Dual Functionality of 6-Methylthioguanine: Synergistic Effects Enhancing the Photolability of DNA Nucleobases.

A small chemical modification of the nucleobase structure can significantly enhance the photoactivity of DNA, which may incur DNA damage, thus holding promising applications in photochemotherapy treatment of cancers or pathogens. However, single substitution confers only limited phototoxicity to DNA. Herein, we combine femtosecond and nanosecond time-resolved spectroscopy with high-level ab initio calculations to disentangle the excited-state dynamics of 6-methylthioguanine (me6-TG) under variable wavelength UVA excitation (310-330 nm). We find that double substitution of nucleobases (thionation and methylation) boosts the photoactivity by introducing more reactive channels. Intriguingly, 1nNπ*, rather than 1nSπ*, acts as the doorway state engendering the formation of the long-lived reactive triplet state in me6-TG. The 1nNπ* induces a low spin-orbit coupling of 8.3 cm-1, which increases the intersystem crossing (ISC) time (2.91 ± 0.14 ns). Despite the slowed ISC, the triplet quantum yield (ΦT) still accounts for a large fraction (0.6 ± 0.1), consistent with the potential energy surface that favors excited-state bifurcation to 1nNπ*min (3.36 ± 0.15 ps) rather than 1ππ*min (5.05 ± 0.26 ps), such that the subsequent ISC to triplet via 1nNπ*min constitutes the main relaxation pathway in me6-TG. Although this ΦT is inferior to its single-substituted predecessor 6-thioguanine (6-TG, 0.8 ± 0.2), the effect of thionation in synergy with methylation opens a unique C-S bond cleavage pathway through crossing to a repulsive 1πσ* state, generating thiyl radicals as highly reactive intermediates that may invoke biological damage. This photodissociation channel is extremely difficult for conventional nucleobases. These findings demonstrate the synergistic effects of double functionality substitution in modulating excited-state dynamics and enhancing the photolabile character of DNA nucleobases, providing inspirations for the rational design of advanced photodynamic and photochemotherapy approaches.

BODIPY-Based Photodynamic Agents for Exclusively Generating Superoxide Radical over Singlet Oxygen.

Developing Type-I photosensitizers is considered as an efficient approach to overcome the deficiency of traditional photodynamic therapy (PDT) for hypoxic tumors. However, it remains a challenge to design photosensitizers for generating reactive oxygen species by the Type-I process. Herein, we report a series of α,ß-linked BODIPY dimers and a trimer that exclusively generate superoxide radical (O2-. ) by the Type-I process upon light irradiation. The triplet formation originates from an effective excited-state relaxation from the initially populated singlet (S1 ) to triplet (T1 ) states via an intermediate triplet (T2 ) state. The low reduction potential and ultralong lifetime of the T1 state facilitate the efficient generation of O2-. by inter-molecular charge transfer to molecular oxygen. The energy gap of T1 -S0 is smaller than that between 3 O2 and 1 O2 thereby precluding the generation of singlet oxygen by the Type-II process. The trimer exhibits superior PDT performance under the hypoxic environment.

Photo-dissociation mechanism of trifluoroacetyl chloride in the gas phase: AIMS dynamic simulations.

In this article, the structures and energies of CF3COCl in the low-lying electronic states have been determined by SA-2-CAS(8,7)/6-31G* and SA-2-MSPT2(8,7)/6-31G* calculations, which include equilibrium geometries, transition states, and three minimum-energy conical intersections (CI-1, CI-2, and CI-3) between S0 and S1 states. The AIMS method was used to carry out non-adiabatic dynamic simulations with the ab initio calculation performed at the SA-2-CAS(8,7)/6-31G* level. Upon irradiation to the S1 state, CF3COCl first relaxes to S1 minimum and then overcomes the ∼10 kcal/mol (TSS1_CCl) or ∼30 kcal/mol (TSS1_CO) barrier to the conical intersection region CI-1 or CI-3 (minor), with the S1 → S0 transition probability of 63:1. After non-adiabatic transition to the S0 state through CI-1, trajectories mainly distribute to three different reaction pathways, with one going back to S0 minimum through shortening of the C-Cl bond, the other forming CF3CO and Cl radicals by continuous elongation of the C-Cl distance, and another dissociating into CF3 + CO + Cl and running into the CI-3 region through elongation of C-C and C-Cl distances. Moreover, we found that the trajectories would recross to the S1 state with the recrossing probability of 13.9% through the CI-3 region due to the extremely sloped topographic character of CI-3. On the basis of time evolution of wavefunctions simulated here, the product ratio of CF3 + CO + Cl and CF3CO + Cl is 53.5%:18.4%, which is consistent with the experimental value of 3:1. We further explain the photo-dissociation wavelength dependence of CF3COCl, and the product ratio of CF3 + CO + Cl increases with the increase in total energy.

