Dissecting biomolecular interactions by integrative modeling
Başlık çevirisi mevcut değil.
- Tez No: 402654
- Danışmanlar: PROF. DR. ALEXANDRE M. J. J. BONVIN
- Tez Türü: Doktora
- Konular: Biyomühendislik, Mikrobiyoloji, Bioengineering, Microbiology
- Anahtar Kelimeler: Belirtilmemiş.
- Yıl: 2013
- Dil: İngilizce
- Üniversite: Universiteit Utrecht
- Enstitü: Yurtdışı Enstitü
- Ana Bilim Dalı: Belirtilmemiş.
- Bilim Dalı: Belirtilmemiş.
- Sayfa Sayısı: 141
Özet
Özet yok.
Özet (Çeviri)
The process of life, which we see, feel, experience or witness in our daily lives, is essentially modulated at the nanoscale (10-9 m). Being more precise, the biological function we observe at the macroscale, is an outcome of the orchestrated communications among biomolecules -particularly proteins- at the nanoscale. Thus, it is by dissecting and grasping the nanoscale world of biomolecules and their interaction networks that we will gain a basic understanding on biological processes. Since the early 50s, this realization has led to the birth and rise of“structural biology”, the science of elucidating structures of biomolecules at atomic scale and relating them to their functions. Classical structural biology is mainly defined by two experimental techniques, X-Ray crystallography and Nuclear Magnetic Resonance (NMR) spectroscopy, through which atomic structures of tens of thousands of biomolecules, mainly proteins, and thousands of biomolecular complexes have been solved in an accurate manner. These methods have helped immensely in the discovery of the structural world of many biomolecular complexes. Still, as a field, structural biology is far from keeping pace with the speed at which biological data are being generated in other disciplines, such as biochemistry. As an example, the current number of known atomic structures of macromolecular complexes is depicted to be considerably smaller than the documented protein-protein interactions. Unfortunately, technical limitations of classical structural biology techniques and/or the nature of the biomolecules under study hamper to close this gap in a rapid manner. As a rescue strategy, structural biologists often resort to different types of biochemical and biophysical experiments that can quickly provide low-resolution information on macromolecular complexes, even for challenging cases. Most of the time, however, the collected data are rather sparse and/or contain limited information compared to that obtained from classical structural methods. These limitations call for integrative computational tools that can, by using a physics-based model, judiciously combine and accurately translate sparse experimental data into structural information. One such tool is macromolecular docking, defined as the process of building a molecular complex from its known individual components. The generic idea behind docking approaches can be comprehended easily via the following analogy (David Goodsell, The Machinery of Life, p.10):“Constructing molecular complexes is much like trying to build machines with Lego blocks: you may build a variety of different objects, but the final form is shaped and limited by the shape and connections of the underlying units.”Current“dockers”have exploited this analogy in two ways: They have either constructed algorithms based on limitations imposed by shape (shape complementarity methods) or connections (optimization methods that minimize a target energy function to satisfy connections –restraints- between supplied molecules). Albeit different in their approaches, both methods are confronted with similar challenges when building multi-component assemblies, addressing large conformational changes upon binding or selecting the“correct”solution among generated models. The work described in this thesis focuses on extending the capabilities of our in-house docking program HADDOCK by developing new protocols and incorporating new types of experimental information for being able to tackle those challenges. In Chapter 1, I provide a comprehensive introduction on the concept of integrative modeling. For that, I first introduce the sources of low-resolution information and then I describe various methods to integrate that information into the modeling process. I also depict the major challenges that the integrative modeling field is faced with: (1) Modeling large assemblies, i.e. dealing with multiple molecules simultaneously, (2) modeling dynamic molecular complexes, i.e. addressing large conformational changes upon binding, (3) constructing accurate scoring functions for fishing out the biologically relevant solution(s) among the generated pool of conformers, and (4) dealing with the degeneracy and ambiguity of methodological advances regarding those challenges taken from literature. I end this chapter by introducing our data-driven docking approach HADDOCK, which can integrate various sources of information to drive the modeling of biomolecular complexes. In the following Chapters 2-5, I position HADDOCK within the described integrative approaches, by presenting the protocols I developed during my PhD studies to cope with the depicted challenges (Chapter 2 refers to challenge (1), Chapter 3 to (2), Chapter 4 to (3) and Chapter 5 to (4)). In Chapter 2 I demonstrate that HADDOCK is able to handle multiple molecules and allows to impose different types of symmetries during docking. All of these options are implemented into a novel web-interface (the Multi-body interface) of HADDOCK, which allows the user to dock up to 6 biomolecules simultaneously, offers inclusion of experimental and/or bioinformatics data and supports several types of cyclic and dihedral symmetries during docking. The performance of the webserver is tested on a benchmark of six cases, containing five symmetric homooligomeric protein complexes and one symmetric protein-DNA complex. The quality of the modeled multimeric assemblies reveal that, in the presence of either bioinformatics and/or experimental data, HADDOCK is able to generate high-ranking near-native solutions (having an interface-RMSD range of 0.8 – 1.8Å) for all cases, demonstrating its ability to model symmetric multi-component assemblies. In