Evolution plays a key role in Cancer as the result of the accumulation of genetic alterations, which provide selective advantages to a tumor cell, allowing resistance to anti-cancer drugs. Unfortunately, however, the identification of the driver mutations and thus the mechanisms underlying anti-cancer drug resistance (ACDR) still remains a challenge. We previously demonstrated that lentiviral vectors, when properly modified, might integrate near specific genes, alter their expression and induce cancer or ACDR in vivo and in vitro. The analysis of vector-cellular genomic junctions in tumor or ACDR cells allowed identifying causative genes of HER2+ breast cancer cell line using a statistical approach defined Common Insertion Sites (CISs) that highlight genomic regions targeted at significantly higher frequency than expected by a random distribution. The reconstruction of cumulative cancer progression from CISs genes has not been yet addressed and may produce causative gene networks. The aim of this project is studying anti-cancer drug resistance from exclusive and co-occurring genes using cumulative cancer progression from our cell line CISs genes and investigating the relation between them. Bioinformatics tools aimed at inferring cancer progression models, in terms of selective advantage relation among relevant genomic alteration from cross-sectional data (Next Generation Sequencing platforms), would allow identifying specific combinations of targeted drugs to overcome the occurrence of resistance. In a new context of vector Integration Sites (ISs), I developed an integrated bioinformatics workflow composed of: (i) an updated and more accurate version of VISPA (Vector Integration Site Parallel Analysis), a pipeline for automated ISs identification and annotation based on a distributed environment with a simple web based interface; (ii) identification of the CISs with a sliding window approach; (iii) a new statistical tool, CAncer PRogression Inference (CAPRI), to infer selective advantage relations among various mutational events in cancer cell genomes, mostly in relation with drug-resistance. The model is based on probabilistic causation and is able to reconstruct our cancer progression Direct Acyclic Graphs (DAGs), involving the CIS genes. With the use of GeneMANIA (http://www.genemania.org) and Enrichr (http://amp.pharm.mssm.edu/Enrichr), I studied the protein-protein interaction, Gene Ontology and Pathway relations between selected genes, collecting and visualizing results in gene networks. By applying our new method to the published ISs dataset from the two cell lines, I was able to generate progression models involving relevant genes (confirming that these are not mutually exclusive genes, by Mutex), which are consistent with previously validated results, confirming the role of PIK3CA-ERBB2 genes in ACDR. Unfortunately, one of the two cell line has a low quality samples. For this reason, CAPRI was not able to generate the progression DAG. I generated the progression DAG for the other cell line, BT474, pre-treatment and post-treatment with lapatinib respectively. The last step is to investigate the relations between genes, produced by the model, trying to find some useful new interactions and confirmations for ACDR studies (i.e. SUMO1-ERBB2-PIK3CA-CSMD3). New insertional mutagenesis data from lung cancer cell lines aimed to induce ACDR in vivo and in vitro are ongoing and will allow to validate and/or identify novel cancer progression models, as well as possible combinatorial therapies.

L'evoluzione nel cancro svolge un ruolo determinante attraverso l'accumulo di alterazioni genetiche, fornendo vantaggi selettivi ad una cellula tumorale, permettendo resistenza ai farmaci anti-cancro. Purtroppo, l'identificazione delle mutazioni driver e i meccanismi sottostanti la resistenza anti-tumorale ai farmaci (ACDR) rimangono ancora irrisolti. Abbiamo precedentemente dimostrato che vettori lentivirali, se adeguatamente modificati, possono integrare nei pressi di geni specifici, modificando la loro espressione e provocate il cancro o ACDR in vivo e in vitro. L'analisi delle integrazioni di questi vettori nei tumori o nelle cellule resistenti alle terapie, ha permesso di identificare geni responsabili di cancro al seno HER2+, su determinate linee cellulari, utilizzando un approccio statistico definito siti di inserzione comuni (CISs) che evidenziano regioni genomiche bersagliate a più alta frequenza di quanto previsto da una distribuzione normale. La ricostruzione della progressione del cancro accumulata dai geni (CISs) non è stata ancora affrontata in letteratura e può produrre reti causali di geni. Lo scopo di questo progetto è di studiare la resistenza anti-tumorale ai farmaci di geni esclusivi e concomitanti con la progressione cumulata del cancro rispetto ai CISs della linea cellulare e indagare il rapporto tra di loro. Strumenti bioinformatici volti a inferire modelli di progressione del cancro, in termini di relazioni di vantaggio selettivo tra alterazioni genomiche, da dati trasversali (piattaforme di sequenziamento di nuova generazione), consentirebbero l'identificazione di specifiche combinazioni di farmaci mirati per superare l'insorgenza della resistenza in fase chemioterapica. In un nuovo contesto di siti di integrazione di vettori lentivirali, ho sviluppato un flusso di lavoro bioinformatico integrato, composto da: (i) una versione aggiornata e più accurata di VISPA (Vector Integration Site Parallel Analysis), una pipeline per l'identificazione automatica dei ISs e la loro annotazione sulla base di un ambiente distribuito con una interfaccia web semplice ed intuitiva; (ii) l'identificazione dei CISs con un approccio a finestra scorrevole; (iii) un nuovo strumento statistico, CAncer PRogression Inference (CAPRI), per dedurre le relazioni di vantaggio selettivo tra i vari eventi mutazionali nei genomi di cellule tumorali, per lo più in relazione con la resistenza ad un farmaco. Il modello si basa sul nesso di causalità probabilistica ed è in grado di ricostruire la progressione del cancro attraverso Direct Acyclic Graphs (DAG), che coinvolge i geni risultanti CISs. Con l'uso di GeneMANIA (http://www.genemania.org) e Enrichr (http://amp.pharm.mssm.edu/Enrichr), ho studiato l'interazione proteina-proteina, Gene Ontology e Pathway tra geni selezionati, raccogliendo e visualizzando i risultati in reti di geni. Applicando il nuovo metodo ai dati di integrazione già pubblicati delle due linee cellulari, sono stato in grado di generare modelli di progressione che coinvolgono i geni rilevanti (confermando che questi non sono geni che si escludono a vicenda, con Mutex). Questi sono in linea con i risultati precedentemente validati, a conferma del ruolo dei geni PIK3CA-ErbB2 in ACDR. Purtroppo, una delle due linee di cellule ha campioni a bassa qualità. Per questo motivo, CAPRI non è stati in grado di generare la progressione. Ho generato la progressione per l'altra linea cellulare, BT474, pre-trattamento e post-trattamento con lapatinib. L'ultimo passo è stato di indagare le relazioni tra geni, prodotte dal modello, cercando di trovare alcune nuove interazioni utili e conferme per gli studi di ACDR (SUMO1-ERBB2-PIK3CA-CSMD3). Nuovi dati di mutagenesi inserzionale da linee cellulari di cancro al polmone, volti a indurre ACDR in vivo e in vitro, sono in corso e permetteranno di validare e/o identificare nuovi modelli di progressione del cancro, così come possibili terapie combinatorie.

