Identifying and Targeting Molecular Deregulations in Uterine Leiomyosarcoma: A Role for mTOR Inhibition-Based Combination Therapies

Uterine leiomyosarcoma (ULMS) is an aggressive malignancy characterized by marked chemoresistance, frequent relapses, and poor outcome. Despite efforts to improve survival over the past several decades, only minimal advances have been made. Hence, there is an urgent and unmet need for better understanding of the molecular deregulations that underlay ULMS and development of more effective therapeutic strategies. This work identified several common deregulations in a large (n=208) tissue microarray of ULMS compared to GI smooth muscle, myometrium, and leiomyoma controls. Our results suggest that significant loss of smooth muscle and gynecological differentiation markers is common in ULMS, a finding that could help render improved ULMS diagnosis, especially for advanced disease. Similarly to reports in other malignancies, we found that several cancer-related proteins were differentially expressed; these could be useful together as biomarkers for ULMS. Notably, we identified significant upregulation and overexpression of the mTOR pathway in ULMS, examined the possible contribution of tyrosine kinase receptor deregulation promoting mTOR activation, and unraveled a role for pS6RP and p4EBP1 as molecular disease prognosticators. The significance of mTOR activation in ULMS and its potential as a therapeutic target were further investigated. Rapamycin abrogated ULMS cell growth and cell cycle progression in vitro but induced only sight growth delay in vivo. Given that effective mTOR therapies likely require combination mTOR blockade with inhibition of other targets, coupled with recent observations suggesting that Aurora A kinase (Aurk A) deregulations commonly occur in ULMS, the preclinical impact of dually targeting both pathways was evaluated. Combined therapy with rapamycin (an mTORC1 inhibitor) and MLN8237 (an investigational Aurk A inhibitor) profoundly and synergistically abrogated ULMS growth in vitro. Interestingly, the superior effects were noted only when MLN8237 was pre-administered. This novel therapeutic combination and scheduling regimen resulted in marked tumor growth inhibition in vivo. Together, these data support further exploration of dual mTOR and Aurk A blockade for the treatment of human ULMS.

INTEGRATIVE BIOMARKER IDENTIFICATION AND CLASSIFICATION USING HIGH THROUGHPUT ASSAYS

It is well accepted that tumorigenesis is a multi-step procedure involving aberrant functioning of genes regulating cell proliferation, differentiation, apoptosis, genome stability, angiogenesis and motility. To obtain a full understanding of tumorigenesis, it is necessary to collect information on all aspects of cell activity. Recent advances in high throughput technologies allow biologists to generate massive amounts of data, more than might have been imagined decades ago. These advances have made it possible to launch comprehensive projects such as (TCGA) and (ICGC) which systematically characterize the molecular fingerprints of cancer cells using gene expression, methylation, copy number, microRNA and SNP microarrays as well as next generation sequencing assays interrogating somatic mutation, insertion, deletion, translocation and structural rearrangements. Given the massive amount of data, a major challenge is to integrate information from multiple sources and formulate testable hypotheses. This thesis focuses on developing methodologies for integrative analyses of genomic assays profiled on the same set of samples. We have developed several novel methods for integrative biomarker identification and cancer classification. We introduce a regression-based approach to identify biomarkers predictive to therapy response or survival by integrating multiple assays including gene expression, methylation and copy number data through penalized regression. To identify key cancer-specific genes accounting for multiple mechanisms of regulation, we have developed the integIRTy software that provides robust and reliable inferences about gene alteration by automatically adjusting for sample heterogeneity as well as technical artifacts using Item Response Theory. To cope with the increasing need for accurate cancer diagnosis and individualized therapy, we have developed a robust and powerful algorithm called SIBER to systematically identify bimodally expressed genes using next generation RNAseq data. We have shown that prediction models built from these bimodal genes have the same accuracy as models built from all genes. Further, prediction models with dichotomized gene expression measurements based on their bimodal shapes still perform well. The effectiveness of outcome prediction using discretized signals paves the road for more accurate and interpretable cancer classification by integrating signals from multiple sources.