order to address the challenge of modeling binding-induced large-scale conformational changes, in Chapter 3, I describe a novel Flexible Multi-domain (FMD) docking protocol. FMD aims at modeling both large-scale domain (hinge) motions and small- to medium-scale interfacial rearrangements: this is achieved by cutting the flexible molecule at a hinge region predicted by an elastic network model, and then by performing a simultaneous docking of all sub-parts/domains making use of HADDOCK's multi-body docking ability, presented in Chapter 2. The cut domains are kept together by connectivity restraints. The performance of this unprecedented range of conformational changes, from 1.5 to 19.5Å. In general, this FMD protocol could reproduce at least a near-native solution (interface-RMSD ≤ 4Å) for each case and rank these at the top. These results demonstrate that this protocol is poised to model conformational changes as large as 20Å! Finally, I show that the cumulative sum of eigenvalues obtained from the elastic network is indicative to predict the extent of the conformational change to be expected.In Chapter 4, I demonstrate that the scoring problem can be eased by incorporating information-based terms into the scoring function. For this purpose, I integrate low-resolution shape information obtained from either Ion Mobility Mass Spectrometry (IM-MS) or Small Angle X-Ray Scattering (SAXS) experiments into HADDOCK's scoring function and systematically assess the strengths and weaknesses of IM-MS- and SAXS-based scoring. For that, I make use of a large docking decoy set composed of 138 heterodimers generated by running HADDOCK in ab initio mode. The statistics calculated on this large decoy set suggests that IM-MS data are of too low resolution for selecting correct models, while scoring with SAXS data leads to a significant performance improvement. The performance of SAXS scoring however depends on the shape and arrangement of the complex and its constituents. In Chapter 5, I present two application examples of integrative modeling using HADDOCK. In the first one, I combine NMR chemical shift perturbations and mutagenesis data in order to predict the topology of the complex between the deubiquitination enzyme Josephin and a di-ubiquitin chain. The results of this integrative modeling suggest that Josephin imposes a preference for the cleavage of K48-linked di-ubiquitin chains, which is validated by biochemical experiments. In the second example, I model the transfer of the histone heterodimer H3-H4 from one histone chaperone, Asf1, to the other one, p48, based on Electron Paramagnetic Resonance data that suggests large structural rearrangements upon complexation and SAXS data, which provides insight on the globular assembly of the complex. This transfer takes place during the initiation of nucleosome assembly. The generated integrative models suggest putative inter-molecular interactions that can be tested by further mutagenesis experiments. In the final chapter of my thesis, Chapter 6, I give a perspective on future directions of HADDOCK. These include incorporating low-resolution shape data (coming from SAXS, cryo-EM) into sampling, implementing coarse-grained representations into HADDOCK in order to allow the modelling of even larger assemblies, and adapting HADDOCK's energy functions in such way that it will enable the modelling of membrane systems. At the end of this chapter, I provide a proof-of-principle example to underpin the fact that the accuracy of integrative models not only depends on the presented methodological advances and future directions, but also on the quality and the distribution of the data supplied. This example nicely pinpoints the fact that integrative modeling should not be seen as an end point, but rather a starting point for generating new hypothesis to be tested experimentally. In conclusion, the recent and ongoing developments in integrative modeling, together with the ones presented in this thesis, are important milestones in the modeling of protein-protein interactions. This should lead in the next decade to a significant growth in the field of structural computational biology and pave the route to increased accuracy, coverage and resolution of integrative models. Comprehending the underlying mechanistic of biomolecular function is not a simple process; computational structural biologists and experimentalists will therefore have to work in a collaborative and synergistic manner, in order to make significant advances in this field.
Benzer Tezler
- Semaforin 3F nin altın nanopartiküllere bağlanması, karakterizasyonu, HUVEC ve A549 hücre hatlarında anti-proliferatif etkisinin incelenmesi
Attachment of Semaphorin 3F onto gold nanoparticles its characterization, investigation of its anti-proliferative effect on HUVEC and A549 cell lines
GAMZE TAN
Doktora
Türkçe
2014
Bilim ve TeknolojiHacettepe ÜniversitesiGenel Biyoloji Ana Bilim Dalı
PROF. DR. MEHMET ALİ ONUR
- Beyaz dut ekstraktlarının antioksidan aktivitelerinin belirlenmesi
Determination of antioxidant activities of white mulberry extracts
TUĞÇE KÜÇÜKYILDIRIM
- Selection, design & applications of solid binding peptides for controlled biomineralization
Kontrollü biyomineralizasyon için katı yüzeylere bağlanan peptitlerin seçimi, tasarımı ve uygulamaları
MUSTAFA GÜNGÖRMÜŞ
Doktora
İngilizce
2012
BiyomühendislikUniversity of WashingtonMalzeme Bilimi ve Mühendisliği Ana Bilim Dalı
PROF. DR. MEHMET SARIKAYA
- İnsan solunum sinsityal virüsünün protein dizisi çeşitliliği dinamiklerinin incelenmesi
Dissecting the dynamics of human respiratory syncytial virus (hrsv) protein sequence diversity
FARUK ÜSTÜNEL
Yüksek Lisans
Türkçe
2022
BiyoteknolojiBezm-i Alem Vakıf ÜniversitesiBiyoteknoloji Ana Bilim Dalı
PROF. DR. ATİLLA AKDEMİR
DOÇ. DR. MOHAMMAD ASİF KHAN
- Dissecting the crosstalk between inhibitory receptor signaling and the T cell receptor in natural killer cells
Doğal öldürücü hücrelerde inhibitör reseptör sinyali ile T hücresi reseptörü arasındaki iletişimin araştırılması
EBRU ZEYNEP ERGÜN
Yüksek Lisans
İngilizce
2022
Allerji ve İmmünolojiBoğaziçi ÜniversitesiMoleküler Biyoloji ve Genetik Ana Bilim Dalı
DR. ÖĞR. ÜYESİ TOLGA SÜTLÜ