Challenges for assessing replicability in preclinical cancer biology

We conducted the Reproducibility Project: Cancer Biology to investigate the replicability of preclinical research in cancer biology. The initial aim of the project was to repeat 193 experiments from 53 high-impact papers, using an approach in which the experimental protocols and plans for data analysis had to be peer reviewed and accepted for publication before experimental work could begin. However, the various barriers and challenges we encountered while designing and conducting the experiments meant that we were only able to repeat 50 experiments from 23 papers. Here we report these barriers and challenges. First, many original papers failed to report key descriptive and inferential statistics: the data needed to compute effect sizes and conduct power analyses was publicly accessible for just 4 of 193 experiments. Moreover, despite contacting the authors of the original papers, we were unable to obtain these data for 68% of the experiments. Second, none of the 193 experiments were described in sufficient detail in the original paper to enable us to design protocols to repeat the experiments, so we had to seek clarifications from the original authors. While authors were extremely or very helpful for 41% of experiments, they were minimally helpful for 9% of experiments, and not at all helpful (or did not respond to us) for 32% of experiments. Third, once experimental work started, 67% of the peer-reviewed protocols required modifications to complete the research and just 41% of those modifications could be implemented. Cumulatively, these three factors limited the number of experiments that could be repeated. This experience draws attention to a basic and fundamental concern about replication – it is hard to assess whether reported findings are credible.

Experiments from unfinished Registered Reports in the Reproducibility Project: Cancer Biology

As part of the Reproducibility Project: Cancer Biology, we published Registered Reports that described how we intended to replicate selected experiments from 29 high-impact preclinical cancer biology papers published between 2010 and 2012. Replication experiments were completed and Replication Studies reporting the results were submitted for 18 papers, of which 17 were accepted and published by eLife with the rejected paper posted as a preprint. Here, we report the status and outcomes obtained for the remaining 11 papers. Four papers initiated experimental work but were stopped without any experimental outcomes. Two papers resulted in incomplete outcomes due to unanticipated challenges when conducting the experiments. For the remaining five papers only some of the experiments were completed with the other experiments incomplete due to mundane technical or unanticipated methodological challenges. The experiments from these papers, along with the other experiments attempted as part of the Reproducibility Project: Cancer Biology, provides evidence about the challenges of repeating preclinical cancer biology experiments and the replicability of the completed experiments.

Challenges for assessing replicability in preclinical cancer biology

We conducted the Reproducibility Project: Cancer Biology to investigate the replicability of preclinical research in cancer biology. The initial aim of the project was to repeat 193 experiments from 53 high-impact papers, using an approach in which the experimental protocols and plans for data analysis had to be peer reviewed and accepted for publication before experimental work could begin. However, the various barriers and challenges we encountered while designing and conducting the experiments meant that we were only able to repeat 50 experiments from 23 papers. Here we report these barriers and challenges. First, many original papers failed to report key descriptive and inferential statistics: the data needed to compute effect sizes and conduct power analyses was publicly accessible for just 4 of 193 experiments. Moreover, despite contacting the authors of the original papers, we were unable to obtain these data for 68% of the experiments. Second, none of the 193 experiments were described in sufficient detail in the original paper to enable us to design protocols to repeat the experiments, so we had to seek clarifications from the original authors. While authors were extremely or very helpful for 41% of experiments, they were minimally helpful for 9% of experiments, and not at all helpful (or did not respond to us) for 32% of experiments. Third, once experimental work started, 67% of the peer-reviewed protocols required modifications to complete the research and just 41% of those modifications could be implemented. Cumulatively, these three factors limited the number of experiments that could be repeated. This experience draws attention to a basic and fundamental concern about replication - it is hard to assess whether reported findings are credible.

Investigating the replicability of preclinical cancer biology

Replicability is an important feature of scientific research, but aspects of contemporary research culture, such as an emphasis on novelty, can make replicability seem less important than it should be. The Reproducibility Project: Cancer Biology was set up to provide evidence about the replicability of preclinical research in cancer biology by repeating selected experiments from high-impact papers. A total of 50 experiments from 23 papers were repeated, generating data about the replicability of a total of 158 effects. Most of the original effects were positive effects (136), with the rest being null effects (22). A majority of the original effect sizes were reported as numerical values (117), with the rest being reported as representative images (41). We employed seven methods to assess replicability, and some of these methods were not suitable for all the effects in our sample. One method compared effect sizes: for positive effects, the median effect size in the replications was 85% smaller than the median effect size in the original experiments, and 92% of replication effect sizes were smaller than the original. The other methods were binary - the replication was either a success or a failure - and five of these methods could be used to assess both positive and null effects when effect sizes were reported as numerical values. For positive effects, 40% of replications (39/97) succeeded according to three or more of these five methods, and for null effects 80% of replications (12/15) were successful on this basis; combining positive and null effects, the success rate was 46% (51/112). A successful replication does not definitively confirm an original finding or its theoretical interpretation. Equally, a failure to replicate does not disconfirm a finding, but it does suggest that additional investigation is needed to establish its reliability.