(2017). Anti-Cancer Drug Resistance Causal Modeling from Lentiviral-Vector Integration Site Studies. (Tesi di dottorato, Università degli Studi di Milano-Bicocca, 2017).

Anti-Cancer Drug Resistance Causal Modeling from Lentiviral-Vector Integration Site Studies

SPINOZZI, GIULIO
2017

Abstract

Evolution plays a key role in Cancer as the result of the accumulation of genetic alterations, which provide selective advantages to a tumor cell, allowing resistance to anti-cancer drugs. Unfortunately, however, the identification of the driver mutations and thus the mechanisms underlying anti-cancer drug resistance (ACDR) still remains a challenge. We previously demonstrated that lentiviral vectors, when properly modified, might integrate near specific genes, alter their expression and induce cancer or ACDR in vivo and in vitro. The analysis of vector-cellular genomic junctions in tumor or ACDR cells allowed identifying causative genes of HER2+ breast cancer cell line using a statistical approach defined Common Insertion Sites (CISs) that highlight genomic regions targeted at significantly higher frequency than expected by a random distribution. The reconstruction of cumulative cancer progression from CISs genes has not been yet addressed and may produce causative gene networks. The aim of this project is studying anti-cancer drug resistance from exclusive and co-occurring genes using cumulative cancer progression from our cell line CISs genes and investigating the relation between them. Bioinformatics tools aimed at inferring cancer progression models, in terms of selective advantage relation among relevant genomic alteration from cross-sectional data (Next Generation Sequencing platforms), would allow identifying specific combinations of targeted drugs to overcome the occurrence of resistance. In a new context of vector Integration Sites (ISs), I developed an integrated bioinformatics workflow composed of: (i) an updated and more accurate version of VISPA (Vector Integration Site Parallel Analysis), a pipeline for automated ISs identification and annotation based on a distributed environment with a simple web based interface; (ii) identification of the CISs with a sliding window approach; (iii) a new statistical tool, CAncer PRogression Inference (CAPRI), to infer selective advantage relations among various mutational events in cancer cell genomes, mostly in relation with drug-resistance. The model is based on probabilistic causation and is able to reconstruct our cancer progression Direct Acyclic Graphs (DAGs), involving the CIS genes. With the use of GeneMANIA (http://www.genemania.org) and Enrichr (http://amp.pharm.mssm.edu/Enrichr), I studied the protein-protein interaction, Gene Ontology and Pathway relations between selected genes, collecting and visualizing results in gene networks. By applying our new method to the published ISs dataset from the two cell lines, I was able to generate progression models involving relevant genes (confirming that these are not mutually exclusive genes, by Mutex), which are consistent with previously validated results, confirming the role of PIK3CA-ERBB2 genes in ACDR. Unfortunately, one of the two cell line has a low quality samples. For this reason, CAPRI was not able to generate the progression DAG. I generated the progression DAG for the other cell line, BT474, pre-treatment and post-treatment with lapatinib respectively. The last step is to investigate the relations between genes, produced by the model, trying to find some useful new interactions and confirmations for ACDR studies (i.e. SUMO1-ERBB2-PIK3CA-CSMD3). New insertional mutagenesis data from lung cancer cell lines aimed to induce ACDR in vivo and in vitro are ongoing and will allow to validate and/or identify novel cancer progression models, as well as possible combinatorial therapies.
MAURI, GIANCARLO
ANTONIOTTI, MARCO
MONTINI, EUGENIO
causal; modeling,; gene; therapy,; IS
causal; modeling,; gene; therapy,; IS
INF/01 - INFORMATICA
English
27-mar-2017
INFORMATICA - 87R
29
2015/2016
open
(2017). Anti-Cancer Drug Resistance Causal Modeling from Lentiviral-Vector Integration Site Studies. (Tesi di dottorato, Università degli Studi di Milano-Bicocca, 2017).
File in questo prodotto:
File Dimensione Formato  
phd_unimib_787765.pdf

accesso aperto

Descrizione: tesi di dottorato
Tipologia di allegato: Doctoral thesis
Dimensione 13.9 MB
Formato Adobe PDF
13.9 MB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/10281/151647
Citazioni
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
Social impact