we could substantially extend the list of c-Myc target genes, particularly those required for ribosome biogenesis. The identification of 38 new c-Myc target genes with nucleolar function, prompted us to investigate processing of ribosomal RNA (rRNA). Using pulse-chase labelling experiments we show that c-Myc regulates the efficiency of rRNA maturation. In serum-stimulated P493-6 cells, only the processing of the 47S rRNA precursor to mature 18S and 28S rRNA, but not the synthesis of the 47S transcript, was dependent on the presence of c-Myc. As processing of rRNA is sensitive to inhibition of cyclin-dependent kinase (cdk) activity by roscovitine, we conclude that c-Myc regulates cell growth and proliferation by the coordinated induction of cdk activity and rRNA processing.

Induction of heat shock protein HSPA6 (HSP70B') upon HSP90 inhibition in cancer cell lines.

Genome-wide transcript profiling to elucidate responses to HSP90 inhibition revealed strong induction of HSPA6 in MCF-7 cells treated with 17-AAG. Time- and dose dependent induction of HSPA6 (confirmed by qPCR and Western Blots) occurred also upon treatment with Radicicol, another HSP90 inhibitor. HSPA6 was not detectable in untreated cells or cells treated with toxins that do not inhibit HSP90, or upon applying oxidative stress. Thus, HSPA6 induction is not a general response to cytotoxic insults. Modulation of HSPA6 levels by siRNA-mediated inhibition or recombinant expression did not influence 17-AAG mediated cell death. HSPA6 induction as a consequence of HSP90 inhibition occurs in various (but not all) cell lines and may be a more specific marker for HSP90 inhibition than induction of other HSP70 proteins.

Identification of rat pancreatic carcinoma genes associated with lymphogenous metastasis.

In order to identify metastasis-associated and promoting genes of pancreatic carcinoma we investigated the transcriptional profile of rat pancreatic carcinoma cell lines BSp73-AS (non-metastatic) and BSp73-ASML (highly metastatic) with Affymetrix GeneChip Array technology. We analyzed the expression profile of 7000 genes. Two hundred and ten genes (3%) were up-regulated and 247 genes (3.5%) were down-regulated in the metastatic cell line based on a fold change of expression of at least 3 and a change factor quality of > or = 2. In order to classify the de-regulated genes we defined the following categories: proteases and protease-related genes, cytokines, receptor tyrosine kinases, other transmembrane proteins/receptors, transcription, cell cycle/apoptosis, signaling, adhesion/extracellular matrix, metabolism, detoxification, protein modification, trafficking, immune response and other genes. We identified de-regulated AP1, FRA-1 and c-myc-mediated transcription in cell line BSp73-ASML. Up-regulation of transmembrane tyrosine kinase receptors c-met, IGFR1, IGFR2 and EGFR family-related ligands such as HB-EGF, TGFa, amphiregulin and neuregulin as well as c-met ligand HGF point to a possible role of this system in metastasis. We identified 56 non-tyrosine kinase transmembrane receptors as new target candidates for inhibition of metastasis, four of them representing already validated targets. In addition, we identified MMP9, uPA, uPAR, cyclin D1 and S100A4 (mts1) as possible contributors of the metastatic phenotype.

Identification of genes associated with metastasis of mammary carcinoma in metastatic versus non-metastatic cell lines.

Cell lines 4A4 and 2C5 are the respective metastatic and non-metastatic variants (nude mouse system) derived from the human mammary carcinoma cell line MDA-MB-435. In order to identify genes associated with or functionally involved in metastasis, we have extended our previous transcriptional profile from 5000 to 12,000 genes using the Affymetrix Gene Chip array technology. Based on a threshold level of a change factor of > or = 2.5, we found that the steady-state level of 40 genes (0.3%) was up-regulated and conversely 89 genes (0.7%) were down-regulated in the metastatic cell line 4A4. The de-regulated genes were classified into categories such as tumor antigens/transmembrane receptors, enzymes, mediators of signaling, cell migration and angiogenesis, cell-cycle-, apoptosis-, differentiation- and growth-factor related genes, tumor suppression, transcription factors and genes encoding components of the extracellular matrix and the cytoskeleton. As possible mediators of invasion we identified DGCR6, osteopontin, autotaxin and the 65 kD phosphoprotein p65. In addition, three sugar-modifying enzymes were up-regulated in cell line 4A4. Profound differences in G-protein-mediated signaling and down-regulation of the tumor-suppressor genes DPC4, BARDI and DLC-1 were noted in the metastatic cell line 4A4.