Theoretical studies on the photochemistry of 2-nitrofluorene in the gas phase and acetonitrile solution.

The photophysical and photochemical mechanisms of 2-nitrofluorene (2-NF) in the gas phase and acetonitrile solution have been studied theoretically. Upon ∼330 nm irradiation to the first bright state (1ππ*), the 2-NF system can decay to triplet excited states via rapid intersystem crossing (ISC) processes through different surface crossing points or to the ground state via an ultrafast internal conversion (IC) process through the S1/S0 conical intersection. The 1nπ* dark state will serve as a bridge when the system leaves the Franck-Condon (FC) region and approaches to the S1 minimum. The molecule maintains a planar geometry during the excited-state relaxation processes. The differences on excitation properties such as electronic configurations and spin-orbit coupling (SOC) interactions between those in the gas phase and acetonitrile solution cannot be neglected, indicating possible changes on the efficiency of the related ISC processes for the 2-NF system in solution. Once arrived at the T1 state, it would further decay to the S0 state or photodegrade into the Ar-O˙ and NO˙ free radicals. During the intramolecular rearrangement process, the twisting of the nitro group out of the aromatic-ring plane is regarded as a critical structural variation for the photodegradation of the 2-NF system. The free radicals finally form through oxaziridine-type intermediate and transition state structures. The present work provides important mechanistic insights to the photochemistry of nitro-substituted polyaromatic compounds.

Selenium substitution effects on excited-state properties and photophysics of uracil: a MS-CASPT2 study.

The photophysics of selenium-substituted nucleobases has attracted recent experimental attention because they could serve as potential photosensitizers in photodynamic therapy. Herein, we present a comprehensive MS-CASPT2 study on the spectroscopic and excited-state properties, and photophysics of 2-selenouracil (2SeU), 4-selenouracil (4SeU), and 2,4-selenouracil (24SeU). Relevant minima, conical intersections, crossing points, and excited-state relaxation paths in the lowest five electronic states (i.e., S0, S1, S2, T2, and T1) are explored. On the basis of these results, their photophysical mechanisms are proposed. Upon photoirradiation to the bright S2 state, 2SeU quickly relaxes to its S2 minimum and then moves in an essentially barrierless way to a nearby S2/S1 conical intersection near which the S1 state is populated. Next, the S1 system arrives at an S1/T2/T1 intersection where a large S1/T1 spin-orbit coupling of 430.8 cm-1 makes the T1 state populated. In this state, a barrier of 6.8 kcal mol-1 will trap 2SeU for a while. In parallel, for 4SeU or 24SeU, the system first relaxes to the S2 minimum and then overcomes a small barrier to approach an S2/S1 conical intersection. Once hopping to the S1 state, there exists an extended region with very close S1, T2, and T1 energies. Similarly, a large S1/T1 spin-orbit coupling of 426.8 cm-1 drives the S1→ T1 intersystem crossing process thereby making the T1 state populated. Similarly, an energy barrier heavily suppresses electronic transition to the S0 state. The present work manifests that different selenium substitutions on uracil can lead to a certain extent of different vertical and adiabatic excitation energies, excited-state properties, and relaxation pathways. These insights could help understand the photophysics of selenium-substituted nucleobases.

QM and ONIOM studies on thermally activated delayed fluorescence of copper(i) complexes in gas phase, solution, and crystal.

