Protein dynamics, structure and function are intimately linked in many protein systems (Karplus et al. 2005, Philos. Trans. A Math. Phys. Eng. Sci., Bahar et al. 2010 Annu. Rev. Biophys.). Thus, the characterization of the protein dynamics in atomistic details can be important to understand functional aspects of these fundamental cellular components. The study of dynamics-structure-function relationship in proteins may also have impact well-beyond fundamental research for biotechnological or pharmacological applications (Ozbabacan et al. 2010 Curr. Opin. Drug Discov. Devel.). In this Ph.D. project, we employed all-atom and explicit solvent Molecular Dynamics (MD) simulations to study protein dynamics of several target proteins that are related to human diseases. MD simulations have the potential to describe motions occurring on different timescales and to characterize protein structural ensembles at the atom-level if integrated to more accurate sampling techniques or to experimental data (Dror et al. 2012 Annu. Rev. Biophys., Sutto et al. 2012 Wiley Interdiscip. Rev. Comput. Mol. Sci.). Indeed, we here integrated our simulations with experimental biophysical data to overcome intrinsic limitations in MD simulations due to force field accuracy (the physical model used to describe the protein and the environment in our simulations) and conformational sampling. The analysis of our MD simulations have been also accompanied by methods inspired by graph theory to describe structural communication occurring between distal sites in the protein during dynamics (Papaleo et al. 2012 PLoS One, Ghosh et al. 2007 Proc. Natl. Acad. Sci. USA). In particular, we focused in our studies on target proteins that belong to the class of intrinsically disordered proteins (IDPs) and transcription factors, which are both often associated with cancer or neurodegenerative diseases. As examples of IDPs we selected the C-terminal domain of yeast Sic1 (Brocca et al. 2011, Biophys. J.) and the disordered regions of human Ataxin-3 (Saunders et al. 2009 Protein Eng. Des. Sel.) to provide a description of their heterogeneous ensemble in solution and to identify structures that resemble the conformations bound to their biological partners (Lambrughi et al. 2012 Front. Physiol.; Invernizzi et al. 2013 BBA Gen. Subj.). To achieve this goal, we integrated MD simulations with biophysical spectroscopies, and especially NMR thanks to a visiting period at the Structural Biology and NMR Laboratory at the University of Copenhagen (Dk) in the group of Prof. Kaare Teilum, under the supervision of Dr. Gaetano Invernizzi. As an example of transcription factor, we focused on p53 and zinc-finger domains to study the effects induced upon DNA-binding and the structural alterations associated with the replacement of zinc with other non-essential metal ions, as cadmium a known toxic metal and human carcinogen (IARC Monogr. Eval. Carcinog. Risks. Hum. 1993). Since no sufficiently accurate parameters are available to describe metal ions in classical MD force field, we have also developed a protocol based on classical molecular mechanics and quantum chemical calculations to derive optimized parameters for zinc and cadmium, which can then be used in MD simulations.

Le dinamiche, la struttura e la funzione sono intimamente legate nei sistemi proteici (Karplus et al. 2005, Philos. Trans. A Math. Phys. Eng. Sci., Bahar et al. 2010 Annu. Rev. Biophys.) e la caratterizzazione a livello atomico delle dinamiche delle proteine è essenziale per comprendere i loro aspetti funzionali. Lo studio della relazione dinamica-struttura-funzione nelle proteine può anche avere un impatto ben oltre la ricerca di base per lo sviluppo di applicazioni biotecnologiche e farmacologiche (Ozbabacan et al. 2010 Curr. Opin. Drug Discov. Devel.). In questo progetto di dottorato, abbiamo impiegato simulazioni di Dinamica Molecolare (MD) in solvente esplicito per studiare le dinamiche di diverse proteine bersaglio associate allo sviluppo di malattie umane. Le simulazioni MD hanno il potenziale per descrivere efficacemente i moti proteici che avvengono su diverse scale temporali e per caratterizzare ensemble strutturali a livello atomico, specialmente se integrate da tecniche di campionamento più accurate o da dati sperimentali (Dror et al. 2012 Annu. Rev. Biophys., Sutto et al. 2012 Wiley Interdiscip. Rev. Comput. Mol. Sci.). In particolare abbiamo integrato le nostre simulazioni con dati biofisici sperimentali in maniera da superare le limitazioni intrinseche nelle simulazioni MD associate all’accuratezza dei forcefield (il modello fisico utilizzato per descrivere la proteina e l'ambiente) e al campionamento conformazionale. Nell'analisi delle simulazioni MD sono stati utilizzati anche metodi ispirati alla teoria dei grafi per descrivere la comunicazione strutturale che occorre tra siti distali della proteina durante la dinamica (Papaleo et al. 2012 PLoS One, Ghosh et al. 2007 Proc. Natl. Acad. Sci. USA). In particolare, nei nostri studi ci siamo concentrati su proteine che appartengono alla classe delle proteine intrinsecamente disordinate (IDP) e fattori di trascrizione, che sono entrambi frequentemente associate con lo sviluppo di cancro o malattie neurodegenerative. Come esempi di IDP abbiamo scelto il dominio C-terminale di Sic1 (Brocca et al. 2011, Biophys. J.) e le regioni disordinate dell’Atassina-3 umana (Saunders et al., 2009 Protein Ing. Des. Sel.) per fornire una descrizione della loro eterogeneità strutturale in soluzione e identificare le variazioni conformazionali associate all’interazione con partner biologici (Lambrughi et al. 2012 Front. Physiol.; Invernizzi et al. 2013 BBA Gen. Subj.). Per raggiungere questo obiettivo, abbiamo integrato simulazioni MD con spettroscopie biofisiche, come la spettroscopia NMR grazie ad un periodo di ricerca presso lo Structural Biology and NMR Laboratory presso l'Università di Copenhagen (DK) nel gruppo del Prof. Kaare Teilum, sotto la supervisione del Dr. Gaetano Invernizzi. Come esempi di fattori di trascrizioni ci siamo concentrati sull’oncosoppressore p53 e su domini zinc-finger per studiare gli effetti indotti dall’interazione con il DNA e le alterazioni strutturali associate alla sostituzione dello zinco con altri ioni metallici non essenziali come il cadmio: un metallo tossico e noto agente cancerogeno nell'uomo (IARC Monogr. Eval. Carcinog. Risks. Hum. 1993). Poiché nei forcefield attualmente esistenti non sono disponibili parametri sufficientemente precisi per descrivere efficacemente la coordinazione con ioni metallici nelle proteine, abbiamo sviluppato un protocollo basato sulla meccanica molecolare classica e calcoli chimici quantistici per derivare nuovi parametri ottimizzati per lo zinco e il cadmio, che possono poi essere utilizzati in simulazioni MD.

