NMR and Structure Analysis Research Unit - Nmrstr

 NMR and Structure Analysis Research Unit Research topic #2 Understanding structure and function of DNAzymes General introduction Catalysis plays an essential role in modern chemical research and is central to the chemistry of life. It allows creating complex molecular structures and enables and accelerates chemical transformations. A billion years of evolution has allowed the development of protein based enzymes as natural catalysts with high efficiency and selectivity. The creation of artificial catalytic systems inspired by nature has now become a timely human endeavor. Many scientists actively search and develop new catalysts for all kinds of reactions, with molecular architectures which, a decade ago, where unthinkable or even impossible. Within the Organic and Biomimetic Chemistry Research group of Prof. Madder, nature itself is inspiring the use of a well‐known molecular architecture to develop new catalysts. Indeed, DNA represents a largely untapped potential for the design and development of novel catalysts, and even seems an unlikely candidate. However, by equipping short DNA strands with protein like functional groups via chemical synthesis, artificial DNA based enzymes can be created with catalytic sites fitted onto the fairly rigid, programmable and predictable DNA duplex structure. Recent developments within OBCR have allow the introduction of amino‐acid like side chain functionalities on the thymine base structure (fig. 1 left), while leaving the stability of the overall DNA duplex unaltered or even slightly improved, thus providing new and unexplored possibilities for the development of synthetic DNA enzymes or ‘DNAzymes’. Figure 1: Overview of the two types of ‘DNAzymes’ and their building blocks Currently, two families of DNAzymes are currently being investigated: Serine protease and esterase like DNAzymes. In the former, the goal is to successfully mimic the active site of α‐chymotrypsine, an enzyme that hydrolyzes peptide bonds by the collaboration between three amino acid functionalities: Serine (Ser), Histidine (His) and Aspartate (Asp). These three entities are engaged in the so‐called ‘catalytic triad’ by an intra‐residual hydrogen bond network, enabling its ultimate functionality. In order to mimic this ‘catalytic traid’, thymine modified building blocks can be introduced into the B‐DNA double helix giving rise to the serine protease DNAzyme (fig.2). The DNA scaffold used here (fig. 1) consists of a 28 base, or 14base‐pair duplex, with the central thymine rich box as the location for the catalytic site. By introducing the protein‐like side‐chains at the 5’ position of the thymine base, a potential catalytic site is created within DNA’s major groove. Figure 2: Serine protease mimicked active site The second type of DNAzymes is based solely on the introduction of multiple THis functionalities onto the predefined DNA scaffold, with the aim to develop esterase‐like DNAzymes. Inspiration for these systems is derived from a synthetic protein architecture developed by Baltzer and co‐workers, where the introduction of several histidine residues affords a 34000 fold enhancement of acid‐base hydrolase activity compared to background hydrolysis (fig. 3). With 7 possible locations for the introduction of THis modified building blocks in the current, thus paving the way for a complete array of possible hydrolase‐like Figure 3: Helix‐loop‐helix motif
DNAzymes (fig.1, right). Since both types of systems are based on the same DNA scaffold and thymine modified building blocks, approaches developed for the characterization of one system can be easily transferred to the study of the other type of DNAzyme. While the synthesis of these DNAzymes is now secured within OBCR, detailed information regarding the conformation, overall structure and dynamics is needed, providing guidance in the further development and future design of these systems. Indeed, do the side chains organize themselves into the desired catalytic site? Are they indeed located in the major groove? Does the scaffold influence the chemical state (pKa) or reactivity? What is the impact of these unnatural nucleotide bases on the local and global structure and stability? In order to answer these questions, a highly integrated structure analysis approach is currently being developed within NMRSTR, in collaboration with Prof. dr. A. Madder (OBCR), Prof. K. Vanhecke (Dept. Inorganic and Physical Chemistry) and Dr. E. Pauwels (Centre for Molecular Modeling) which relies on NMR, X‐ray, molecular modeling and isotope labeled synthesis of building blocks, as the main pillars supporting this project. For the academic year 2012‐2013 several thesis subjects are available within each discipline, either dedicated or in a combined fashion. Preferential base pairing modes of T‐T mismatches in DNA Promoter: Prof. J. Martins ([email protected]), Mentor: D.Buyst ([email protected]) Introduction The two new proposed catalytic DNAzymes, serine protease like and esterase‐like DNAzymes, are based on a 14‐mer B‐DNA helix. As can be seen in the sequence (fig. 1, middle) this scaffold features a central non‐Watson‐Crick base pair in the form of a T‐T mismatch. 5’-GACCATTTGCAGCG-3’