Herein, we have employed B3LYP and TD-B3LYP methods with the QM/MM approach to study the thermally activated delayed fluorescence (TADF) phenomenon of two Cu(i) complexes bearing 5-(2-pyridyl)-tetrazolate (PyrTet) and phosphine (POP) ligands in the gas phase, solution, and crystal form. On the basis of spectroscopic properties, ground- and excited-state geometric and electronic structures, and related radiative and nonradiative rates, we have found that (1) the S1 and T1 excited states have clear metal-to-ligand charge transfer character from the Cu(i) atom to the PyrTet group; (2) the S1 and T1 states have a very small energy gap ΔES1-T1, less than 0.18 eV, which makes the forward and reverse intersystem crossing ISC and rISC processes between the S1 and T1 states very efficient; and (3) the low-frequency vibrational modes related to the torsional motion of the POP and PyrTet groups are found to have significant Huang-Rhys factors and are responsible for the efficient ISC and rISC rates. However, the corresponding Huang-Rhys factors are remarkably suppressed in the crystal compared with those in the gas phase and in solution due to the rigidity of the crystal surroundings; as a result, the ISC and rISC rates are accordingly reduced slightly in the crystal. This comparison also demonstrates that the surrounding effects are very important for modulating the photophysical properties of the Cu(i) complexes. Finally, our work gives helpful insights into the TADF mechanism of the Cu(i) compounds, which could assist in rationally designing TADF materials with excellent performance.

How Photoisomerization Drives Peptide Folding and Unfolding: Insights from QM/MM and MM Dynamics Simulations.

Photoswitchable azobenzene cross-linkers can control the folding and unfolding of peptides by photoisomerization and can thus regulate peptide affinities and enzyme activities. Using quantum mechanics/molecular mechanics (QM/MM) methods and classical MM force fields, we report the first molecular dynamics simulations of the photoinduced folding and unfolding processes in the azobenzene cross-linked FK-11 peptide. We find that the interactions between the peptide and the azobenzene cross-linker are crucial for controlling the evolution of the secondary structure of the peptide and responsible for accelerating the folding and unfolding events. They also modify the photoisomerization mechanism of the azobenzene cross-linker compared with the situation in vacuo or in solution.

Mechanism for the nonadiabatic photooxidation of benzene to phenol: orientation-dependent proton-coupled electron transfer.

An efficient catalytic one-step conversion of benzene to phenol was achieved recently by selective photooxidation under mild conditions with 2,3-dichloro-5,6-dicyano-p-benzoquinone (DDQ) as the photocatalyst. Herein, high-level electronic structure calculations in the gas phase and in acetonitrile solution are reported to explore the underlying mechanism. The initially populated (1)ππ* state of DDQ can relax efficiently through a nearby dark (1)nπ* doorway state to the (3)ππ* state of DDQ, which is found to be the precursor state involved in the initial intermolecular electron transfer from benzene to DDQ. The subsequent triplet-state reaction between DDQ radical anions, benzene radical cations, and water is computed to be facile. The formed DDQH and benzene-OH radicals can undergo T1→S0 intersystem crossing and concomitant proton-coupled electron transfer (PCET) to generate the products DDQH2 and phenol. Two of the four considered nonadiabatic pathways involve an orientation-dependent triplet PCET process, followed by intersystem crossing to the ground state (S0). The other two first undergo a nonadiabatic T1→S0 transition to produce a zwitterionic S0 complex, followed by a barrierless proton transfer. The present theoretical study identifies novel types of nonadiabatic PCET processes and provides detailed mechanistic insight into DDQ-catalyzed photooxidation.

Photodecarbonylation mechanism of cyclopropenone in the gas phase: electronic structure calculation and AIMS dynamics simulation.

In this article, structures and energies of cyclopropenone in the low-lying electronic states have been determined by the CASSCF and MS-CASPT2 calculations with different basis sets. Two minimum-energy conical intersections (CI-1 and CI-2) between S0 and S1 were obtained and their topographic characters were characterized by the SA4-CAS(10,9) calculated energy gradients and nonadiabatic coupling vectors. The AIMS method was used to carry out nonadiabatic dynamics simulation with ab initio calculation performed at the SA4-CAS(10,9) level. On the basis of time evolution of wave functions simulated here, the S1 lifetime is fitted to be 125 fs with a pure exponential decay for the S1 electronic population. The CI-1 intersection is mainly responsible for ultrafast S1→S0 nonadiabatic transition and the photoinduced decarbonylation is a sequential process, where the first C-C bond is broken in the S1 state and fission of the second C-C bond occurs in the S0 state as a result of the S1→S0 internal conversion via the CI-1 region. As a minor channel through the CI-2 region, the decarbonylation proceeds in an asynchronous concerted way. Effects of the S1 excess energies and the S1-S0 energy gap on the nonadiabatic dynamics were examined, which reveals that the S1→S0 nonadiabatic transition occurs within a small energy gap and high-energy conical intersection regions can play an important role. The present study provides new insights into mechanistic photochemistry of cyclopropenones and reveals that the AIMS dynamics simulation at a high-accuracy ab initio level is a powerful tool for exploring a mechanism of an ultrafast photochemical reaction.