Replication study: androgen receptor splice variants determine taxane sensitivity in prostate cancer

In 2015, as part of the Prostate Cancer Foundation–Movember Foundation Reproducibility Initiative, we published a Registered Report (Shan et al., 2015) that described how we intended to replicate selected experiments from the paper “Androgen Receptor Splice Variants Determine Taxane Sensitivity in Prostate Cancer” (Thadani-Mulero et al., 2014). Here we report the results of those experiments. Growth of tumor xenografts from two prostate cancer xenograft lines, LuCaP 86.2, which expresses wild-type androgen receptor (AR) and AR variant 567, and LuCaP 23.1, which expresses wild-type AR and AR variant 7, were not affected by docetaxel treatment. The LuCaP 23.1 tumor xenografts grew slower than in the original study. This result is different from the original study, which reported significant reduction of tumor growth in the LuCaP 86.2. Furthermore, we were unable to detect ARv7 in the LuCaP 23.1, although we used the antibody as stated in the original study and ensured that it was detecting ARv7 via a known positive control (22rv1, Hörnberg et al., 2011). Finally, we report a meta-analysis of the result.

Registered report: the androgen receptor induces a distinct transcriptional program in castration-resistant prostate cancer in man

The Prostate Cancer Foundation-Movember Foundation Reproducibility Initiative (PCFMFRI) seeks to address growing concerns about reproducibility in scientific research by conducting replications of recent papers in the field of prostate cancer. This Registered Report describes the proposed replication plan of key experiments from “The Androgen Receptor Induces a Distinct Transcriptional Program in Castration-Resistant Prostate Cancer in Man” by Sharma and colleagues (2013), published in Cancer Cell in 2013. Of thousands of targets for the androgen receptor (AR), the authors elucidated a subset of 16 core genes that were consistently downregulated with castration and re-emerged with castration resistance. These 16 AR binding sites were distinct from those observed in cells in culture. The authors suggested that cellular context can have dramatic effects on downstream transcriptional regulation of AR binding sites. The present study will attempt to replicate Fig. 7C by comparing gene expression of the 16 core genes identified by Sharma and colleagues in xenograft tumor tissue compared to androgen treated LNCaP cells in vitro. The Prostate Cancer Foundation-Movember Foundation Reproducibility Initiative is a collaboration between the Prostate Cancer Foundation, the Movember Initiative, and Science Exchange, and the results of the replications will be published by PeerJ.

Registered report: androgen receptor splice variants determine taxane sensitivity in prostate cancer

The Prostate Cancer Foundation-Movember Foundation Reproducibility Initiative seeks to address growing concerns about reproducibility in scientific research by conducting replications of recent papers in the field of prostate cancer. This Registered Report describes the proposed replication plan of key experiments from “Androgen Receptor Splice Variants Determine Taxane Sensitivity in Prostate Cancer” by Thadani-Mulero and colleagues (2014) published in Cancer Research in 2014. The experiment that will be replicated is reported in Fig. 6A. Thadani-Mulero and colleagues generated xenografts from two prostate cancer cell lines; LuCaP 86.2, which expresses predominantly the ARv567 splice variant of the androgen receptor (AR), and LuCaP 23.1, which expresses the full length AR as well as the ARv7 variant. Treatment of the tumors with the taxane docetaxel showed that the drug inhibited tumor growth of the LuCaP 86.2 cells but not of the LuCaP 23.1 cells, indicating that expression of splice variants of the AR can affect sensitivity to docetaxel. The Prostate Cancer Foundation-Movember Foundation Reproducibility Initiative is a collaboration between the Prostate Cancer Foundation, the Movember Foundation and Science Exchange, and the results of the replications will be published by PeerJ.

Replication attempt: "Effect of BMAP-28 antimicrobial peptides on Leishmania major promastigote and amastigote growth: role of leishmanolysin in parasite survival"

This study describes an attempt to replicate experiments from the paper "Effect of BMAP-28 Antimicrobial Peptides on Leishmania major Promastigote and Amastigote Growth: Role of Leishmanolysin in Parasite Survival," which was submitted to the Reproducibility Initiative for independent validation. The cathelicidin bovine myeloid antimicrobial peptide 28 (BMAP-28) and its isomers were previously shown to have potent antiparasitic activity against Leishmania major. We tested the effectiveness of L-BMAP-28 and two of its isomers, the D-amino acid form (D-BMAP-28) and the retro-inverso form (RI-BMAP-28), in both unamidated and amidated forms, as anti-leishmanial agents against Leishmania major promastigotes in vitro. We observed that L-BMAP-28, as well as its D and RI isomers, demonstrate anti-leishmanial activity against L. major promastigotes in vitro. The inhibitory effect was lower than what was seen in the original study. At 2 µM of amidated peptides, the viability was 94%, 36%, and 66% with L-, D- and RI-peptides, versus 57%, 6%, and 18% in the original study.

