Molecular probes - Biochemical Society Transactions

46

Biochemical Society Transactions (2008) Volume 36, part 1

Molecular probes: insights into design and analysis from computational and physical chemistry Felicity L. Mitchell, Gabriel E. Marks, Elena V. Bichenkova, Kenneth T. Douglas and Richard A. Bryce1 Wolfson Centre for Structure-Based Design of Molecular Diagnostics, School of Pharmacy and Pharmaceutical Sciences, University of Manchester, Manchester M13 9PL, U.K.

Abstract The application of new molecular diagnostics to probe cellular process in vivo is leading to a greater understanding of molecular cytology at a sub-nanoscale level and is opening the way to individualized medicines. We review here three distinct fluorescence-based molecular probes, HyBeaconsTM , split-probe exciplexes and GFP (green fluorescent protein)-based FRET (fluorescence resonance energy transfer) systems. Through this, we highlight the insights into the mechanism and design that a combined computational and experimental approach can yield.

Introduction The confluence of modern molecular biology with biophysical chemistry has led to a growing number of powerful tools in the study of biological structure and properties. In the pharmaceutical industry, modern molecular design, computational biology, novel organic synthesis approaches and new instrumentation all mean that the discovery of lead drug candidates is no longer a limiting factor in developing new medicines. By analogy, the linkage of such technologies with modern cell biology can lead to a process that might be called rational, structure-based design of molecular diagnostics. Underpinning a large number of molecular probes and diagnostics in this growing arsenal is the modulation of fluorescence properties. In the present review, we consider three distinct fluorescence-based molecular detectors and consider the insights that computational and physical methods can provide in mechanistic analysis and design.

Detection of SNPs (single nucleotide polymorphisms) Driven by the success of genome projects, there has been considerable interest in the design of tools to detect specific nucleic acid sequences and, in particular, SNPs, where there is allelic variation in nucleic acid sequence at a single base-pair position. These types of sequence variations, which account for 90 % of those observed in the human genome, can act as markers for genetic conditions [1,2]. A number of tools exist for the detection of SNPs [3–8]. Key words: computational chemistry, fluorescence resonance energy transfer (FRET), fluorescence-based molecular probe, Molecular Dynamics simulation, single nucleotide polymorphism (SNP), structure-based design of molecular diagnostics. Abbreviations used: dsDNA, double-stranded DNA; FRET, fluorescence resonance energy transfer; GB/SA, generalized born/implicit solvent; GFP, green fluorescent protein; SNP, single nucleotide polymorphism; ssDNA, single-stranded DNA; TetR , tetracycline repressor. 1

To whom correspondence should be addressed (email [email protected]). Biochem. Soc. Trans. (2008) 36, 46–50; doi:10.1042/BST0360046

C The

C 2008 Biochemical Society Authors Journal compilation

One example is the HyBeaconTM probe, developed by LGC [4]. This system involves a fluorophore covalently attached to an oligonucleotide with a sequence complementary to that of the targeted DNA (Figure 1). The attachment point is an internal nucleotide. The function of the probe is essentially independent of any secondary structure. On hybridization to the target DNA and formation of the duplex, the HyBeaconTM system exhibits increased fluorescence. Thus these systems are a simple, cost-effective means to detect SNPs, successfully distinguishing between homozygous and heterozygous samples in PCR assays. In an effort to understand the mechanistic basis of HyBeaconsTM , computational, spectroscopic and melting temperature analyses [9] were performed for the HyBeaconTM system FVG1, which has the dye fluorescein covalently attached to a uracil residue (Figure 1). Simulated annealing using Molecular Dynamics of the hybridized duplex system indicated quite clearly the absence of any specific interaction of the attached fluorophore with the DNA moiety, either with the backbone or with the major or minor grooves (Figure 2). Similarly, room temperature Molecular Dynamics simulations revealed principally conformations with the fluorophore projecting away from the DNA into the solvent. The solvent was modelled using the GB/SA (generalized born/implicit solvent) model. Consistent with this lack of dye–DNA interaction in the duplex, UV–visible and fluorescence spectroscopic studies indicated similar emission and excitation λmax and absorption properties for the unconjugated dye and for the dye incorporated into the dsDNA (double-stranded DNA) as a part of the oligonucleotide probe hybridized to the complementary target DNA. By comparison, the dye within the single-stranded unhybridized probe displayed red-shifted emission and absorption bands suggestive of dye–ssDNA (single-stranded DNA) interactions in this system, potentially indicating quenching of the dye by