Photo-induced isomerization of ethylene-bridged azobenzene explored by ab initio based non-adiabatic dynamics simulation: a comparative investigation of the isomerization in the gas and solution phases.

Azobenzene is one of the most widely used photoactive units and recently an ethylene-bridged azobenzene (BAB) was reported to have greatly enhanced conversion efficiency, quantum yield, and other favorable properties. As the first step towards exploring its photo-switchable character in real systems, we report here a systematic study on the photoisomerization dynamics between trans (E) and cis (Z) isomers in the gas phase and the CH3OH solution, using ab initio based surface hopping and molecular dynamics, which is the first report of dynamics simulation to reveal the environmental effects on BAB photoreactions. Results show that while the relatively faster S1 relaxation of the photo-induced E → Z process is only mildly affected by the solvent effect, the relatively slower S1 relaxation of the reverse reaction becomes even slower in the solution compared to the gas phase. The subsequent S0 dynamics from the conical intersection between S1 and S0 (CI_E) to Z is accelerated in solution compared to the gas phase because of avoided re-crossing to the S1 state, while the S0 dynamics from the conical intersection between S1 and S0 (CI_Z) to E are basically the same in both phases. Overall, the solvent effect was found to enhance the back-and-forth photo-switch efficiency between the Z and E isomers compared to the gas phase, while the quantum yields are reduced. But the solution yields of both the forward and backward photoreactions are still around 0.4. Therefore, BAB may have good photo-responsive properties if used as a photoactive unit in real systems. These results will facilitate future experimental and theoretical studies in this area to help design new azobenzene derivatives as photoactive units in biological processes, nanoscale devices, and photo-responsive materials.

State-specific heavy-atom effect on intersystem crossing processes in 2-thiothymine: a potential photodynamic therapy photosensitizer.

Thiothymidine has a potential application as a photosensitizer in cancer photodynamic therapy (PDT). As the chromophore of thiothymidine, 2-thiothymine exhibits ultrahigh quantum yield of intersystem crossing to the lowest triplet state T(1) (ca. 100%), which contrasts with the excited-state behavior of the natural thymine that dissipates excess electronic energy via ultrafast internal conversion to the ground state. In this work, we employed high-level complete-active space self-consistent field and its second-order perturbation methods to explore the photophysical mechanism of a 2-thiothymine model. We have optimized the minimum energy structures in the low-lying seven electronic states, as well as ten intersection points. On the basis of the computed potential energy profiles and spin-orbit couplings, we proposed three competitive, efficient nonadiabatic pathways to the lowest triplet state T(1) from the initially populated singlet state S(2). The suggested mechanistic scenario explains well the recent experimental phenomena. The origin responsible for the distinct photophysical behaviors between thymine and 2-thiothymine is ascribed to the heavy-atom effect, which is significantly enhanced in the latter. Additionally, this heavy-atom effect is found to be state-specific, which could in principle be used to tune the photophysics of 2-thiothymine. The present high-level electronic structure calculations also contribute to understand the working mechanism of thiothymidine in PDT.

Theoretical investigation of astacin proteolysis.

Astacin, acting as the prototype of astacin family and of the metzincin superfamily, is a mononuclear zinc enzyme that catalyzes the hydrolysis of polypeptides and proteins. In the present article, the reaction mechanism of astacin has been investigated using density functional theory (DFT) method. Using a model of the active site constructed on the basis of the X-ray crystal structure of astacin, the potential energy surface for catalytic reaction has been mapped and the transition states and intermediates along the reaction pathway are characterized. The calculations give general support to the previously proposed mechanism by experiments, which mainly involve nucleophilic attack on carbonyl group of substrate and C-N bond cleavage, and reveal more detailed mechanistic features. The Glu93 functions as a crucial general base to activate the zinc-bound water and the resulting zinc-bound hydroxide acts as the real nucleophile. It is demonstrated that there exists a "reactant region", where the interconversion between zinc-bound water and zinc-bound hydroxide occurs rapidly. The rate-limiting step is predicted to be the nucleophilic attack, which leads to an anionic gem-diolate tetrahedral intermediate. During the catalysis, zinc ion provides main catalytic power by stabilizing the developing charge of tetrahedral intermediate, while the Tyr149 residue also partially contributes to the catalysis by stabilizing the anionic intermediate. In addition, it is shown that the Cys64 plays roles in assisting in binding and orientating the substrate via electrostatic interaction.

