Paulo Nuin

EMT transcription factors snail and slug directly contribute to cisplatin resistance in ovarian cancer

BackgroundThe epithelial to mesenchymal transition (EMT) is a molecular process through which an epithelial cell undergoes transdifferentiation into a mesenchymal phenotype. The role of EMT in embryogenesis is well-characterized and increasing evidence suggests that elements of the transition may be important in other processes, including metastasis and drug resistance in various different cancers.MethodsAgilent 4 × 44 K whole human genome arrays and selected reaction monitoring mass spectrometry were used to investigate mRNA and protein expression in A2780 cisplatin sensitive and resistant cell lines. Invasion and migration were assessed using Boyden chamber assays. Gene knockdown of snail and slug was done using targeted siRNA. Clinical relevance of the EMT pathway was assessed in a cohort of primary ovarian tumours using data from Affymetrix GeneChip Human Genome U133 plus 2.0 arrays.ResultsMorphological and phenotypic hallmarks of EMT were identified in the chemoresistant cells. Subsequent gene expression profiling revealed upregulation of EMT-related transcription factors including snail, slug, twist2 and zeb2. Proteomic analysis demonstrated up regulation of Snail and Slug as well as the mesenchymal marker Vimentin, and down regulation of E-cadherin, an epithelial marker. By reducing expression of snail and slug, the mesenchymal phenotype was largely reversed and cells were resensitized to cisplatin. Finally, gene expression data from primary tumours mirrored the finding that an EMT-like pathway is activated in resistant tumours relative to sensitive tumours, suggesting that the involvement of this transition may not be limited to in vitro drug effects.ConclusionsThis work strongly suggests that genes associated with EMT may play a significant role in cisplatin resistance in ovarian cancer, therefore potentially leading to the development of predictive biomarkers of drug response or novel therapeutic strategies for overcoming drug resistance.

PTEN Genomic Deletions that Characterize Aggressive Prostate Cancer Originate Close to Segmental Duplications

Deletion of PTEN at 10q23.3 occurs in ∼40% of human prostate cancers and is associated with aggressive metastatic potential, poor prognosis, and androgen‐independence. This high frequency of recurrent PTEN deletions in prostate cancer suggests there may be unusual genomic features close to this locus that facilitate DNA alteration at 10q23.3. To explore possible mechanisms for deletions in the PTEN region, a meta‐analysis of 311 published human genome array datasets was conducted and determined that the minimal prostate cancer‐associated deletion at 10q23.3 corresponds to ∼2.06 MB region flanked by BMPR1A and FAS. On a separate cohort comprising an additional 330 tumors, four‐color fluorescence in situ hybridization analysis using probes for BMPR1A, FAS, cen(10), and PTEN showed that 132 of 330 (40%) tumors had PTEN loss, 50 (15%) of which were homozygous losses (comprising in total 100 deletion events). Breakpoints between PTEN and BMPR1A or FAS were subsequently mapped in 100 homozygous and 82 hemizygous PTEN losses, revealing that 125/182 PTEN microdeletions occurred within the 940 kB interval between BMPR1A and PTEN. Furthermore, this breakpoint interval coincides with a repeat‐rich region of 414 kB containing the SD17 and SD18 segmental duplications, which contain at least 13 homologous inverted repeat sequences. Together, these data suggest that a strong selective growth advantage for loss of PTEN and upregulation of PI3K/AKT, combined with the close proximity of PTEN to a large unstable segment of repeated DNA comprising SD17 and SD18, can lead to recurrent microdeletions of the PTEN gene in prostate cancer.

Identification of the IGF1/PI3K/NF κB/ERK gene signalling networks associated with chemotherapy resistance and treatment response in high-grade serous epithelial ovarian cancer