(2015). Network analysis and molecular dynamics simulations to investigate the link between structure and function in intrinsically disordered proteins and transcription factors.. (Tesi di dottorato, Università degli Studi di Milano-Bicocca, 2015).

Network analysis and molecular dynamics simulations to investigate the link between structure and function in intrinsically disordered proteins and transcription factors.

LAMBRUGHI, MATTEO
2015

Abstract

Protein dynamics, structure and function are intimately linked in many protein systems (Karplus et al. 2005, Philos. Trans. A Math. Phys. Eng. Sci., Bahar et al. 2010 Annu. Rev. Biophys.). Thus, the characterization of the protein dynamics in atomistic details can be important to understand functional aspects of these fundamental cellular components. The study of dynamics-structure-function relationship in proteins may also have impact well-beyond fundamental research for biotechnological or pharmacological applications (Ozbabacan et al. 2010 Curr. Opin. Drug Discov. Devel.). In this Ph.D. project, we employed all-atom and explicit solvent Molecular Dynamics (MD) simulations to study protein dynamics of several target proteins that are related to human diseases. MD simulations have the potential to describe motions occurring on different timescales and to characterize protein structural ensembles at the atom-level if integrated to more accurate sampling techniques or to experimental data (Dror et al. 2012 Annu. Rev. Biophys., Sutto et al. 2012 Wiley Interdiscip. Rev. Comput. Mol. Sci.). Indeed, we here integrated our simulations with experimental biophysical data to overcome intrinsic limitations in MD simulations due to force field accuracy (the physical model used to describe the protein and the environment in our simulations) and conformational sampling. The analysis of our MD simulations have been also accompanied by methods inspired by graph theory to describe structural communication occurring between distal sites in the protein during dynamics (Papaleo et al. 2012 PLoS One, Ghosh et al. 2007 Proc. Natl. Acad. Sci. USA). In particular, we focused in our studies on target proteins that belong to the class of intrinsically disordered proteins (IDPs) and transcription factors, which are both often associated with cancer or neurodegenerative diseases. As examples of IDPs we selected the C-terminal domain of yeast Sic1 (Brocca et al. 2011, Biophys. J.) and the disordered regions of human Ataxin-3 (Saunders et al. 2009 Protein Eng. Des. Sel.) to provide a description of their heterogeneous ensemble in solution and to identify structures that resemble the conformations bound to their biological partners (Lambrughi et al. 2012 Front. Physiol.; Invernizzi et al. 2013 BBA Gen. Subj.). To achieve this goal, we integrated MD simulations with biophysical spectroscopies, and especially NMR thanks to a visiting period at the Structural Biology and NMR Laboratory at the University of Copenhagen (Dk) in the group of Prof. Kaare Teilum, under the supervision of Dr. Gaetano Invernizzi. As an example of transcription factor, we focused on p53 and zinc-finger domains to study the effects induced upon DNA-binding and the structural alterations associated with the replacement of zinc with other non-essential metal ions, as cadmium a known toxic metal and human carcinogen (IARC Monogr. Eval. Carcinog. Risks. Hum. 1993). Since no sufficiently accurate parameters are available to describe metal ions in classical MD force field, we have also developed a protocol based on classical molecular mechanics and quantum chemical calculations to derive optimized parameters for zinc and cadmium, which can then be used in MD simulations.
DE GIOIA, LUCA
Protein, Dynamics, Molecular Dynamics, Nuclear Magnetic Resonance, Intrinsically Disordered Protein, Transcription Factors, Sic1, p53, metals, Ataxin-3, zinc-finger, zinc, cadmium
CHIM/03 - CHIMICA GENERALE E INORGANICA
English
13-feb-2015
Scuola di dottorato di Scienze
BIOTECNOLOGIE INDUSTRIALI - 15R
27
2013/2014
open
(2015). Network analysis and molecular dynamics simulations to investigate the link between structure and function in intrinsically disordered proteins and transcription factors.. (Tesi di dottorato, Università degli Studi di Milano-Bicocca, 2015).
File in questo prodotto:
File Dimensione Formato  
PhD_unimib_075794.pdf

accesso aperto

Descrizione: Tesi di dottorato
Tipologia di allegato: Doctoral thesis
Dimensione 8.49 MB
Formato Adobe PDF
8.49 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/10281/71187
Citazioni
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
Social impact