Bringing Together Biomolecular Simulation and Experimental Studies

Figure 1 Oligonucleotide sequences of the HyBeaconTM probes, their unmodified counterparts, reverse complements and structures of the fluorophore 6

Figure 2 Six superimposed annealed structures of the (FVG1 + FVG1RC) duplex using GB/SA potential The fluorophore is highlighted in different colours for each structure. All conformations show the fluorophore extending into the solvent and not bound to DNA.

neighbouring bases in the DNA. In general, the intensities of the fluorophore absorption and emission bands substantially decreased when the fluorophore was incorporated into the

single-stranded unhybridized probe. This was particularly evident for guanine-rich systems, where quenching would be expected to be greatest. The hybridized system possessing covalently attached fluorophore exhibited a lower melting temperature than that of the corresponding duplex DNA with the dye absent. This could also point to favourable dye–DNA interactions stabilizing ssDNA in the unhybridized probe. This combination of calculation and experiment therefore sheds light on the action of HyBeaconsTM where the basis of increased fluorescence on recognition of target DNA is due to disruption of the quenching interactions between the dye and nucleobases of the ssDNA, and provides insights into design, the basis of new directions. Another system for detection of DNA sequences and, in particular, SNPs is based on exciplex formation, resulting in emission of characteristic long-wavelength fluorescence [3]. In this new approach, the two distinct exciplex-forming partners, an electron donor and an acceptor, are fused to two oligonucleotide probes complementary to the adjacent sites of the target DNA. Hybridization of these split-probes with the target DNA sequence assembles the probes and juxtaposes the exci-partners closely in space, which leads to exciplex emission upon specific excitation (Figure 3a). Exciplex-based detectors, because of the ability to exploit different exci-partners (unlike excimers, where the partners must always be identical), permit the modulation of spectral properties. However, exciplexes have, traditionally, only been observed in low dielectric solvents, and so it was not immediately evident how to make such a system operate in aqueous solutions for DNA detection. To provide insights into the operation of these systems and to explore the issue of solvent polarity, simple intramolecular exciplex systems such as 1–5 (Figure 3b) were constructed and characterized computationally and spectroscopically in various solvents C The


47

48


Figure 3 Exciplex-based systems

Figure 4 Two putative, minimum energy conformations of 1

(a) Schematic representation of the split-probe approach showing self-assembly of exciplex components (X and Y) induced by hybridization

calculated from conformational analysis using the MMFF (Merck Molecular Force Field) in GB/SA solvent

of oligonucleotide-probes with complementary nucleic acid target. (b) Chemical structures of model intramolecular exciplex systems 1–5.

[10]. It was found that these systems exhibited exciplex emission even at very high solvent polarity, for example in N-methylformamide, which has a dielectric constant of 182. Structural analysis involved NMR and Monte Carlo simulations at 300 K to identify representative conformations of small, flexible molecule 1. Interestingly, the conformation of 1 associated with the calculated lowest energy cluster of structures, both in water and chloroform, gave good agreement with NOE (nuclear Overhauser effect) information from NMR and indicated a non-stacked orientation (Figure 4). Higher energy folded forms of 1 were also identified by the simulations, which would provide the orbital overlap expected for exciplex formation. These folded forms could be readily formed by internal torsion angle rotations, as might be expected on photoexcitation.

Protease-sensitive signalling by mutants of GFP (green fluorescent protein) Finally, we turn to consider a molecular probe involving a protein as the fluorescent tag. Due to its unique spectral properties, GFP and its mutants have been widely deployed as biological tools [11,12], often as genetic fusion constructs, to study a wide range of cellular phenomena, for example in measuring intracellular Ca2+ levels [13]. GFP’s fluorescence properties stem from its p-hydroxybenzylideneimidazolidinone chromophore, which is formed by posttranslational cyclization of the polypeptide backbone between residues Ser65 , Tyr66 and Gly67 , and by α,βdehydrogenation of Tyr66 . GFP can be used as a FRET (fluorescence resonance energy transfer) partner for visualization C The