Antibody mediated CDCP1 degradation as mode of action for cancer targeted therapy.

CUB-domain-containing-protein-1 (CDCP1) is an integral membrane protein whose expression is up-regulated in various cancer types. Although high CDCP1 expression has been correlated with poor prognosis in lung, breast, pancreas, and renal cancer, its functional role in tumor formation or progression is incompletely understood. So far it has remained unclear, whether CDCP1 is a useful target for antibody therapy of cancer and what could be a desired mode of action for a therapeutically useful antibody. To shed light on these questions, we have investigated the cellular effects of a therapeutic antibody candidate (RG7287). In focus formation assays, prolonged RG7287 treatment prevented the loss of contact inhibition caused by co-transformation of NIH3T3 cells with CDCP1 and Src. In a xenograft study, MCF7 cells stably overexpressing CDCP1 reached the predefined tumor volume faster than the parental MCF7 cells lacking endogenous CDCP1. This tumor growth advantage was abolished by RG7287 treatment. In vitro, RG7287 induced rapid tyrosine phosphorylation of CDCP1 by Src, which was accompanied by translocation of CDCP1 to a Triton X-100 insoluble fraction of the plasma membrane. Triggering these effects required bivalency of the antibody suggesting that it involves CDCP1 dimerization or clustering. However, this initial activation of CDCP1 was only transient and prolonged RG7287 treatment induced internalization and down-regulation of CDCP1 in different cancer cell lines. Antibody stimulated CDCP1 degradation required Src activity and was proteasome dependent. Also in three different xenograft models with endogenous CDCP1 expression RG7287 treatment resulted in significant tumor growth inhibition concomitant with substantially reduced CDCP1 levels as judged by immunohistochemistry and Western blotting. Thus, despite transiently activating CDCP1 signaling, the RG7287 antibody has a therapeutically useful mode of action.

Preclinical activity of the type II CD20 antibody GA101 (obinutuzumab) compared with rituximab and ofatumumab in vitro and in xenograft models.

We report the first preclinical in vitro and in vivo comparison of GA101 (obinutuzumab), a novel glycoengineered type II CD20 monoclonal antibody, with rituximab and ofatumumab, the two currently approved type I CD20 antibodies. The three antibodies were compared in assays measuring direct cell death (AnnexinV/PI staining and time-lapse microscopy), complement-dependent cytotoxicity (CDC), antibody-dependent cell-mediated cytotoxicity (ADCC), antibody-dependent cell-mediated phagocytosis (ADCP), and internalization. The models used for the comparison of their activity in vivo were SU-DHL4 and RL xenografts. GA101 was found to be superior to rituximab and ofatumumab in the induction of direct cell death (independent of mechanical manipulation required for cell aggregate disruption formed by antibody treatment), whereas it was 10 to 1,000 times less potent in mediating CDC. GA101 showed superior activity to rituximab and ofatumumab in ADCC and whole-blood B-cell depletion assays, and was comparable with these two in ADCP. GA101 also showed slower internalization rate upon binding to CD20 than rituximab and ofatumumab. In vivo, GA101 induced a strong antitumor effect, including complete tumor remission in the SU-DHL4 model and overall superior efficacy compared with both rituximab and ofatumumab. When rituximab-pretreated animals were used, second-line treatment with GA101 was still able to control tumor progression, whereas tumors escaped rituximab treatment. Taken together, the preclinical data show that the glyoengineered type II CD20 antibody GA101 is differentiated from the two approved type I CD20 antibodies rituximab and ofatumumab by its overall preclinical activity, further supporting its clinical investigation.

Established breast cancer stem cell markers do not correlate with in vivo tumorigenicity of tumor-initiating cells.