An open investigation of the reproducibility of cancer biology research

It is widely believed that research that builds upon previously published findings has reproduced the original work. However, it is rare for researchers to perform or publish direct replications of existing results. The Reproducibility Project: Cancer Biology is an open investigation of reproducibility in preclinical cancer biology research. We have identified 50 high impact cancer biology articles published in the period 2010-2012, and plan to replicate a subset of experimental results from each article. A Registered Report detailing the proposed experimental designs and protocols for each subset of experiments will be peer reviewed and published prior to data collection. The results of these experiments will then be published in a Replication Study. The resulting open methodology and dataset will provide evidence about the reproducibility of high-impact results, and an opportunity to identify predictors of reproducibility.

DOI:http://dx.doi.org/10.7554/eLife.04333.001

Abstract 5053: Preliminary results from the Reproducibility Project: Cancer Biology - a replication of 50 high-impact cancer cell biology papers from 2010-2012

The lack of reproducibility of preclinical research is a significant and growing problem which slows basic research and leads to fruitless clinical trials. An increasing number of reports have found discrepancies in published preclinical studies across scientific disciplines. For example:

* Amgen found that 47 of 53 “landmark” oncology publications could not be reproduced.

* Bayer found that their internal results contradicted academic publications in 43 of 67 oncology and cardiovascular projects.

* Dr. John Ioannidis and his colleagues found that of 432 publications purporting sex differences in hypertension, multiple sclerosis, or lung cancer, only one data set was reproducible.

These studies, and the many others that report similar results, highlight a significant problem in the development of new therapies to treat disease. With increasing reports of discrepancies in preclinical publications, pharmaceutical companies are being forced to re-evaluate their reliance on academic research. In fact, Bayer recently decided to halt nearly two-thirds of target-validation projects.In this session, we'll report the work we have done to address this issue. We will bring together researchers and representatives from funders and publishers to discuss the issue of reproducibility. We will share funder and publisher initiatives to promote reproducibility and also discuss research practices that lead to more reproducible research such as using proper statistical tests, reporting all experimental data, experimental blinding, and identification and validation of research reagents. We'll also share preliminary results from an Arnold Foundation-funded study to replicate the 50 most high impact cancer biology studies from 2010-1012.

Reproducibility ResearchAmgen47 of 53 “landmark” oncology publications could not be reproducedBayerinternal results contradicted academic publications in 43 of 67 oncology and cardiovascular projectsDr. John Ioannidis and colleagues432 publications purporting sex differences in hypertension, multiple sclerosis, or lung cancer, only one data set was reproducible

Note: This abstract was not presented at the meeting.

Citation Format: Gunn William, Elizabeth Iorns, Elizabeth Silva, Brian Nosek. Preliminary results from the Reproducibility Project: Cancer Biology - a replication of 50 high-impact cancer cell biology papers from 2010-2012. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 5053. doi:10.1158/1538-7445.AM2014-5053

New forms of checks and balances are needed to improve research integrity

Recent attempts at replicating highly-cited peer-reviewed studies demonstrate that the “reproducibility crisis” is indeed upon us. However, punitive measures against individuals committing research misconduct are neither sufficient nor useful because this is a systemic issue stemming from a lack of positive incentive. As an alternative approach, here we propose a system of checks and balances for the publishing process that involves 1) technical review of methodology by publishers, and 2) incentivizing direct replication of key experimental results. Together, these actions will help restore the self-correcting nature of scientific discovery.

Truncated p110 ERBB2 induces mammary epithelial cell migration, invasion and orthotopic xenograft formation, and is associated with loss of phosphorylated STAT5