of cellular processes, either with another protein (e.g. blue fluorescent protein) or with a chemical partner. Recently, an intramolecular FRET probe based on GFP mutants was developed [14,15]. In the GFP mutant, a dye is linked to a cysteine residue introduced by site-directed mutagenesis. Also introduced is a tryptic cleavage sequence, which lies between the location of the dye and the GFP chromophore (Figure 5). Thus the system operates as a sensitive detector of proteolytic enzymes, as FRET is abolished by cleavage at this intervening amino acid sequence. The starting point for design of this probe involved constructing an in silico homology model of GFP mutant, GFPuv5, using the crystal structure of wild-type GFP [14]. GFPuv5 differs from wildtype GFP at seven points in sequence and fluoresces 240 times more strongly. The locations of various amino acids in this model were examined for their suitability for introducing a cysteine residue. Interestingly, Glu6 and Ile229 appeared to be suitable candidates, as modelling indicated that an attached ˚ from the eosin dye would be separated by approx. 21 and 23 A GFP chromophore when attached at I229C and E6C respectively. However, significantly stronger FRET was observed for the I229C mutant. Glu6 resides in the N-terminal region, whereas Ile229 lies in the C-terminal region. Subsequent Molecular Dynamics simulations of the GFPuv5 E6C and I229C constructs suggested structures of the N-terminal region where the eosin dye is more distant than indicated from static modelling via the homology model (Figure 6).

Bringing Together Biomolecular Simulation and Experimental Studies

Figure 5 Schematic representation of a GFP construct

Figure 6 Molecular Dynamics simulations of GFPuv5 mutants (left) I229C and (right) E6C suggest shorter average chromophore–dye distances for I229C (the remainder of the protein’s N- and C-termini omitted for clarity)

Discussion and conclusions Consideration of these three fluorescence-based molecular probes, HyBeaconsTM , split-probe exciplexes and GFP-based FRET systems, clearly illustrates the strength of a combined computational and experimental approach in providing molecular insights and design directions. We observe that with the seemingly inexorable advance in computer power, the systems being addressed by computational methods are becoming increasingly ambitious. This applies in the first instance to the scale of systems studied: Zhao et al. [16] combined experimental with theoretical techniques to investigate the use of restriction enzymes in identifying SNPs using a synthetic nanopore. A single mutation in the recognition site for the restriction enzyme, such as an SNP, can easily be detected as a change in the threshold voltage. Molecular Dynamics simulations revealed that the voltage threshold for permeation through a synthetic nanopore of dsDNA bound to a restriction enzyme is associated with a nanonewton force required to rupture the DNA–protein complex. Secondly, increased computational resource permits models of greater sophistication. FRET between tetracycline and tryptophan can occur in the complex of tetracycline and the protein, TetR (tetracycline repressor). The time-

resolved fluorescence spectra of this system were studied by a combination of Molecular Dynamics simulations with subsequent single-point hybrid QM/MM (quantum mechanical/molecular mechanical) calculations. Using the configuration interaction level of theory, Beierlein et al. [17] thus calculated the tryptophan vertical absorption and fluorescence energies and intensities as well as relative FRET rate constants in the tetracycline–TetR complex, and were able to show that the typical rotamer model employed for tryptophan can be applied to systems where FRET acceptors are attached to the protein, but that care was required in interpretation of the fitted lifetimes. Interestingly, in addition to the synergy between experiment and increasingly powerful computational approaches, theoretical methods can also serve to challenge the assumptions of experiment. This is typified by a recent Molecular Dynamics study of hen’s-egg white lysozyme with two fluorescent probes covalently attached [18]. These simulations suggest that the orientational factor, κ 2 , on which efficient FRET depends, deviates significantly from the value of 2/3 typically assumed in FRET studies. This κ 2 value of 2/3 stems from the assumption of one free probe and one fixed probe. Thus the findings of these Molecular Dynamics simulations C The


49

50


may prove to have quite general implications for the interpretation of FRET experiments and will lead to greater interplay between experiment and theory in the pursuit of new and improved fluorescence-based biological probes.