Nonlinear dimension reduction in pathway-based genome-wide association analysis

Pathway based genome wide association study evolves from pathway analysis for microarray gene expression and is under rapid development as a complementary for single-SNP based genome wide association study. However, it faces new challenges, such as the summarization of SNP statistics to pathway statistics. The current study applies the ridge regularized Kernel Sliced Inverse Regression (KSIR) to achieve dimension reduction and compared this method to the other two widely used methods, the minimal-p-value (minP) approach of assigning the best test statistics of all SNPs in each pathway as the statistics of the pathway and the principal component analysis (PCA) method of utilizing PCA to calculate the principal components of each pathway. Comparison of the three methods using simulated datasets consisting of 500 cases, 500 controls and100 SNPs demonstrated that KSIR method outperformed the other two methods in terms of causal pathway ranking and the statistical power. PCA method showed similar performance as the minP method. KSIR method also showed a better performance over the other two methods in analyzing a real dataset, the WTCCC Ulcerative Colitis dataset consisting of 1762 cases, 3773 controls as the discovery cohort and 591 cases, 1639 controls as the replication cohort. Several immune and non-immune pathways relevant to ulcerative colitis were identified by these methods. Results from the current study provided a reference for further methodology development and identified novel pathways that may be of importance to the development of ulcerative colitis.^

Network-based analysis of genome-wide scans of natural selection

The genomic era brought by recent advances in the next-generation sequencing technology makes the genome-wide scans of natural selection a reality. Currently, almost all the statistical tests and analytical methods for identifying genes under selection was performed on the individual gene basis. Although these methods have the power of identifying gene subject to strong selection, they have limited power in discovering genes targeted by moderate or weak selection forces, which are crucial for understanding the molecular mechanisms of complex phenotypes and diseases. Recent availability and rapid completeness of many gene network and protein-protein interaction databases accompanying the genomic era open the avenues of exploring the possibility of enhancing the power of discovering genes under natural selection. The aim of the thesis is to explore and develop normal mixture model based methods for leveraging gene network information to enhance the power of natural selection target gene discovery. The results show that the developed statistical method, which combines the posterior log odds of the standard normal mixture model and the Guilt-By-Association score of the gene network in a naïve Bayes framework, has the power to discover moderate/weak selection gene which bridges the genes under strong selection and it helps our understanding the biology under complex diseases and related natural selection phenotypes.^

A Genomic Approach to Identify The Notch Pathway as a Putative Tumor Suppressor in Endometrial Cancer

Endometrial cancer is the most common gynecological malignancy and the fourth most frequently diagnosed cancer among women. The molecular changes that distinguish normal endometrium from endometrial carcinoma are not thoroughly understood. Identification of these changes could potentially aid in identifying at-risk women who are especially prone to develop endometrial cancer, such as obese women and women with Lynch Syndrome. A microarray analysis was performed using normal endometrium from thin and obese women and cancerous endometrium from obese women. We validated the differential expression of ten genes whose expression was significantly up-regulated or down-regulated using qRT-PCR. All of the genes had distinct expression levels depending on the endometrial carcinoma histotype. As a result, they could serve as molecular markers to distinguish between normal endometrium and endometrial cancer, as well as between low grade endometrial carcinomas and high grade endometrial carcinomas. Two of the ten genes validated, HEYL and HES1, are down-stream targets of the Notch signaling pathway. HEYL and HES1 were identified by microarray and qRT-PCR to have a significant decrease in expression in endometrial carcinomas compared to normal endometrium. We further analyzed the differential expression of other components of the Notch signaling pathway, Notch4 and Jagged1. They were also identified by qRT-PCR to be significantly down-regulated in endometrial carcinomas compared to normal endometrium. Therefore, we believe the Notch signaling pathway to act as a tumor suppressor in endometrial carcinomas.