Truncated-ERBB2 isoforms (t-ERBB2s), resulting from receptor proteolysis or alternative translation of the ERBB2 mRNA, exist in a subset of human breast tumors. t-ERBB2s lack the receptor extracellular domain targeted by therapeutic anti-ERBB2 antibodies and antibody–drug conjugates, including trastuzumab, trastuzumab-DM1 and pertuzumab. In clinical studies, expression of t-ERBB2 in breast tumors correlates with metastasis as well as trastuzumab resistance. By using a novel immuno-microarray method, we detect a significant t-ERBB2 fraction in 18 of 31 (58%) of immunohistochemistry (IHC)3+ ERBB2+ human tumor specimens, and further show that t-ERBB2 isoforms are phosphorylated in a subset of IHC3+ samples (10 of 31, 32%). We investigated t-ERBB2 biological activity via engineered expression of full-length and truncated ERBB2 isoforms in human mammary epithelial cells (HMECs), including HMEC and MCF10A cells. Expression of p110 t-ERBB2, but not p95m (m=membrane, also 648CTF) or intracellular ERBB2s, significantly enhanced cell migration and invasion in multiple cell types. In addition, only expression of the p110 isoform led to human breast epithelial cell (HMLE) xenograft formation in vivo. Expression of t-ERBB2s did not result in hyperactivation of the phosphoinositide kinase-3/AKT or mitogen-activated protein kinase signaling pathways in these cells; rather, phosphoproteomic array profiling revealed attenuation of phosphorylated signal transducer and activator of transcription 5 (STAT5) in p110-t-ERBB2-expressing cells compared to controls. Short hairpin-mediated silencing of STAT5 phenocopied p110-t-ERBB2-driven cell migration and invasion, while expression of constitutively active STAT5 reversed these effects. Thus, we provide novel evidence that (1) expression of p110 t-ERBB2 is sufficient for full transformation of HMEC, yielding in vivo xenograft formation, and (2) truncated p110 t-ERBB2 expression is associated with decreased phosphorylation of STAT5.

Whole genome in vivo RNAi screening identifies the leukemia inhibitory factor receptor as a novel breast tumor suppressor

Cancer is caused by mutations in oncogenes and tumor suppressor genes, resulting in the deregulation of processes fundamental to the normal behavior of cells. The identification and characterization of oncogenes and tumor suppressors has led to new treatment strategies that have significantly improved cancer outcome. The advent of next generation sequencing has allowed the elucidation of the fine structure of cancer genomes, however, the identification of pathogenic changes is complicated by the inherent genomic instability of cancer cells. Therefore, functional approaches for the identification of novel genes involved in the initiation and development of tumors are critical. Here we report the first whole human genome in vivo RNA interference screen to identify functionally important tumor suppressor genes. Using our novel approach, we identify previously validated tumor suppressor genes including TP53 and MNT, as well as several novel candidate tumor suppressor genes including leukemia inhibitory factor receptor (LIFR). We show that LIFR is a key novel tumor suppressor, whose deregulation may drive the transformation of a significant proportion of human breast cancers. These results demonstrate the power of genome wide in vivo RNAi screens as a method for identifying novel genes regulating tumorigenesis.

PD01-09: Identifying Novel Mechanisms of Resistance to Lapatinib in ERBB2+ Breast Cancer Cells through Whole Genome Mutational Analysis

Background: Overexpression of the epidermal growth factor receptor ERBB2 (HER2) is found in 20% of human breast cancers. Therapies targeting ERBB2 including trastuzumab and lapatinib have significantly improved the outlook for women with ERBB2+ breast cancer. However, resistance to these agents occurs frequently and remains a significant clinical problem. In the case of lapatinib resistance, the mechanism(s) of resistance remain poorly understood, since the current proposed rationale thought to limit lapatinib's anti-tumor effects has been difficult to reconcile with clinical data. Therefore, we hypothesize that novel mechanisms of resistance could be identified by mapping genomic variations in ERBB2+ cells with acquired resistance to lapatinib. The identification of such mutations may provide insights into mechanisms of resistance and may indicate therapeutic strategies to overcome lapatinib resistance in ERBB2+ breast cancer.

Material and Methods: SKBR3 breast cancer cells resistant to lapatinib were generated through serial passage by exposure of drug sensitive parental SKBR3 cells to increasing concentrations of lapatinib up to the peak plasma concentration observed in human subjects (2.6 uM (SK-lapR)). Multiple signaling pathways in lapatinib sensitive and resistant cells were interrogated by Reverse Phase Protein Array (RPPA) and western blot analysis. To identify genome wide somatic mutations, the Exome of lapatinib resistant and sensitive SKBR3 cells was sequenced utilizing next generation deep sequencing. Following exclusion of germline variants, the acquired gene mutations in lapatinib resistant SKBR3 cells were confirmed by DNA re-sequencing of PCR amplified DNA segments.