3’-CTGGTAATCGTCGC-5’
Figure 1: serine catalytic site (left), 14‐mer DNA scaffold (middle), , T8HisT6His Hydrolase DNAzyme (right)
The origin and function of this mismatch can be tracked down to the initial design stages of these new catalytic systems: in silico model building (performed by Prof. Dr. G. Bifulco, University of Salerno, fig. 1, left) showed that the catalytic triad of a serine protease could be successfully imitated when using the core 5’‐TTT‐3’, 3’‐AAT‐5’ section. Here, the T‐T mismatch is necessary to provide sufficient spatial proximity between the three functionalities in order to form a hydrogen bond network. Although this was not a prerequisite in the case of the esterase‐like DNAzymes, this central T‐T mismatch nevertheless provides a unique opportunity to introduce multiple THis residues, bearing histidine‐like functionalities in close vicinity to each other, with the aim at acid‐base co‐operation and catalytic activity. As an added bonus, the destabilisation caused by this mismatch is confined to its nearest‐neighbouring base pairs, with very little distortion of the sugar‐phosphate backbone of the DNA duplex [1]. Goals Considering the previous, it may be clear that the T‐T mismatch and its conformational behavior is of importance and interest in the further development of the proposed catalytic systems. In this context, it has already been Figure 2: W(↑) (le ), W(↓) (right) determined earlier that a T‐T mismatch can adopt two different modes of exchangeable wobble base pairs: W↑ (αβ) and W↓ (βα) also known as ‘wobble up’ and ‘wobble down’ respectively. In the past, it was generally accepted that these two conformations are in fast exchange with each other [2]. Nevertheless, it’s becoming clear that some degree of preferential pairing mode does exist in this T‐T mismatch, depending on the type of its flanking base pairs [1], [3]. In order to clarify this behavior, a systematic NMR investigation will be performed in order to study the sequence context effect on the pairing modes of the T‐T mismatch present in the DNA scaffold sequence. Here, multiple non‐modified DNA sequences will be studied where the neighboring base pairs of the mismatch will be altered in each sequence. This study will be carried out by applying both 1D 1H temperature studies and more advanced 2D 1H‐1H spectroscopic techniques of a series of DNA constructs, using the high‐field 700MHz Avance II spectrometer of the Interuniversitary NMR Facility at UGent. Ultimately, given sufficient time and material available, this approach could be applied on modified ‘catalytic’ DNA to investigate whether the same or different behavior can be observed. The results obtained do not stand alone. Similar investigations are planned via X‐ray diffraction (see topic with Prof. K. Vanhecke) and modeling (collaboration with Dr. E. Pauwels). Thus comparison of the behavior in solution and the solid state of the preferred wobble base pair state, and the level at which current simulation methods are capable of reproducing these, can be obtained in this fashion, providing for the first time an in depth, detailed look at this interesting ‘defect’. References 1. 2. 3. He, G.Y., C.K. Kwok, and S.L. Lam, Febs Letters, 2011. 585(24): p. 3953‐3958. Gervais, V., et al., European Journal of Biochemistry, 1995. 228(2): p. 279‐290. Gantchev, T.G., S. Cecchini, and D.J. Hunting, Journal of Molecular Modeling, 2005. 11(2): p. 141‐159. Structurele karakterisering van gemodificeerd DNA: predictie en confrontatie met NMR data Keywords: Structurele predictie, QM/MM, simulatie van NMR grootheden Probleemstelling Modificaties aan DNA treden in vivo op als gevolg van fysische of chemische schade en kunnen leiden tot kanker. Vanuit katalytisch oogpunt kan de synthetische introductie van modificaties in DNA echter een interessante opportuniteit bieden: de modulaire opbouw en relatief rigide structuur van de dubbele helix maken DNA immers een interessante templaatmolecule om functionele groepen gecontroleerd dicht bij elkaar te brengen. In het laboratorium voor Organische en Biomimetische Chemie (o.l.v. Prof. Annemieke Madder) zijn enkele indrukwekkende ontwikkelingen verwezenlijkt in dit opzicht: een nieuwe methodologie voor het plaatsspecifiek cross‐linken van nucleïnezuren en een nieuwe synthesemethodologie voor het inbouwen van meerdere aminozuur‐achtige zijketenfunctionaliteiten in een DNA duplex. Dit effent het pad voor de constructie van synthetische, gemodificeerde DNA systemen met een katalytische werking vergelijkbaar