BackgroundResistance to platinum-based chemotherapy remains a major impediment in the treatment of serous epithelial ovarian cancer. The objective of this study was to use gene expression profiling to delineate major deregulated pathways and biomarkers associated with the development of intrinsic chemotherapy resistance upon exposure to standard first-line therapy for ovarian cancer.MethodsThe study cohort comprised 28 patients divided into two groups based on their varying sensitivity to first-line chemotherapy using progression free survival (PFS) as a surrogate of response. All 28 patients had advanced stage, high-grade serous ovarian cancer, and were treated with standard platinum-based chemotherapy. Twelve patient tumours demonstrating relative resistance to platinum chemotherapy corresponding to shorter PFS (< eight months) were compared to sixteen tumours from platinum-sensitive patients (PFS > eighteen months). Whole transcriptome profiling was performed using an Affymetrix high-resolution microarray platform to permit global comparisons of gene expression profiles between tumours from the resistant group and the sensitive group.ResultsMicroarray data analysis revealed a set of 204 discriminating genes possessing expression levels which could influence differential chemotherapy response between the two groups. Robust statistical testing was then performed which eliminated a dependence on the normalization algorithm employed, producing a restricted list of differentially regulated genes, and which found IGF1 to be the most strongly differentially expressed gene. Pathway analysis, based on the list of 204 genes, revealed enrichment in genes primarily involved in the IGF1/PI3K/NF κB/ERK gene signalling networks.ConclusionsThis study has identified pathway specific prognostic biomarkers possibly underlying a differential chemotherapy response in patients undergoing standard platinum-based treatment of serous epithelial ovarian cancer. In addition, our results provide a pathway context for further experimental validations, and the findings are a significant step towards future therapeutic interventions.

MicroRNA Signature Helps Distinguish Early from Late Biochemical Failure in Prostate Cancer

PURPOSE Prostate-specific antigen testing has led to overtreatment of prostate cancer (PCa). Only a small subset of PCa patients will have an aggressive disease that requires intensive therapy, and there is currently no biomarker to predict disease aggressiveness at the time of surgery. MicroRNAs (miRNAs) are reported to be involved in PCa pathogenesis. METHODS This study involved 105 participants. For the discovery phase, prostatectomy samples were dichotomized to high-risk (n = 27, biochemical failure <36 months after prostatectomy) and low-risk groups (n = 14, ≥ 36 months without biochemical failure). Expression of 754 mature miRNAs was compared between the 2 groups. Linear regression models were built to accurately predict biochemical failure risk. miRNA mimics were transfected into PCa model cell lines to test effects on proliferation and to deduce responding signaling pathways. RESULTS We identified 25 differentially expressed miRNAs between the biochemical failure risk groups. Based on the expression of 2-3 miRNAs, 3 logistic regression models were developed, each with a high positive predictive value. Candidate miRNAs and the best-performing model were also verified on an independent PCa set. miRNA-152, featured in the models, was further investigated by using cell line models and was shown to affect cell proliferation. Predicted interaction between miR-152 and (mRNA)ERBB3 (erythroblastic leukemia viral oncogene homolog 3) was experimentally validated in vitro. CONCLUSIONS miRNAs can help to predict biochemical failure risk at the time of prostatectomy.

No correlation between the molecular subtype of COL1A1-PDGFB fusion gene and the clinico-histopathological features of dermatofibrosarcoma protuberans.