Results and Discussion: Analysis of activated signaling pathways in lapatinib resistant and sensitive SKBR3 cells did not confirm any of the previously proposed mechanisms of resistance. In particular, these cells show no activation of AKT or alternative receptor tyrosine kinases such as IGF-IR, ERBB3 or c-Met. However they exhibit sustained activation of mTORC1 and ERK1/2, as well as phosphorylation of STAT3, STAT5, rpS6 and CREB. Initial sequence analysis of exome and transcriptome reveals the presence of 76 single nucleotide variants/Indels differing between sensitive and resistant cells with 34/76 validated as true mutations present in the genome of lapatinib resistant SKBR3 cells, including mutations in LATS2, MAP3K5, SMAD3 and PDGFRA. This is the first exome sequence analysis to be reported which defines a drug resistant phenotype in ERBB2+ breast cancer. Ongoing work includes investigation of mutations as drug resistance mediators and analysis of copy number variations and gene fusions/translocations to systematically search for molecular alterations, with the goal of providing a rationale for the design of new combination therapies aimed at lapatinib resistance for ERBB2+ breast cancer.

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr PD01-09.

P2-01-25: Truncated p110 ERBB2 (CTF611) Increases Migration and Invasion of Breast Epithelial Cells by Inhibiting STAT5b Activation

Background: Truncated ERBB2 receptors are present in a subset of human ERBB2+ amplified/overexpressing breast tumors, and are associated with trastuzumab resistance, metastasis, and poor clinical prognosis. However, whether truncated ERBB2 receptors are drivers of metastasis has not been well defined. In this study, we examined effects of full-length (p185) and truncated (p110) ERBB2 on the migration and invasion of human mammary epithelial cells, including HMLE and MCF10A cells.

Material and Methods: Recombinant p185 and p110 ERBB2 were stably expressed in human mammary epithelial cells (HMLE) and MCF10A cells via retroviral vector. Expression of comparable levels of p185 and p110 in cells was confirmed by western blot. The phosphorylation states of downstream signaling proteins including STAT5 were assayed via phosphoproteomics and Collaborative Enzyme Enhanced Reactive (CEER™) immunoassay. The effects of the p110 constructs on cell migration and invasion were investigated by transwell assays. shRNA-encoding lentivirus was used for specific silencing of STAT5b in HMLE cells, and STAT5b silencing was confirmed at the protein level using western blot.

Results and Discussion: Expression of p110 ERBB2 increased cell migration (HMLE, p = 0.04; MCF10A, p< 0.01) and invasion (HMLE, p= 0.03) when compared to expression of p185. Furthermore, expression of p110 in HMLE cells was associated with reduced phosphorylation of STAT5b. shRNA mediated silencing of STAT5b was sufficient to increase the migration (p < 0.01) and invasion of HMLE cells, phenocopying the p110 driven effects on HMLE cells. In clinical studies, loss of activated STAT5 protein correlates with breast cancer progression and is a negative predictor of survival. By analyzing publicly available gene expression datasets, we found that STAT5b mRNA expression is also significantly decreased in breast cancer compared to normal breast tissues in several studies, as well as in ERBB2 amplified vs. nonamplified samples. To our knowledge, this is the first reported perturbation of STAT signaling by truncated ERBB2 receptor, and suggests a mechanism by which truncated p110 ERBB2 (CTF611) increases migration and invasion of breast epithelial cells. This study extends the available data regarding STAT5 loss in breast cancer progression.

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P2-01-25.

Abstract 1211: Novel detection and characterization of truncated ERBB2 demonstrates potent oncogenicity of the p110 isoform in human mammary epithelial cells

The full-length p185 ERBB2 (HER2) receptor tyrosine kinase is a validated target for breast cancer (BC) therapy. It is now appreciated that a portion of ERBB2+ BC expresses truncated isoforms of the receptor, including the membrane localized forms p95 and p110. Importantly, t-ERBB2s lack the extracellular domain epitope of trastuzumab and act as a resistance mechanism to therapy. Data from transgenic mouse models implicate expression of t-ERBB2s in tumor formation and metastasis. Furthermore, human clinical studies have revealed that expression of t-ERBB2s correlates with poor patient outcome, increased nodal involvement, and increased metastasis.

In this study, we used partially transformed breast epithelial cells (HMLEs) to extend the studies of t-ERBB2 expression in human cells. Stable expression of cDNAs coding for p110 t-ERBB2 led to increased migration (79 cells/ well versus 39, p = 0.04), and invasion (583 cell/ well vs. 290) p = 0.03) of p110 versus p185 expressing HMLE cells, respectively. Importantly, expression of p110 t-ERBB2, but not other isoforms, led to xenograft formation by HMLE cells in NOD/SCID mice (p110, 10 of 12 possible xenografts formed, p185, no xenografts formed).

Accurate clinical profiling of t-ERBB2 expression may aid in selection of ERBB2-targeted therapy. To this end, we used a novel proximity-mediated immunoassay (CoPIA) to detect and quantify t-ERBB2 receptors in patient samples. Analysis of a cohort of 74 primary breast tumors revealed that 52% of breast tumor samples with a high ERBB2-IHC score (3+) contained t-ERBB2, with 32% containing phosphorylated t-ERBB2. In several samples t-ERBB2 accounted for 10-20% of the total ERBB2 receptor population in tumor cells. Finally, fine-needle aspirates of ERBB2+ metastases also revealed expression and phosphorylation of t-ERBB2s.