References 1 Brookes, A.J. (1999) The essence of SNPs. Gene 234, 177–186 2 Collins, F.S., Patrinos, A., Jordan, E., Chakravarti, A., Gesteland, R. and Walters, L. (1998) New goals for the U.S. Human Genome Project: 1998–2003. Science 282, 682–689 3 Bichenkova, E.V., Gbaj, A., Walsh, L., Savage, H.E., Rogert, C., Sardarian, A.R., Etchells, L.L. and Douglas, K.T. (2007) Detection of nucleic acids in situ: novel oligonucleotide analogues for target-assembled DNA-mounted exciplexes. Org. Biomol. Chem. 5, 1039–1051 4 French, D.J., Archard, C.L., Brown, T. and McDowell, D.G. (2001) HyBeaconTM probes: a new tool for DNA sequence detection and allele discrimination. Mol. Cell. Probes 15, 363–374 5 Holland, P.M., Abramson, R.D., Watson, R. and Gelfand, D.H. (1991) Detection of specific polymerase chain reaction product by utilizing the 5 →3 exonuclease activity of Thermus aquaticus DNA polymerase. Proc. Natl. Acad. Sci. U.S.A. 88, 7276–7280 6 Kong, D.M., Gu, L., Shen, H.X. and Mi, H.F. (2002) A modified molecular beacon combining the properties of TaqMan probe. Chem. Commun., 854–855 7 Tyagi, S. and Kramer, F.R. (1996) Molecular beacons: probes that fluoresce upon hybridization. Nat. Biotechnol. 14, 303–308 8 Whitcombe, D., Theaker, J., Guy, S.P., Brown, T. and Little, S. (1999) Detection of PCR products using self-probing amplicons and fluorescence. Nat. Biotechnol. 17, 804–807 9 Marks, A.H.R., Bhadra, P.K., McDowell, D.G., French, D.J., Douglas, K.T., Bichenkova, E.V. and Bryce, R.A. (2005) Molecular basis of action of HyBeaconTM fluorogenic probes: a spectroscopic and molecular dynamics study. J. Biomol. Struct. Dyn. 23, 49–62

C The


10 Bichenkova, E.V., Sardarian, A.R., Wilton, A.N., Bonnet, P., Bryce, R.A. and Douglas, K.T. (2006) Exciplex fluorescence emission from simple organic intramolecular constructs in non-polar and highly polar media as model systems for DNA-assembled exciplex detectors. Org. Biomol. Chem. 4, 367–378 11 Chalfie, M. (1995) Green fluorescent protein. Photochem. Photobiol. 62, 651–656 12 Heim, R. and Tsien, R.Y. (1996) Engineering green fluorescent protein for improved brightness, longer wavelengths and fluorescence resonance energy transfer. Curr. Biol. 6, 178–182 13 Romoser, V.A., Hinkle, P.M. and Persechini, A. (1997) Detection in living cells of Ca2+ -dependent changes in the fluorescence emission of an indicator composed of two green fluorescent protein variants linked by a calmodulin-binding sequence. A new class of fluorescent indicators. J. Biol. Chem. 272, 13270–13274 14 Suzuki, M., Ito, Y., Sakata, I., Sakai, T., Husimi, Y. and Douglas, K.T. (2005) Caspase-3 sensitive signaling in vivo in apoptotic HeLa cells by chemically engineered intramolecular fluorescence resonance energy transfer mutants of green fluorescent protein. Biochem. Biophys. Res. Commun. 330, 454–460 15 Suzuki, M., Ito, Y., Savage, H.E., Husimi, Y. and Douglas, K.T. (2004) Protease-sensitive signalling by chemically engineered intramolecular fluorescent resonance energy transfer mutants of green fluorescent protein. Biochim. Biophys. Acta 1679, 222–229 16 Zhao, Q., Sigalov, G., Dimitrov, V., Dorvel, B., Mirsaidov, U., Sligar, S., Aksimentiev, A. and Timp, G. (2007) Detecting SNPs using a synthetic nanopore. Nano Lett. 7, 1680–1685 17 Beierlein, F., Lanig, H., Schurer, G., Horn, A.H.C. and Clark, T. (2003) Quantum mechanical/molecular mechanical (QM/MM) docking: an evaluation for known test systems. Mol. Phys. 101, 2469–2480 18 VanBeek, D.B., Zwier, M.C., Shorb, J.M. and Krueger, B.P. (2007) Fretting about FRET: correlation between κ and R. Biophys. J. 92, 4168–4178

Received 18 September 2007 doi:10.1042/BST0360046