TO THE EDITOR Dermatofibrosarcoma protuberans (DFSP) is a dermal tumor of intermediate malignancy. Propensity for metastases is low but local recurrence risk is high unless a wide surgical excision is performed. In addition to the classical presentation of DFSP, several clinical and histological variants have been described, such as DFSP-containing fibrosarcomatous areas or malignant fibrous histiocytoma areas, Bednar tumor, giant cell fibroblastoma, and superficial adult fibrosarcoma (Simon et al., 1997; Sheng et al., 2001; Maire et al., 2002a; Sirvent et al., 2003; Bianchini et al., 2007; Szollosi et al., 2007). DFSP and related tumors are molecularly characterized by the presence of an abnormal chimeric gene that fuses COL1A1 (17q21.3) with platelet-derived growth factor-B chain (PDGFB) (22q13.1) (Simon et al., 1997). The detection of COL1A1– PDGFB is mandatory in case of unusual clinical or histological presentation, as well as in pediatric cases (Sirvent et al., 2003; Bianchini et al., 2007; Maire et al., 2007). A striking feature of COL1A1–PDGFB fusion consists of the high variability of the break point location between the exons 6 and 49 of COL1A1 (Sirvent et al., 2003; Bianchini et al., 2007). In contrast, the break point in PDGFB is consistently located in the first intron. We have investigated whether the COL1A1 break point location is associated with clinical or histological features. In such a case, the precise detection of the COL1A1 break point detection by using multiplex reverse transcriptase–PCR and sequencing could be useful for diagnosis or evaluation of prognosis. For this purpose we compiled a series of 172 DFSP cases with both molecular and clinico-histological data (Supplementary Table S1), including (i) 35 cases of DFSP or related tumors addressed for molecular diagnosis referral to the laboratory of Solid Tumors Genetics (Nice, France) from nine different French Hospitals and previously unreported; (ii) 22 DFSP cases previously published by our group (Pedeutour et al., 1994, 1995, 1996; Simon et al., 1997; Greco et al., 1998; Navarro et al., 1998; Maire et al., 2002a, b, c, 2007; Terrier-Lacombe et al., 2003; Jouary et al., 2007); (iii) 115 cases published between 1997 and 2008 (O’Brien et al., 1998; Wang et al., 1999, 2000; Vanni et al., 2000; Sheng et al., 2001; Gokden et al., 2003; Sandberg et al., 2003; Sirvent et al., 2003; Saeki et al., 2003a, b, 2005, 2006; Martin et al., 2005; Craver et al., 2006; Kashima et al., 2006; Llombart et al., 2006, 2008; Nakanishi et al., 2007; Szollosi et al., 2007; Takahira et al., 2007; Patel et al., 2008). The description of the clinical and histological features of the 172 cases is compiled in Supplementary Table S1. Continuous data (age, tumor size) were described with mean, SD, and range (minimum–maximum); nominal scale data (gender, anatomic location, histological variant) were described by frequencies and relative frequencies (expressed as percentage). A Fisher exact test showed no significant differences in terms of gender, clinical presentation, and histological variant between our series and tumors studied by other groups (Table 1). Similarly, an unpaired Student’s t-test showed no significant differences in the two series in terms of tumor size. In contrast, a slight difference between the two series was observed for the anatomic location and the mean age. The first could be explained by reports of exceptional anatomical locations in the literature, and the later by a more frequent recruitment of pediatric patients in our laboratory. The very high variability of the COL1A1 break point position was confirmed, as 38 different COL1A1 exons have been identified. The most frequently rearranged COL1A1 exons were exon 25 (n1⁄4 19), 32 (n1⁄418), and 47 (n1⁄4 15) (Figure 1). A standard analysis of variance of the COL1A1 break point position with respect to the clinico-histological parameters was performed by grouping the break points into four distinct groups of 11 exons each (group A: exons 6–16; group B: exons 17–27; group C: exons 28–38; and group D: exons 39–49). The following parameters were tested for the COL1A1 break point position: age, histological variant, anatomic location, and gender. None of the parameters tested showed a significant association with the either group. Only the anatomic location showed a slight trend, P1⁄40.01843 (Supplementary Table S2). Also, no significant correlation between the most represented COL1A1 break points (n45) (exons 7, 25, 29, 32, 33–34, 40, 42, 46, and 47) and clinical or histological features was observed. This result is consistent with the role of the COL1A1 side of the fusion gene as a simple cis element for PDGFB overexpression (Greco et al., 1998; Simon et al., 2001). The spectacular clinical effect of tyrosine kinase receptor inhibitor (imatinib mesylate) also supports the major role of the PDGFR activation through the overexpression of PDGFB Abbreviations: COL1A1, collagen type I-a 1; DFSP, dermatofibrosarcoma protuberans; PDGFB, plateletderived growth factor-B chain

Variability of origin for the neocentromeric sequences in analphoid supernumerary marker chromosomes of well-differentiated liposarcomas