Two way clustering of Microarray Data using a Hybrid Approach

The Microarray technique is rather powerful, as it allows to test up thousands of genes at a time, but this produces an overwhelming set of data files containing huge amounts of data, which is quite difficult to pre-process, separate, classify and correlate for interesting conclusions to be extracted. Modern machine learning, data mining and clustering techniques based on information theory, are needed to read and interpret the information contents buried in those large data sets. Independent Component Analysis method can be used to correct the data affected by corruption processes or to filter the uncorrectable one and then clustering methods can group similar genes or classify samples. In this paper a hybrid approach is used to obtain a two way unsupervised clustering for a corrected microarray data.

Mouse p53-deficient cancer models as platforms for obtaining genomic predictors of human cancer clinical outcomes

Mutations in the TP53 gene are very common in human cancers, and are associated with poor clinical outcome. Transgenic mouse models lacking the Trp53 gene or that express mutant Trp53 transgenes produce tumours with malignant features in many organs. We previously showed the transcriptome of a p53-deficient mouse skin carcinoma model to be similar to those of human cancers with TP53 mutations and associated with poor clinical outcomes. This report shows that much of the 682-gene signature of this murine skin carcinoma transcriptome is also present in breast and lung cancer mouse models in which p53 is inhibited. Further, we report validated gene-expression-based tests for predicting the clinical outcome of human breast and lung adenocarcinoma. It was found that human patients with cancer could be stratified based on the similarity of their transcriptome with the mouse skin carcinoma 682-gene signature. The results also provide new targets for the treatment of p53-defective tumours.

Identifying the semantics of eligibility criteria of clinical trials based on relevant medical ontologies

An important objective of the INTEGRATE project1 is to build tools that support the efficient execution of post-genomic multi-centric clinical trials in breast cancer, which includes the automatic assessment of the eligibility of patients for available trials. The population suited to be enrolled in a trial is described by a set of free-text eligibility criteria that are both syntactically and semantically complex. At the same time, the assessment of the eligibility of a patient for a trial requires the (machineprocessable) understanding of the semantics of the eligibility criteria in order to further evaluate if the patient data available for example in the hospital EHR satisfies these criteria. This paper presents an analysis of the semantics of the clinical trial eligibility criteria based on relevant medical ontologies in the clinical research domain: SNOMED-CT, LOINC, MedDRA. We detect subsets of these widely-adopted ontologies that characterize the semantics of the eligibility criteria of trials in various clinical domains and compare these sets. Next, we evaluate the occurrence frequency of the concepts in the concrete case of breast cancer (which is our first application domain) in order to provide meaningful priorities for the task of binding/mapping these ontology concepts to the actual patient data. We further assess the effort required to extend our approach to new domains in terms of additional semantic mappings that need to be developed.

Is Microarray Analysis Really Useful and Sufficient to Diagnose Nut Allergy in the Mediterranean Area?

Background: Component-based diagnosis on multiplex platforms is widely used in food allergy but its clinical performance has not been evaluated in nut allergy. Objective: To assess the diagnostic performance of a commercial protein microarray in the determination of specific IgE (sIgE) in peanut, hazelnut, and walnut allergy. Methods: sIgE was measured in 36 peanut-allergic, 36 hazelnut-allergic, and 44 walnut-allergic patients by ISAC 112, and subsequently, sIgE against available components was determined by ImmunoCAP in patients with negative ISAC results. ImmunoCAP was also used to measure sIgE to Ara h 9, Cor a 8, and Jug r 3 in a subgroup of lipid transfer protein (LTP)-sensitized nut-allergic patients (positive skin prick test to LTP-enriched extract). sIgE levels by ImmunoCAP were compared with ISAC ranges. Results: Most peanut-, hazelnut-, and walnut-allergic patients were sensitized to the corresponding nut LTP (Ara h 9, 66.7%; Cor a 8, 80.5%; Jug r 3, 84% respectively). However, ISAC did not detect sIgE in 33.3% of peanut-allergic patients, 13.9% of hazelnut-allergic patients, or 13.6% of walnut-allergic patients. sIgE determination by ImmunoCAP detected sensitization to Ara h 9, Cor a 8, and Jug r 3 in, respectively, 61.5% of peanut-allergic patients, 60% of hazelnut-allergic patients, and 88.3% of walnut-allergic patients with negative ISAC results. In the subgroup of peach LTP?sensitized patients, Ara h 9 sIgE was detected in more cases by ImmunoCAP than by ISAC (94.4% vs 72.2%, P<.05). Similar rates of Cor a 8 and Jug r 3 sensitization were detected by both techniques. Conclusions: The diagnostic performance of ISAC was adequate for hazelnut and walnut allergy but not for peanut allergy. sIgE sensitivity against Ara h 9 in ISAC needs to be improved.

