Theoretical Studies On The Structures And Spectra Of DNA-ligands And Bases In Condensed Phase

It has aroused general interests to accurately describe the structures and spectroscopic properties of complex condensed phase systems with the theoretical calculations.However,due to the large size,flexible structure,and complicated chemical environment of complex systems,the traditional quantum chemical methods are difficult to calculate their structures and spectral properties.It is because that the computational costs of these methods increase rapidly with the increasing of system sizes.The multiscale or low-scale quantum chemistry method,combined with the molecular mechanics(MM)and quantum mechanics(QM),can be used to reveal the influence of environmental effects,such as solvent and hydrogen bond(HB),on the structure and spectral properties of complex systems.In this thesis,the molecular dynamics(MD)simulation,QM/MM method,the generalized energy-based fragmentation(GEBF)approach,and experiment have been employed to reveal the relationship between the structure and spectra of complex systems at the molecular level.The studies include the structure and electronic spectra of DNA-ligand complexes in the condensed-phase,the influence of structural changes and chemical descriptors on the spectral properties,and the prediction of DNA-ligand binding modes and sites.The main research contents and innovations of this thesis are summarized as follows:The DNA-binding dyes have unique fluorescent properties,which can be widely used as fluorescent probes for sensing,such as chemical detection and biological imaging.In Chapter 3,based on the QM and machine learning results of thiazole orange homodimeric(TOTO)dye binding to a single-strand DNA(ss DNA),poly(d G)_n(n=2,4,6,8),we show that the n=6 complex shows 300-fold stronger fluorescent intensity than n=2,4,and 8 complexes,which is due to the length match between TOTO and poly(d G)6.Using the machine learning to analyze the MD trajectories suggests that TOTO intramolecular end-to-end distance and folded dihedral are two important features,which can be used to classify TOTO structure into stretch or hairpin type(the end-to-end distance greater or less than 18.0(?)).The QM low-lying excited-states calculations reveal that thep-pend-stacking binding structures with stretch TOTO have significant fluorescence enhancement,which can be attributed to the localizedp-p^*orbital transitions.However,most intercalative binding structures with hairpin TOTO have larger binding energies(predicted by GEBF method),which can bring about large intermolecular charge transfers and lead to fluorescence quenching.Therefore,the degree of charge transfer induced by the DNA-dye binding mode can be used to regulate the fluorescence enhancement and quenching,which suggests that the dyes can be screened or designed to probe the DNA(or RNA)fragments with specific sequences.The calculations of excited states of DNA or RNA help to understand the photoexcitation process of complex systems in photochemistry and photophysics.To reduce the computational cost and consider the environmental effects,such as explicit solvent,bases can be used to study the excited-stated properties because that the light absorption of nucleic acid occurs on the base.In Chapter 4,combinations of GEBF,GEBF-QM/MM,and QM/MM approaches have been executed to investigate the electronic absorption spectra ofp-p⁽⁹⁾transition for uracil in aqueous solution,amorphous solid,and uracil crystal.The results show that all the absorption spectra of uracil in the condensed phase can well reproduce the experimental values,and indicate that the intermolecular interactions in terms of molecular packing have a great influence on the absorption spectra.The absorption spectra of amorphous uracil have large redshifts due to the HB and strongp-pstacking interactions.However,the spectra of uracil in aqueous solution only have relatively small redshifts,which arises from the synergy of HB,and the long-range electrostatic and polarization interactions.The uracil crystal also has a small red-shifted peak due to the HB and weakp-pstackings.It suggests that the GEBF,combined with QM/MM method,can be carried out to accurately calculate the absorption spectra of condensed phase systems,and can enable us to deeply understand the influence of structures and intermolecular interactions on the spectra.Understanding the binding between DNA and small ligand molecules can help design new drugs and fluorescent probes,which have potential and important applications in chemical and biological fields.However,it remains as a challenge to accurately predict the binding modes and sites of DNA-ligand at the molecular level through QM methods.In Chapter 5,the ab initio high-throughput screening using a combination of molecular docking,MD simulation,and GEBF approach(a low-scale quantum mechanics method)are performed to predict the binding modes and sites of three complexes between DNA and drugs or photoelectric molecules.It was found that the DNA and each ligand forms a small groove binding pattern,which is consistent with the experimental results.The GEBF binding energy can screen out the reasonable structure that matches the experimental binding site,which is better than the MM or semi-empirical QM screened structures.The GEBF predicted structure can also reproduce the binding sites of surrounding water molecules,which form HB networks with DAPI molecule.Furthermore,the calculated absorption spectra of DNA-DAPI complexes are very close to the experimental results by taking the solvent,electrostatic and polarization interactions into account.Therefore,the high-throughput computational screening based on the ab initio calculation with GEBF approach can accurately and effectively predict and understand the binding sites and corresponding properties of DNA-ligands.