In addition to its role in prognosis and therapeutic guidance, p110 t-ERBB2 represents an attractive target for therapy. We have generated several mouse monoclonal antibodies targeting the N-termini of p110 t-ERBB2s. Binding of p110 on the surface of living cells was assayed by flow cytometry, and the biological activity of these antibodies is currently being assessed in vitro and in vivo.

Together, these data imply a functional role for the p110 t-ERBB2 isoform in tumorigenesis, invasion, and migration. Furthermore, expression of t-ERBB2 in the majority of IHC 3+ tumor samples reveals a significant incidence of t-ERBB2 in human cancers. Targeting of p110 t-ERBB2 may allow for novel therapeutic intervention in ERBB2+ primary and metastatic breast cancer.

Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 1211. doi:10.1158/1538-7445.AM2011-1211

Abstract P5-06-03: Truncated ERBB2 Receptors: Diagnostic and Therapeutic Targets

Background:

Patients with ERBB2+ breast cancer have aggressive disease and poor prognoses. It is now apparent that many ERBB2+ tumors also express truncated ERBB2 receptors (t-ERBB2s), namely p110 and p95. Increased expression of t-ERBB2s by breast tumors correlates with increased nodal involvement, distant metastasis, and poor clinical outcome in patients. Because t-ERBB2s lack the epitope bound by trastuzumab, expression of high levels of these isoforms may designate patients who would be better treated with alternative anti-ERBB2 therapy such as lapatinib; unfortunately, there is currently no clinical method to distinguish full-length p185- versus t-ERBB2 in patient tumor samples. Materials and Methods:

Recombinant forms of p185-, p110- and p95-ERBB2 were constructed using standard cloning techniques and expressed in human mammary epithelial cells (HMLE) via retroviral vector. The expression and subcellular localization of constructs were confirmed by western blot analysis and confocal microscopy. The ability of p185- and t-ERBB2 constructs to transform HMLE cells was evaluated using soft agar assays, and the effects on migration and invasion of these cells were investigated by transwell assays. Finally, the in vivo tumor formation by p185- vs. t-ERBB2 expressing cells was evaluated in immunodeficient mice. Additionally, a novel proximity-based antibody-capture method method to discern full-length versus t-ERBB2 in patient tumor samples was assessed (COPIA). Block tumors and fine-needle aspirates from patient tumor samples were used for quantifying total and phosphorylated ERBB2 receptors.

Results and Discussion:

Recombinant p185- and t-ERBB2 constructs were stably expressed in HMLE cells, and were correctly targeted to the cell membrane, as shown by confocal immunofluorescence microscopy and immunoblot. Expression of p110 t-ERBB2 increased migration and invasion of HMLE cells compared to p185 ERBB2 (P<0.0001), while p110, p95m and p185 ERBB2s were equally effective at enhancing anchorage-independent growth. In vivo, expression of p110 t-ERBB2 but not other isoforms led to increased tumor formation in mice compared to controls (P<0.005). No apparent phenotypes were elicited by expression of intracellular t-ERBB2 isoforms. Using COPIA testing, t-ERBB2 isoforms were detected in strongly ERBB2- positive tumors (16 of 31 samples, 52%) and were phosphorylated in 10 of 31 (32%). As expected, t-ERBB2s were not detected in ERBB2-negative tumor samples.

Truncated ERBB2s, particularly p110, may be major pathogenic drivers in ERBB2+ cancers. These isoforms may accelerate disease progression by promoting invasion and metastasis, and likely mediate resistance to trastuzumab and other therapies. Thus, t-ERBB2s represent attractive novel targets for diagnosis and treatment of ERBB2+ breast cancers.

Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P5-06-03.

Abstract P5-05-02: Whole Genome In Vivo RNA Interference Screening Identifies the Leukemia Inhibitory Factor Receptor as a Novel Breast Tumor Suppressor