Terpenoid-based defenses in conifers: cDNA cloning, characterization, and functional expression of wound-inducible (E)-α-bisabolene synthase from grand fir (Abies grandis)

(E)-α-Bisabolene synthase is one of two wound-inducible sesquiterpene synthases of grand fir (Abies grandis), and the olefin product of this cyclization reaction is considered to be the precursor in Abies species of todomatuic acid, juvabione, and related insect juvenile hormone mimics. A cDNA encoding (E)-α-bisabolene synthase was isolated from a wound-induced grand fir stem library by a PCR-based strategy and was functionally expressed in Escherichia coli and shown to produce (E)-α-bisabolene as the sole product from farnesyl diphosphate. The expressed synthase has a deduced size of 93.8 kDa and a pI of 5.03, exhibits other properties typical of sesquiterpene synthases, and resembles in sequence other terpenoid synthases with the exception of a large amino-terminal insertion corresponding to Pro81–Val296. Biosynthetically prepared (E)-α-[3H]bisabolene was converted to todomatuic acid in induced grand fir cells, and the time course of appearance of bisabolene synthase mRNA was shown by Northern hybridization to lag behind that of mRNAs responsible for production of induced oleoresin monoterpenes. These results suggest that induced (E)-α-bisabolene biosynthesis constitutes part of a defense response targeted to insect herbivores, and possibly fungal pathogens, that is distinct from induced oleoresin monoterpene production.

Octamer-based genome scanning distinguishes a unique subpopulation of Escherichia coli O157:H7 strains in cattle

Multilocus-genotyping methods have shown that Escherichia coli O157:H7 is a geographically disseminated clone. However, high-resolution methods such as pulse-field gel electrophoresis demonstrate significant genomic diversity among different isolates. To assess the genetic relationship of human and bovine isolates of E. coli O157:H7 in detail, we have developed an octamer-based genome-scanning methodology, which compares the distance between over-represented, strand-biased octamers that occur in the genome. Comparison of octamer-based genome-scanning products derived from >1 megabase of the genome demonstrated the existence of two distinct lineages of E. coli O157:H7 that are disseminated within the United States. Human and bovine isolates are nonrandomly distributed among the lineages, suggesting that one of these lineages may be less virulent for humans or may not be efficiently transmitted to humans from bovine sources. Restriction fragment length polymorphism analysis with lambdoid phage genomes indicates that phage-mediated events are associated with divergence of the lineages, thereby providing one explanation for the degree of diversity that is observed among E. coli O157:H7 by other molecular-fingerprinting methods.