Theoretical studies on photoisomerizations of (6-4) and Dewar photolesions in DNA.

The (6-4) photoproduct ((6-4) PP) is one of the main lesions in UV-induced DNA damage. The (6-4) PP and its valence isomer Dewar photoproduct (Dewar PP) can have a great threat of mutation and cancer but gained much less attention to date. In this study, with density functional theory (DFT) and the complete active space self-consistent field (CASSCF) methods, the photoisomerization processes between the (6-4) PP and the Dewar PP in the gas phase, the aqueous solution, and the photolyase have been carefully examined. Noticeably, the solvent effect is treated with the CASPT2//CASSCF/Amber (QM/MM) method. Our calculations show that the conical intersection (CI) points play a crucial role in the photoisomerization reaction between the (6-4) PP and the Dewar PP in the gas and the aqueous solution. The ultrafast internal conversion between the S(2) ((1)ππ*) and the S(0) states via a distorted intersection point is found to be responsible for the formation of the Dewar PP lesion at 313 nm, as observed experimentally. For the reversed isomeric process, two channels involving the "dark" excited states have been identified. In addition to the above passages, in the photolyase, a new electron-injection isomerization process as an efficient way for the photorepair of the Dewar PP is revealed.

Why iron? A spin-polarized conceptual density functional theory study on metal-binding specificity of porphyrin.

Heme is a key cofactor of hemoproteins in which porphyrin is often found to be preferentially metalated by the iron cation. In our previous work [Feng, X. T.; Yu, J. G.; Lei, M.; Fang, W. H.; Liu, S. B. J. Phys. Chem. B 2009, 113, 13381], conceptual density functional theory (CDFT) descriptors have been applied to understand the metal-binding specificity of porphyrin. We found that the iron-porphyrin complex significantly differs in many aspects from porphyrin complexes with other metal cations except Ru, for which similar behaviors for the reactivity descriptors were discovered. In this study, we employ the spin-polarized version of CDFT to investigate the reactivity for a series of (pyridine)(n)-M(ll)-porphyrin complexes-where M = Mg, Ca, Cr, Mn, Co, Ni, Cu, Zn, Ru, and Cd, and n = 0, 1, and 2-to further appreciate the metal-binding specificity of porphyrin. Both global and local descriptors were examined within this framework. We found that, within the spin resolution, not only chemical reactivity descriptors from CDFT of the iron complex are markedly different from that of other metal complexes, but we also discovered substantial differences in reactivity descriptors between Fe and Ru complexes. These results confirm that spin properties play a highly important role in physiological functions of hemoproteins. Quantitative reactivity relationships have been revealed between global and local spin-polarized reactivity descriptors. These results contribute to our better understanding of the metal binding specificity and reactivity for heme-containing enzymes and other metalloproteins alike.

Peptide hydrolysis by the binuclear zinc enzyme aminopeptidase from Aeromonas proteolytica: a density functional theory study.

Aminopeptidase from Aeromonas proteolytica (AAP) is a binuclear zinc enzyme that catalyzes the cleavage of the N-terminal amino acid residue of peptides and proteins. In this study, we used density functional methods to investigate the reaction mechanism of this enzyme. A model of the active site was constructed on the basis of the X-ray crystal structure of the native enzyme, and a model dipeptide was used as a substrate. It was concluded that the hydroxide is capable of performing a nucleophilic attack at the peptide carbonyl from its bridging position without the need to first become terminal. The two zinc ions are shown to have quite different roles. Zn2 binds the amino group of the substrate, thereby orienting it toward the nucleophile, while Zn1 stabilizes the alkoxide ion of the tetrahedral intermediate, thereby lowering the barrier for the nucleophilic attack. The rate-limiting step is suggested to be the protonation of the nitrogen of the former peptide bond, which eventually leads to the cleavage of the C-N bond.