Background: Cancer is caused by mutations in oncogenes and tumor suppressor genes resulting in the deregulation of processes fundamental to the normal behavior of cells. The identification and characterization of oncogenes and tumor suppressors has led to new treatment strategies that have significantly improved cancer outcome. The advent of next generation sequencing has allowed the elucidation of the fine structure of cancer genomes, however, the identification of pathogenic changes is complicated by the inherent genomic instability of cancer cells. Therefore, functional approaches for the identification of novel genes involved in the initiation and development of tumors are critical. Methods: In order to identify functionally important tumor suppressor genes we have conducted the first human whole genome in vivo RNA interference (RNAi) screen. Partially transformed human mammary epithelial cells (HMLEs), which do not form tumors in immunodeficient mice, were infected with the Expression Arrest™ GIPZ lentiviral shRNA library consisting of 62,000 shRNAs targeting the whole human genome, and injected into the mammary fat pad of immunodeficient mice. shRNAs that silenced tumor suppressor genes fully transformed the mammary epithelial cells resulting in tumor formation. Candidate tumor suppressor genes were identified by PCR amplification and sequencing of tumor integrated shRNAs. For validation, candidate tumor suppressor genes were silenced in HMLEs and ectopically expressed in fully transformed breast cancer cells. The effect of modifying gene expression on the transformed phenotype was assessed using soft agar colony formation assays. Clinical significance was determined by comparing expression in normal and cancerous human breast tissue using Oncomine Research. Results and Discussion: Using our novel approach, we identify previously validated tumor suppressor genes including TP53 and MNT, as well as several novel candidate tumor suppressor genes including leukemia inhibitory factor receptor (LIFR). Silencing LIFR expression with multiple shRNA constructs fully transformed human mammary epithelial cells resulting in enhanced colony formation in soft agar (P<0.05). Furthermore, overexpression of LIFR significantly inhibited colony formation in soft agar of fully transformed MDA231 and MCF7 breast cancer cells (P<0.01). In addition, our analysis of clinical data revealed that LIFR expression is significantly decreased in a large percentage of human cancers including breast (P<0.0001), lung (P<0.0001), hepatocellular (P<0.0001) and gastrointestinal tumors (P<0.0001). These results validate LIFR as a previously unidentified highly significant tumor suppressor, and also demonstrate the power of whole genome in vivo RNAi screens as a method for identifying novel genes regulating tumorigenesis.

Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P5-05-02.

The role of SATB1 in breast cancer pathogenesis

BACKGROUND

SATB1 has been previously proposed as a key protein that controls the development and progression of breast cancer. We explored the potential of the SATB1 protein as a therapeutic target and prognostic marker for human breast cancer.

METHODS

We used aggressive (MDA-MB-231 and BT549) and nonaggressive (SKBR3 and MCF7) breast cancer cell lines to investigate the potential of SATB1 as a therapeutic target. SATB1 mRNA expression was silenced in aggressive cells by use of short hairpin RNAs against SATB1. SATB1 was overexpressed in nonaggressive cells by use of SATB1 expression vectors. We assessed the effect of modifying SATB1 expression on the transformed phenotype by examining anchorage-independent cell proliferation, acinar morphology on matrigel, and migration by wound healing in cultured cells. We examined tumor formation and metastasis, respectively, by use of orthotopic mammary fat pad and tail vein xenograft mouse models (mice were used in groups of six, and in total, 96 mice were used). SATB1 mRNA expression was compared with outcome for patients with primary breast cancer from six previous microarray studies that included a total of 1170 patients. All statistical tests were two-sided.

RESULTS

The transformed phenotype was not suppressed by SATB1 silencing in aggressive cells and was not enhanced by ectopic expression of SATB1 in nonaggressive cells. Modifying SATB1 expression did not alter anchorage-independent cell proliferation, invasive acinar morphology, or cell migration in cultured cells and did not affect tumor formation or metastasis in xenograft mouse models. In addition, SATB1 expression was not associated with decreased overall survival of patients with primary breast cancer in six previous independent microarray studies (overall odds ratio = 0.80, 95% confidence interval = 0.62 to 1.03, P = .10).

CONCLUSION

In contrast to previous studies, we found that SATB1 expression did not promote breast cancer progression and was not associated with breast cancer outcome.

FGFR1 amplification drives endocrine therapy resistance and is a therapeutic target in breast cancer

Amplification of fibroblast growth factor receptor 1 (FGFR1) occurs in approximately 10% of breast cancers and is associated with poor prognosis. However, it is uncertain whether overexpression of FGFR1 is causally linked to the poor prognosis of amplified cancers. Here, we show that FGFR1 overexpression is robustly associated with FGFR1 amplification in two independent series of breast cancers. Breast cancer cell lines with FGFR1 overexpression and amplification show enhanced ligand-dependent signaling, with increased activation of the mitogen-activated protein kinase and phosphoinositide 3-kinase-AKT signaling pathways in response to FGF2, but also show basal ligand-independent signaling, and are dependent on FGFR signaling for anchorage-independent growth. FGFR1-amplified cell lines show resistance to 4-hydroxytamoxifen, which is reversed by small interfering RNA silencing of FGFR1, suggesting that FGFR1 overexpression also promotes endocrine therapy resistance. FGFR1 signaling suppresses progesterone receptor (PR) expression in vitro, and likewise, amplified cancers are frequently PR negative, identifying a potential biomarker for FGFR1 activity. Furthermore, we show that amplified cancers have a high proliferative rate assessed by Ki67 staining and that FGFR1 amplification is found in 16% to 27% of luminal B-type breast cancers. Our data suggest that amplification and overexpression of FGFR1 may be a major contributor to poor prognosis in luminal-type breast cancers, driving anchorage-independent proliferation and endocrine therapy resistance.