aan die van enzymes ‐ “DNAzymes”. Een dergelijk construct, geïnspireerd op de serine protease katalytische site, is momenteel in ontwikkeling (zie figuur). De performantie hiervan hangt echter zeer sterk af van de uiteindelijke conformatie en dynamica: zijn de drie aminozuur‐achtige modificaties voldoende in elkaars omgeving en zijn ze stabiel binnen die conformatie? Om een antwoord op deze vragen te vinden wordt in het laboratorium voor NMR en Structuuranalyse (o.l.v. Prof. José Martins) een nauwgezette analyse uitgevoerd van de structuur van verscheidene sequenties met behulp van NMR spectroscopie. De complexiteit van de resulterende spectra belemmert echter een eenduidige analyse, verder bemoeilijkt doordat sommige gemodificeerde nucleobasen kunnen voorkomen in meerdere ionisatietoestanden. Via simulaties biedt moleculaire modellering rechtstreeks inzicht in de conformatie en dynamica van deze molecules. Bovendien kan het ook aangewend worden om de NMR spectroscopische grootheden te simuleren voor een bepaalde structuur. Doelstelling Met behulp van moleculaire modellering wordt nagegaan in welke mate de drie‐dimensionale structuur van een DNAzyme kan worden voorspeld uitgaande van de sequentie. Daarbij wordt vertrokken van een beschrijving van het hele DNA oligomeer in een geschikt krachtveld, waarbij voor de aminozuur‐achtige modificaties bruikbare parameters zullen worden afgeleid. Voor de modificaties en hun directe moleculaire omgeving zal bovendien een kwantumchemische DFT methode worden gekozen in een zogenaamde QM/MM benadering. Op basis van deze beschrijving kan de optimale structuur worden afgeleid, aangevuld met een bepaling van de dynamica, gebruik makend van moleculaire dynamica simulaties. Terugkoppeling met het NMR experiment is mogelijk door voor de optimale structuren de NMR grootheden te voorspellen: proton en koolstof chemische shifts. Bovendien kunnen deze simulaties ook worden uitgevoerd voor verscheidene structuren uit de dynamica, hetgeen een uniek beeld biedt op de invloed van dynamica en temperatuur op de verdeling van de chemische shifts. Eenmaal een werkbaar protocol is uitgewerkt, kan deze methodiek worden toegepast op verscheidene sequenties waarvoor experimentele gegevens voorhanden zijn. Promotoren: Prof. Dr. J. Martins – [email protected] 09‐264.44.69 Dr. E. Pauwels – [email protected] 09‐264.65.62 Begeleiding: Dr. E. Pauwels – [email protected] 09‐264.65.62 Website: http://molmod.ugent.be/student‐corner Structural investigation of a T‐T mismatched DNA oligonucleotide by X‐ray diffraction. Promotors: Prof. K. Van Hecke ([email protected]), Prof. J. Martins Introduction The proposed project exists as a close collaboration between the Department of Inorganic and Physical Chemistry (prof. K. Van Hecke) and the Department of Organic Chemistry (prof. J. Martins, NMR and Structure Analysis unit; Prof. A. Madder, OBCR unit). Enzymes are of great importance in nature because of their ability to catalyse reactions in a selective and efficient way. For example, protease enzymes such as the serine protease α‐
chymotrypsin, are able to hydrolyse amide and peptide bonds under mild conditions [1]. Because of their well‐defined and predictable structure, DNA duplexes are currently being developed and used as predictable structural scaffolds onto which amino acid like functionalities are grafted in order to mimic an enzyme’s active site (DNAzymes). This can be achieved by equipping the oligonucleotides with the functional tools of peptide chains by modification of the nucleotide building blocks [2]. This results in a powerful combination of both predefined structural organization (duplex formation) and presence of catalytic entities. In this way, hydrolase type activity such as esterase and serine protease sites are being engineered (Figure 1). The DNA sequence, being considered as scaffold, was developed to afford optimal stability while maintaining spectral resolution for NMR investigation. We now wish to explore the possibility to obtain high‐resolution structures via X‐ray diffraction. These must provide suitable models for interpretation of functional data, and assist in further design and optimization of the DNAzymes. Of considerable interest in this context is the conformation of the T‐T mismatch, which represents a non‐Watson Crick (i.e. GC or AT) base pair. It has been shown by NMR in a DNA hairpin model that the T‐T mismatch adopts two conformations in rapid equilibrium, whose population ratio depends on the nature of the flanking base pairs, and is measurable using NOE spectroscopy. Whether this equilibrium and context dependence also occurs in the solid‐state is currently unknown. Goals The goal of the proposed project is threefold: 1. To elucidate the local conformation of the central T‐T mismatch in the native 14 basepair synthetic DNA duplex and its influence on the overall duplex structure. 