Well-differentiated liposarcomas (WDLPS) and dedifferentiated liposarcomas are cytogenetically characterized by the presence of supernumerary ring or giant chromosomes containing amplified material from the 12q14-15 region. These chromosomes contain neocentromeres, which are able to bind the kinetochore proteins and to ensure a stable mitotic transmission although they do not show detectable alpha-satellite sequences. WDLPS is the sole solid tumor for which the presence of a neocentromere is a consistent and specific feature. By immunostaining with anti-centromere antibodies in combination with FISH analysis (immunoFISH) in four cases of WDLPS, we have shown that sequences from the region 12q14-21 region were not located at the neocentromere site. In addition, we have microdissected the neocentromeric region from a giant supernumerary chromosome in the 94T778 WDLPS cell line. By using immunoFISH and positional cloning we have shown that the neocentromere of this cell line originated from a region at 4p16.1, rich in AT sequences and in long interspersed nucleotide element (LINE)1, that was co-amplified with 12q14-15. We have observed that this 4p sequence was not involved in the neocentromere of the supernumerary giant chromosome present in the 93T449 WDLPS cell line derived from a metachronous recurrence of the same primary WDLPS than 94T778. Altogether, these results indicate that the neocentromeres in WDLPS originate from amplified chromosomal regions other than 12q14-15 and do not involve a specific and recurrent DNA sequence. These sequences might be activated for centromeric function by epigenetic mechanisms.

Analysis of segmental duplications, mouse genome synteny and recurrent cancer-associated amplicons in human chromosome 6p21-p12.

It has been proposed that regions of microhomology in the human genome could facilitate genomic rearrangements, copy number transitions, and rapid genomic change during tumor progression. To investigate this idea, this study examines the role of repetitive sequence elements, and corresponding syntenic mouse genomic features, in targeting cancer-associated genomic instability of specific regions of the human genome. Automated database-mining algorithms designed to search for frequent copy number transitions and genomic breakpoints were applied to 2 publicly-available online databases and revealed that 6p21-p12 is one of the regions of the human genome most frequently involved in tumor-specific alterations. In these analyses, 6p21-p12 exhibited the highest frequency of genomic amplification in osteosarcomas. Analysis of repetitive elements in regions of homology between human chromosome 6p and the syntenic regions of the mouse genome revealed a strong association between the location of segmental duplications greater than 5 kilobase-pairs and the position of discontinuities at the end of the syntenic region. The presence of clusters of segmental duplications flanking these syntenic regions also correlated with a high frequency of amplification and genomic alteration. Collectively, the experimental findings, in silico analyses, and comparative genomic studies presented here suggest that segmental duplications may facilitate cancer-associated copy number transitions and rearrangements at chromosome 6p21-p12. This process may involve homology-dependent DNA recombination and/or repair, which may also contribute towards the overall plasticity of the human genome.

Exploring high dimensional data with Butterfly: a novel classification algorithm based on discrete dynamical systems

MOTIVATION We introduce a novel method for visualizing high dimensional data via a discrete dynamical system. This method provides a 2D representation of the relationship between subjects according to a set of variables without geometric projections, transformed axes or principal components. The algorithm exploits a memory-type mechanism inherent in a certain class of discrete dynamical systems collectively referred to as the chaos game that are closely related to iterative function systems. The goal of the algorithm was to create a human readable representation of high dimensional patient data that was capable of detecting unrevealed subclusters of patients from within anticipated classifications. This provides a mechanism to further pursue a more personalized exploration of pathology when used with medical data. For clustering and classification protocols, the dynamical system portion of the algorithm is designed to come after some feature selection filter and before some model evaluation (e.g. clustering accuracy) protocol. In the version given here, a univariate features selection step is performed (in practice more complex feature selection methods are used), a discrete dynamical system is driven by this reduced set of variables (which results in a set of 2D cluster models), these models are evaluated for their accuracy (according to a user-defined binary classification) and finally a visual representation of the top classification models are returned. Thus, in addition to the visualization component, this methodology can be used for both supervised and unsupervised machine learning as the top performing models are returned in the protocol we describe here. RESULTS Butterfly, the algorithm we introduce and provide working code for, uses a discrete dynamical system to classify high dimensional data and provide a 2D representation of the relationship between subjects. We report results on three datasets (two in the article; one in the appendix) including a public lung cancer dataset that comes along with the included Butterfly R package. In the included R script, a univariate feature selection method is used for the dimension reduction step, but in the future we wish to use a more powerful multivariate feature reduction method based on neural networks (Kriesel, 2007). AVAILABILITY AND IMPLEMENTATION A script written in R (designed to run on R studio) accompanies this article that implements this algorithm and is available at http://butterflygeraci.codeplex.com/. For details on the R package or for help installing the software refer to the accompanying document, Supporting Material and Appendix.