The tumor-initiating capacity of primary human breast cancer cells is maintained in vitro by culturing these cells as spheres/aggregates. Inoculation of small cell numbers derived from these non-adherent cultures leads to rapid xenograft tumor formation in mice. Accordingly, injection of more differentiated monolayer cells derived from spheres results in significantly decelerated tumor growth. For our study, two breast cancer cell lines were generated from primary tumors and cultured as mammospheres or as their adherent counterparts. We examined the in vivo tumorigenicity of these cells by injecting serial dilutions into immunodeficient mice. Inoculation of 106 cells per mouse led to rapid tumor formation, irrespective of cell line or culture conditions. However, after injection of only 103 cells, solely sphere cells were highly tumorigenic. In vitro, we investigated differentiation markers, established breast CSC markers and conducted mRNA profiling. Cytokeratin 5 and 18 were increased in both monolayer cell types, indicating a more differentiated phenotype. All cell lines were CD24(-)/CD44(+) and did not express CD133, CD326 or E-cadherin. ALDH1 activity was not detectable in any cell line. A verapamil­sensitive Hoechst side population was present in sphere cells, but there was no correlation with tumorigenicity in vivo. mRNA profiling did not reveal upregulation of relevant transcription factors. In vitro cell cycle kinetics and in vivo tumor doubling times displayed no difference between sphere and monolayer cultures. Our data indicate that intrinsic genetic and functional markers investigated are not indicative of the in vivo tumori-genicity of putative breast tumor-initiating cells.

Modulation of activated leukocyte cell adhesion molecule-mediated invasion triggers an innate immune gene response in melanoma.

Activated leukocyte cell adhesion molecule (ALCAM/CD166) is a progression marker of a variety of cancers, including melanoma, and is a marker for mesenchymal stem cells. ALCAM expression triggers matrix metalloproteinase activity and correlates with the transition between superficial melanoma growth and deep dermal invasion in vivo. We previously showed that manipulating ALCAM functionality could both decrease and increase melanoma invasion, depending on the manner by which ALCAM function was altered. How ALCAM exerts these opposing invasive phenotypes remained elusive. In the present study, we analyzed differences in melanoma cell gene expression in two- and three-dimensional cultures as function of ALCAM-mediated adhesion. We identified a cluster of genes highly responsive to ALCAM functionality and relevant for melanoma invasion. This cluster is characterized by known invasion-related genes similar to L1 neuronal cell adhesion molecule and showed a remarkable induction of several innate immune genes. Unexpectedly, we identified major variations in the expression of genes related to an immunological response when modulating ALCAM function, including complement factors C1r and C1s. The expression and function of these proteinases were confirmed in protein assays and in vivo. Together, our results demonstrate a link between ALCAM functionality and the immune transcriptome, and support the assumption that ALCAM-ALCAM interactions could function as a cell signaling complex to promote melanoma tumor invasion.

Targeted siRNA Delivery and mRNA Knockdown Mediated by Bispecific Digoxigenin-binding Antibodies.

Bispecific antibodies (bsAbs) that bind to cell surface antigens and to digoxigenin (Dig) were used for targeted small interfering RNA (siRNA) delivery. They are derivatives of immunoglobulins G (IgGs) that bind tumor antigens, such as Her2, IGF1-R, CD22, and LeY, with stabilized Dig-binding variable domains fused to the C-terminal ends of the heavy chains. siRNA that was digoxigeninylated at its 3'end was bound in a 2:1 ratio to the bsAbs. These bsAb-siRNA complexes delivered siRNAs specifically to cells that express the corresponding antigen as demonstrated by flow cytometry and confocal microscopy. The complexes internalized into endosomes and Dig-siRNAs separated from bsAbs, but Dig-siRNA was not released into the cytoplasm; bsAb-targeting alone was thus not sufficient for effective mRNA knockdown. This limitation was overcome by formulating the Dig-siRNA into nanoparticles consisting of dynamic polyconjugates (DPCs) or into lipid-based nanoparticles (LNPs). The resulting complexes enabled bsAb-targeted siRNA-specific messenger RNA (mRNA) knockdown with IC(50) siRNA values in the low nanomolar range for a variety of bsAbs, siRNAs, and target cells. Furthermore, pilot studies in mice bearing tumor xenografts indicated mRNA knockdown in endothelial cells following systemic co-administration of bsAbs and siRNA formulated in LNPs that were targeted to the tumor vasculature.Molecular Therapy - Nucleic Acids (2012) 1, e45; doi:10.1038/mtna.2012.39; published online 18 September 2012.