Truncated p95erbB2 Isoforms Are Capable of Transforming Human Mammary Epithelial Cells

Background:Objective clinical response to trastuzumab monotherapy in erbB2-amplified first line metastatic breast cancer is 34% (Vogel, et al., JCO 20: 719-26, 2002). Amongst patients who respond, most develop resistance (defined by disease progression on trastuzumab). One proposed mechanism of trastuzumab resistance is proteolytic cleavage of erbB2 receptor from its full-length (p185) form into truncated, constitutively active p95. Increased expression of p95erbB2 correlates with increased nodal involvement and poor clinical outcome. Because p95erbB2 lacks the trastuzumab binding epitope, expression may designate patients who would be suitable for treatment with erbB2 kinase inhibitors.Materials and Methods:Recombinant p185erbB2 and p95erbB2 constructs were stably expressed in several cell types via retroviral vector. Additionally, an intracellular form of p95erbB2 that arises via alternative translation and an intracellular p95erbB2 construct containing two copies of a nuclear localization sequence were also expressed. Expression and proper subcellular localization of constructs were confirmed by cell fractionation, western blot analysis and confocal microscopy. Transformation of human mammary epithelial (HMEC) and NIH3T3 cells by p185erbB2 and p95erbB2 isoforms was evaluated by anchorage independent growth using a quantitative fluorescent soft agar assay, and effects on migration and invasion of these cells were investigated by wound-healing and transwell assays. Cells transfected with oncogenic Ras or empty vector were used as positive and negative controls in these experiments.Results and Discussion:Recombinant p185erbB2 and p95erbB2 constructs were stably expressed in HMEC and NIH3T3 cells, and were correctly directed to the cell membrane; nuclear targeted intracellular p95erbB2 was correctly localized to the cell nucleus. Both p185 erbB2 and membrane-bound truncated p95erbB2 were sufficient to transform HMEC cells as compared to empty vector control transfected cells [mean fluorescence intensity empty vector control 2480 ± 464 (1 standard deviation); mean fluorescence intensity p185erbB2 9208 ± 2528, p= 0.0106; mean fluorescence intensity p95erbB2 6615 ± 1588, p= 0.0124)] as was the positive control oncogenic Ras (mean fluorescence intensity 4350 ± 433, p=0.0069). Interestingly, nuclear-targeted p95erbB2 was also sufficient to transform HMEC cells (mean fluorescence intensity 6492 ± 818, p=0.0018). These data support the hypothesis that truncated p95erbB2 species may be major pathogenic drivers in erbB2-amplified cancers. P95erbB2 therefore represents an attractive target for diagnosis and treatment of erbB2+ breast cancer.

Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 3136.

Dissecting resistance to endocrine therapy in breast cancer

Drugs that target the reliance of tumor cells upon estrogen signalling have revolutionised the treatment of breast cancer. Despite this, resistance to these endocrine therapies limits their utility. While the study of individual genes has contributed greatly to understanding drug resistance, relatively unbiased screening approaches may also be illuminating. The results of a high-throughput RNA interference screen identifying novel determinants of tamoxifen resistance support this conjecture and demonstrate that such approaches can identify clinically relevant genes, such as CDK10.

Identification of CDK10 as an important determinant of resistance to endocrine therapy for breast cancer

Therapies that target estrogen signaling have transformed the treatment of breast cancer. However, the effectiveness of these agents is limited by the development of resistance. Here, an RNAi screen was used to identify modifiers of tamoxifen sensitivity. We demonstrate that CDK10 is an important determinant of resistance to endocrine therapies and show that CDK10 silencing increases ETS2-driven transcription of c-RAF, resulting in MAPK pathway activation and loss of tumor cell reliance upon estrogen signaling. Patients with ER alpha-positive tumors that express low levels of CDK10 relapse early on tamoxifen, demonstrating the clinical significance of these observations. The association of low levels of CDK10 with methylation of the CDK10 promoter suggests a mechanism by which CDK10 expression is reduced in tumors.

Utilizing RNA interference to enhance cancer drug discovery

With the development of RNA interference (RNAi) libraries, systematic and cost-effective genome-wide loss-of-function screens can now be carried out with the aim of assessing the role of specific genes in neoplastic phenotypes, and the rapid identification of novel drug targets. Here, we discuss the existing applications of RNAi in cancer drug discovery and highlight areas in this process that may benefit from this technology in the future.