Cross-tolerance between osmotic and freeze-thaw stress in microbial assemblages from temperate lakes

Osmotic stress can accompany increases in solute concentrations because of freezing or high-salt environments. Consequently, microorganisms from environments with a high-osmotic potential may exhibit cross-tolerance to freeze stress. To test this hypothesis, enrichments derived from the sediment and water of temperate lakes with a range of salt concentrations were subjected to multiple freeze-thaw cycles. Surviving isolates were identified and metagenomes were sampled prior to and following selection. Enrichments from alkali lakes were typically the most freeze-thaw resistant with only 100-fold losses in cell viability, and those from freshwater lakes were most susceptible, with cell numbers reduced at least 100,000-fold. Metagenomic analysis suggested that selection reduced assemblage diversity more in freshwater samples than in those from saline lakes. Survivors included known psychro-, halo- and alkali-tolerant bacteria. Characterization of freeze-thaw-resistant isolates from brine and alkali lakes showed that few isolates had ice-associating activities such as antifreeze or ice nucleation properties. However, all brine- and alkali-derived isolates had high intracellular levels of osmolytes and/or appeared more likely to form biofilms. Conversely, these phenotypes were infrequent amongst the freshwater-derived isolates. These observations are consistent with microbial cross-tolerance between osmotic and freeze-thaw stresses.

Abstract 63: Incorporation of flanking probes reduces truncation losses for fluorescence in situ hybridization analysis of recurrent genomic deletions in tumor sections.

Proceedings: AACR 104th Annual Meeting 2013; Apr 6-10, 2013; Washington, DC Fluorescence in situ hybridization (FISH) is a robust technique when used with appropriate interpretative guidelines, and the assay can yield consistent results that provide diagnostically and prognostically useful information to help guide patient therapy. The establishment of quality control standards in clinical laboratories using routine FISH analysis of formalin fixed paraffin embedded (FFPE) sections in different tumor types for fusion and break-apart strategies and gene amplifications have been developed for both analysis and reporting, but at the present time there are no comprehensive guidelines for the analysis of genomic deletions using FFPE sections. The use of FISH on archival FFPE samples is technically demanding and becomes more challenging when applied to paraffin-embedded tissue microarrays. The evaluation of FISH signals in interphase nuclei of FFPE sections is affected by truncation and overlapping of the nuclei due to varying cell density, tissue architecture differences and histological forms. In this study we report a generalizable four-color deletion FISH approach to assist interpretation problems arising when evaluating FISH signals. The guidelines will help address interpretative dilemmas associated with overlapping and truncated nuclei in FFPE prostate cancer sections. The four-color FISH approach was developed using the PTEN tumor suppressor gene deletion model in prostate cancer. The PTEN assay was based on a robust bioinformatics analysis of 311 published human genome array datasets and comprises a centromeric “chromosome enumeration” probe, a specific PTEN gene probe, and control flanking probes either side of the target probe. The sensitivity and specificity parameters of the four-color PTEN probe set were further characterized using a large number of well-characterized tumors and stringent scoring criteria. The incorporation of flanking control probes allowed the analysts to determine if the chromosomal region was subject to truncation loss. A minimum threshold for apparent deletion frequency was set to address the heterogeneous and homogeneous nature of tumor histology. In addition the approach facilitated analysis of genotypic heterogeneity and varying clonality within different foci of tumor in the prostate. Overall the approach provided robust and highly reproducible results that minimized inter- and intra-assay variability. The four-color FISH deletion assay reduced the frequency of misinterpretation and improved both the quality and throughput of FISH analyses using clinical samples. Moreover the use of established controls and conservative cut-offs for assigning deletions will facilitate more coherent approach to developing reporting standards for deletion assays as more tumor suppressor genes of clinical importance are discovered by next generation sequencing methods. Citation Format: Maisa Yoshimoto, Olga Ludkovski, Jennifer Good, Robert J. Gooding, Jean McGowan-Jordan, Alexander Boag, Andrew Evans, Ming-Sound Tsao, Paulo Nuin, Jeremy A. Squire. Incorporation of flanking probes reduces truncation losses for fluorescence in situ hybridization analysis of recurrent genomic deletions in tumor sections. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 63. doi:10.1158/1538-7445.AM2013-63