Antibodies against CD20 or B-cell receptor induce similar transcription patterns in human lymphoma cell lines.

BACKGROUND: CD20 is a cell surface protein exclusively expressed on B cells. It is a clinically validated target for Non-Hodgkin's lymphomas (NHL) and autoimmune diseases. The B cell receptor (BCR) plays an important role for development and proliferation of pre-B and B cells. Physical interaction of CD20 with BCR and components of the BCR signaling cascade has been reported but the consequences are not fully understood. METHODOLOGY: In this study we employed antibodies against CD20 and against the BCR to trigger the respective signaling. These antibodies induced very similar expression patterns of up- and down-regulated genes in NHL cell lines indicating that CD20 may play a role in BCR signaling and vice versa. Two of the genes that were rapidly and transiently induced by both stimuli are CCL3 and CCL4. 4 hours after stimulation the concentration of these chemokines in culture medium reaches a maximum. Spleen tyrosine kinase Syk is a cytoplasmic tyrosine kinase and a key component of BCR signaling. Both siRNA mediated silencing of Syk and inhibition by selective small molecule inhibitors impaired CCL3/CCL4 protein induction after treatment with either anti-CD20 or anti-BCR antibodies. CONCLUSION: Our results suggest that treatment with anti-CD20 antibodies triggers at least partially a BCR activation-like response in NHL cell lines.

The Roche Cancer Genome Database 2.0.

BACKGROUND: Cancer is a disease of genome alterations that arise through the acquisition of multiple somatic DNA sequence mutations. Some of these mutations can be critical for the development of a tumor and can be useful to characterize tumor types or predict outcome. DESCRIPTION: We have constructed an integrated biological information system termed the Roche Cancer Genome Database (RCGDB) combining different human mutation databases already publicly available. This data is further extended by hand-curated information from publications.The current version of the RCGDB provides a user-friendly graphical interface that gives access to the data in different ways: (1) Single interactive search by genes, samples, cell lines, diseases, as well as pathways, (2) batch searches for genes and cell lines, (3) customized searches for regularly occurring requests, and (4) an advanced query interface enabling the user to query for samples and mutations by various filter criteria. CONCLUSION: The interfaces of the presented database enable the user to search and view mutations in an intuitive and straight-forward manner. The database is freely accessible at http://rcgdb.bioinf.uni-sb.de/MutomeWeb/.

Human variation databases.

More than 100,000 human genetic variations have been described in various genes that are associated with a wide variety of diseases. Such data provides invaluable information for both clinical medicine and basic science. A number of locus-specific databases have been developed to exploit this huge amount of data. However, the scope, format and content of these databases differ strongly and as no standard for variation databases has yet been adopted, the way data is presented varies enormously. This review aims to give an overview of current resources for human variation data in public and commercial resources.

Strongly enhanced antitumor activity of trastuzumab and pertuzumab combination treatment on HER2-positive human xenograft tumor models.

The human epidermal growth factor receptor (HER) family plays an important role in cell survival and proliferation, and is implicated in oncogenesis. Overexpression of HER2 is associated with aggressive disease and poor prognosis. Trastuzumab is a humanized monoclonal antibody targeting HER2 and has proven survival benefit for women with HER2-positive early and metastatic breast cancer. Pertuzumab, another monoclonal antibody, is a HER2 dimerization inhibitor that binds to a different epitope on HER2 than trastuzumab and inhibits HER2 dimer formation with other HER family members such as HER3 and HER1. We investigated the antitumor activity of these agents alone and in combination in HER2-positive breast and non-small cell lung cancer xenografts. Our data show that the combination of trastuzumab and pertuzumab has a strongly enhanced antitumor effect and induces tumor regression in both xenograft models, something that cannot be achieved by either monotherapy. The enhanced efficacy of the combination was also observed after tumor progression during trastuzumab monotherapy. Near-IR fluorescence imaging experiments confirm that pertuzumab binding to tumors is not impaired by trastuzumab pretreatment. Furthermore, we show by in vitro assay that both trastuzumab and pertuzumab potently activate antibody-dependent cellular cytotoxicity. However, our data suggest that the strongly enhanced antitumor activity is mainly due to the differing but complementary mechanisms of action of trastuzumab and pertuzumab, namely inhibition of HER2 dimerization and prevention of p95HER2 formation.