Structures and cis-to-trans photoisomerization of hexafluoro-1,3-butadiene radical cation: electron spin resonance and computational studies.

The s-cis and s-trans isomer radical cations of hexafluoro-1,3-butadiene (s-cis-HFBD+ and s-trans-HFBD+) were generated by a gamma-irradiated solid solution of the neutral HFBD molecule in solid matrix at 77 K and observed by means of electron spin resonance (ESR) and electronic spectroscopies. In comparing the experimental isotropic and anisotropic 19F hyperfine splittings with the computational ones by the DFT B3LYP and MP2 methods, the generated s-cis-HFBD+ and s-trans-HFBD+ radical cations were confirmed to be 2A2 and 2Bg electronic ground states in C2v and C2h symmetries, respectively. The present spectroscopic study revealed that the relative abundance of s-cis-HFBD+ to s-trans-HFBD+ was 4.0 immediately after being formed by gamma-irradiation, and subsequently most s-cis-HFBD+ was isomerized to s-trans-HFBD+ by visible-light illumination with 500-600 nm wavelength. The process of nonplanar HFBD ionizing to form stable planar s-cis- and s-trans-HFBD+ and the reaction mechanism of the cis-to-trans photoisomerization were discussed by (MS-)CASPT2//CASSCF calculated vertical excitation energies (Tv) and torsional potential energy curves (TPECs) of HFBD and HFBD+.

Water-assisted transamination of glycine and formaldehyde.

A computational study on the transamination reaction of molecular complexes that consist of NH2CH2COOH + CH2O + nH2O, where n = 0, 1, 2, is presented. This work has allowed the description of the geometries of all the intermediates and transition states of the reactions, which can be described by five steps: carbinolamine formation, dehydration, 1,3 proton transfer, hydrolysis, and carbinolamine elimination. Among the five steps of the reaction, hydrolysis and elimination occur with the existence of general acid catalysis related to the carboxylic group. The water molecules can be involved in the reaction by performing as a proton-transfer carrier and a stabilizing zwitterion. It can be predicted from our calculations that in the transamination between alpha-amino acids and alpha-keto acids, the carbinolamine is formed with small barrier or even barrierless while the dehydration occurs easily at room temperature. However, without heating the 1,3 proton transfer could not occur as the barrier is 26.7 kcal/mol relative to the reactant complex when including two water molecules. Our results are in good agreement with experimental conclusions.

Probing mechanistic photochemistry of glyoxal in the gas phase by ab initio calculations of potential-energy surfaces and adiabatic and nonadiabatic rates.

In the present work, the wavelength-dependent mechanistic photochemistry of glyoxal in the gas phase has been explored by ab initio calculations of potential-energy surfaces, surface crossing points, and adiabatic and nonadiabatic rates. The CHOCHO molecules in S1 by photoexcitation at 393-440 nm mainly decay to the ground state via internal conversion, which is followed by molecular eliminations to form CO, H2CO,H2, and HCOH. Upon photodissociation of CHOCHO at 350-390 nm, intersystem crossing to T1 followed by the C-C bond cleavage is the dominant process in this wavelength range, which is responsible for the formation of the CHO radicals. The C-C and C-H bond cleavages along the S1 pathway are energetically accessible upon photodissociation of CHOCHO at 290-310 nm, which can compete with the S1-->T1 intersystem crossing process. The present study predicts that the C-H bond cleavage on the S1 surface is probably a new photolysis pathway at high excitation energy, which has not been observed experimentally. In addition, the trans-cis isomerization is predicted to occur more easily in the ground state than in the excited states.

Photodissociation of acetic acid in the gas phase: an ab initio study.

Photodissociation of acetic acid in the gas phase was investigated using ab initio molecular orbital methods. The stationary structures on the ground-state potential energy surfaces were mainly optimized at the MP2 level of theory, while those on the excited-state surfaces were determined by complete active space SCF calculations with a correlation-consistent basis set of cc-pVDZ. The reaction pathways leading to different photoproducts are characterized on the basis of the computed potential energy surfaces and surface crossing points. The calculations reproduce the experimental results well and provide additional insight into the mechanism of the ultraviolet photodissociation of acetic acid and related compounds.