2. To start exploring the variation in T‐T mismatch conformation caused by varying the flanking base‐pairs, and match these with NMR data. 3. To investigate the possibility of Hg incorporation into the DNA duplex as a potential crystallographic phasing tool, which will lead to the first structure of this novel metal‐DNA motif. Figure 1: Creation of an active site via duplex formation between two complementary base‐modified oligonucleotide strands [2]. Methods The 14 basepair DNA duplex, featuring the central TT‐mismatch, that is used as the basis for modification, will be investigated to provide a suitable model to characterize the structure, stability and reactivity of such modified DNA structures. The sequence of the duplex is 5’‐
GACC ATXTYTZGCAGCG‐3’/3’‐CTGGTAATWCGTCGC‐5’, with x/y/z/w the possible positions, where future protein side chain like modifications will be introduced. Crystals, suitable for X‐ray diffraction will be obtained by high‐throughput automated screening (Mosquito® crystallization robot) of the oligonucleotide. Initial hit conditions can be subsequently optimized in order to obtain diffraction‐quality crystals. These will subsequently be used to collect diffraction data at the beamline PXI/PXIII of the SLS synchrotron (Swiss Light Source, Villigen, Switserland). Once diffraction data are available, the phase problem will be solved by exploiting the selective incorporation of Hg ions by T‐T mismatched base pairs. This is possible by simple titration with HgClO4 and already demonstrated [3]. Solving this structure would thus allow the first determination of this novel metal‐DNA motif. References 1. 2. 3. Hedstrom, L. (2002) Chem. Rev., 102, 4501‐4523. Catry, M.A. & Madder, A. (2007) Molecules, 12, 114‐129. Anichina, J., Dobrusin, Z. & Bohme, D.K. (2010) J. Phys. Chem. B, 114, 15106‐15112. Isotope labeled synthesis of DNAzyme building blocks Promotor: Prof. A. Madder ([email protected]) Introduction DNA not only plays an important role in live as carrier of the genetic material; it has further been shown to possess features making the double helix structure an attractive tool for enantioselective catalysis. Based on the chemically stable and well defined chiral structure, recent literature reports on the capacity of DNA for introducing chirality into chemical reactions. We therefore currently develop DNA duplexes grafted with extra amino acid‐like functionalities to obtain so‐
called catalytic DNAzymes. In order to gain insight into the exact positioning of catalytic functionalities within the duplex, an active collaboration was set up with the NMR and Structure Analysis Research Group (NMRSTR). Figure 1: Serine protease mimicked active site In this master thesis you would thus be involved in the synthesis of 2H, 13C or 15N enriched analogues of histamine that will be incorporated into a nucleoside building block, further to be assembled into a suitable DNA sequence using the automated DNA synthesizer. According to your own interests you can opt for a purely synthetic work or also be involved in the advanced NMR structural investigation of the modified duplex. Furthermore you could choose to use the methodologies of Molecular Modeling to obtain further insight into the conformation and dynamics of the modified duplexes (in collaboration, Dr. Ewald Pauwels (Centre for Molecular Modeling, UGent). Detailed synthetic info The scheme below depicts the conventional synthetic route followed for the construction of the regular THis building block. Starting from 5‐iododeoxyuridine, a CO insertion reaction using histamine allows to introduce the desired imidazole moiety on the 5‐position of the base. Suitable protection is foreseen on the imidazole unit to prevent side reactions at the DNA synthesis stage. Boc
N
O
I
NH
N
HO
O
N
a
O
OH
N
c
N
O
O
NH
N
O
O
N P
O
NC
O
b
O
O
N
H
NH
N
DMTrO
O
O
OH
O
N
H
DMTrO
N
NH
HO
OH
Boc
N
O
N
H
O
Boc
N
O
a. histamine.2HCl, (PPh3)4Pd(0), Et3N,
DMF, CO (7bar), 20h, 70°C; tert-butyl dicarbonate, Et 3N,
DMF, 15min, rt, 58%;
b. DMTrCl, pyridine, 7h, 0°C
c. 2-cyano-N,N' -diisopropylchlorophosphoramidite,
DIPEA, DCM, 0°C to rt, 1h
The building block is further converted into a phosphoramidite, required for automated incorporation into the desired sequence using an automated DNA synthesizer. In the context of the current master thesis we will synthesize a labeled analogue of histamine to be then used in the CO insertion reaction. For this purpose, we will start from commercially available 13C,15N‐L‐Histidine and apply and compare two described procedures, an enzymatic one and a chemical one, for conversion to labeled histamine through a decarboxylation reaction. The generated labeled histamine will then be used in the building block synthesis depicted above. Alternatively routes will be explored to prepare the labeled histidine in‐house, starting from cheap and simple building blocks.