Detection of HIV-1 RNA by nucleic acid sequence-based amplification combined with fluorescence correlation spectroscopy

Nucleic acid sequence-based amplification (NASBA) has proved to be an ultrasensitive method for HIV-1 diagnosis in plasma even in the primary HIV infection stage. This technique was combined with fluorescence correlation spectroscopy (FCS) which enables online detection of the HIV-1 RNA molecules amplified by NASBA. A fluorescently labeled DNA probe at nanomolar concentration was introduced into the NASBA reaction mixture and hybridizing to a distinct sequence of the amplified RNA molecule. The specific hybridization and extension of this probe during amplification reaction, resulting in an increase of its diffusion time, was monitored online by FCS. As a consequence, after having reached a critical concentration of 0.1–1 nM (threshold for unaided FCS detection), the number of amplified RNA molecules in the further course of reaction could be determined. Evaluation of the hybridization/extension kinetics allowed an estimation of the initial HIV-1 RNA concentration that was present at the beginning of amplification. The value of initial HIV-1 RNA number enables discrimination between positive and false-positive samples (caused for instance by carryover contamination)—this possibility of discrimination is an essential necessity for all diagnostic methods using amplification systems (PCR as well as NASBA). Quantitation of HIV-1 RNA in plasma by combination of NASBA with FCS may also be useful in assessing the efficacy of anti-HIV agents, especially in the early infection stage when standard ELISA antibody tests often display negative results.

Quantifying single gene copy number by measuring fluorescent probe lengths on combed genomic DNA

An approach was developed for the quantification of subtle gains and losses of genomic DNA. The approach relies on a process called molecular combing. Molecular combing consists of the extension and alignment of purified molecules of genomic DNA on a glass coverslip. It has the advantage that a large number of genomes can be combed per coverslip, which allows for a statistically adequate number of measurements to be made on the combed DNA. Consequently, a high-resolution approach to mapping and quantifying genomic alterations is possible. The approach consists of applying fluorescence hybridization to the combed DNA by using probes to identify the amplified region. Measurements then are made on the linear hybridization signals to ascertain the region's exact size. The reliability of the approach first was tested for low copy number amplifications by determining the copy number of chromosome 21 in a normal and trisomy 21 cell line. It then was tested for high copy number amplifications by quantifying the copy number of an oncogene amplified in the tumor cell line GTL-16. These results demonstrate that a wide range of amplifications can be accurately and reliably quantified. The sensitivity and resolution of the approach likewise was assessed by determining the copy number of a single allele (160 kb) alteration.

t(11;22)(q23;q11.2) in acute myeloid leukemia of infant twins fuses MLL with hCDCrel, a cell division cycle gene in the genomic region of deletion in DiGeorge and velocardiofacial syndromes

We examined the MLL genomic translocation breakpoint in acute myeloid leukemia of infant twins. Southern blot analysis in both cases showed two identical MLL gene rearrangements indicating chromosomal translocation. The rearrangements were detectable in the second twin before signs of clinical disease and the intensity relative to the normal fragment indicated that the translocation was not constitutional. Fluorescence in situ hybridization with an MLL-specific probe and karyotype analyses suggested t(11;22)(q23;q11.2) disrupting MLL. Known 5′ sequence from MLL but unknown 3′ sequence from chromosome band 22q11.2 formed the breakpoint junction on the der(11) chromosome. We used panhandle variant PCR to clone the translocation breakpoint. By ligating a single-stranded oligonucleotide that was homologous to known 5′ MLL genomic sequence to the 5′ ends of BamHI-digested DNA through a bridging oligonucleotide, we formed the stem–loop template for panhandle variant PCR which yielded products of 3.9 kb. The MLL genomic breakpoint was in intron 7. The sequence of the partner DNA from band 22q11.2 was identical to the hCDCrel (human cell division cycle related) gene that maps to the region commonly deleted in DiGeorge and velocardiofacial syndromes. Both MLL and hCDCrel contained homologous CT, TTTGTG, and GAA sequences within a few base pairs of their respective breakpoints, which may have been important in uniting these two genes by translocation. Reverse transcriptase-PCR amplified an in-frame fusion of MLL exon 7 to hCDCrel exon 3, indicating that an MLL-hCDCrel chimeric mRNA had been transcribed. Panhandle variant PCR is a powerful strategy for cloning translocation breakpoints where the partner gene is undetermined. This application of the method identified a region of chromosome band 22q11.2 involved in both leukemia and a constitutional disorder.