摘要:

The thermal stability of S-nitrosothiols experimental studies and ab initio calculations on model compounds Nicola Bainbrigge,a Anthony R. Butler a and Carl Henrik Görbitz *,b a b School of Chemistry University of St. Andrews Fife KY16 9ST UK Department of Chemistry University of Oslo N-0315 Norway One factor responsible for the enhanced thermal stability of S-nitroso-N-acetylpenicillamine (SNAP) compared with S-nitroso-N-acetylcysteine (SNAC) has been shown by ab initio calculations on model compounds to be steric interactions in the dimerisation reaction leading to disulfide formation. Studies using DSC and TGA indicate that the two gem methyl groups in SNAP do not have a substantial effect on the strength of the ]S]NO bond. SNO CO2H Introduction One of the many consequences of the discovery of the roles of nitric oxide (NO) in animal physiology 1 has been heightened interest in NO derivatives particularly S-nitrosothiols (RSNO) which decompose2 with formation of NO and a disulfide.S- 2RSNO æÆ RS]SR + 2NO Nitrosocysteine was proposed3 as an alternative to NO during early attempts to identify the endothelium-derived relaxing factor (EDRF).4 This view has recently been refuted.5 The formation of S-nitrosothiols by reaction of nitrous acid with the appropriate thiol is readily detected spectroscopically but most S-nitrosothiols are far too reactive to isolate and characterise.6 There are two notable exceptions to this generalisation S-nitroso-N-acetylpenicillamine (SNAP) and S-nitro- SNO CH3COHN CO2H SNAC Me Me CH3COHN SNAP soglutathione (GSNO).The former can be readily obtained as an analytically pure pink solid which is stable enough for the determination of its structure by X-ray crystallography.7 In contrast P26/04691F/B1 S-nitroso-N-acetylcysteine (SNAC) appears to have a halflife of only seconds in aqueous solution.8 The difference in stability between SNAP and SNAC in view of their similar chemical structures has been a matter of interest and speculation. Its origin has been ascribed to the two gem methyl groups in SNAP and we report now an attempt to understand the effect. Before continuing it is necessary to describe a complication which occurs when S-nitrosothiols are in aqueous solution. Our initial efforts to study the kinetics of S-nitrosothiol decomposition in solution in which we compared inter alia the rates of reaction of SNAC prepared in situ and SNAP were frustrated by erratic and irreproducible results.The explanation was that copper ions are very powerful catalysts of S-nitrosothiol decomposition and even the concentration of copper ions in good distilled water is sufficient to effect reaction.9 A further complication appeared when it was confirmed that copper(I) ions rather than the more readily available copper(II) ions are the effective catalyst. In an extensive study of this effect 10 we proposed that copper(I) ions act by forming a complex with SNAP or SNAC from which NO is readily lost. Calculations to elucidate the sites of Cu+ complexation in S-nitrosothiol are now in progress.In the copper-catalysed reaction the effect of the two gem methyl groups could be merely to lower the formation constant of the reactive intermediate complex and thus give to SNAP an enhanced stability. However careful examination of the results reported in refs. 8 and 10 indicate that in addition to the copper-catalysed reaction there is thermal decomposition and that SNAP is more stable than SNAC also in the latter pathway to NO release which becomes dominant at very low copper concentrations (1028 M). There is EPR evidence 11 that thiyl radicals are formed during S-nitrosothiol decomposition and so the mechanism must be a two-step process eqns. (1) and (2). The two gem methyl groups RSNO æÆ RS? + NO 2RS? æÆ RS]SR could affect either of these two steps (1) by increasing the strength of the ]S]NO bond or (2) by influencing the ability of the two thiyl radicals to dimerise.We have addressed both these possibilities (1) through studies of S-nitrosothiol decomposition by differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA) and (2) by comparing the amounts of steric conflict in the disulfide products formed from dimerisation of cysteinyl and penicillamyl radicals. In previous ab initio calculations 12 diethyl disulfide (CH3–CH2–S–S–CH2– CH3 ESSE) was used as a model compound for the L-cystine disulfide bridge. It was shown that the molecule has six different conformational minima with energies varying within 6.9 kJ mol21 due to small inherent differences in the electronic stabilities as well as to various amounts of steric conflict between the terminal methyl groups.The scope of the ab initio calculations presented here was to establish the level of steric strain in the penicillamine disulfide through calculations using di(tert-butyl) disulfide (tBSStB) as a model compound. (1) (2) Experimental SNAP4 and GSNO13 were prepared by literature methods. The DSC was performed on a Perkin-Elmer DSC 7 and the TGA on a TA Instruments SDT 2960. All an initio optimisations were carried with the GAUSSIAN94 program14 at the MP2/6-31G*//MP2/6-31G* level of theory with inclusion of second-order electron correlation effects. Previous calculations on ESSE12 and ethyl hydrodisulfide (CH3CH2SSH ESSH)15 indicated that this is an adequate choice for studies of energy minima for the selected compounds.The minimisation of the fully relaxed tBSStB J. Chem. Soc. Perkin Trans. 2 1997 351 structure required 49 h of CPU-time on a DEC Alpha 3000/900 computer and the use of still larger basis sets was not feasible. The amount of steric strain of a particular molecular conformation is normally estimated by calculating first the energy (absolute or relative) of a conformation with negligible steric interactions and then the energy of the strained conformer interpreting the difference as a measure of steric conflict. One thus ignores differences in the inherent electronic stabilities which may in many cases be a good approximation.For tBSStB this technique is not possible since the molecule has just one single minimum conformation. Hence we decided to obtain an ab initio estimate in an indirect manner. Calculations for two ESSE minima were also included for comparison with previous results. These are the absolute energy minimum with C]C]S]S and C]S]S]C torsion angles all gauche+ (code GGG) and the least favourable minimum with the two C]C]S]S torsion angles gauche2 and C]S]S]C gauche+ (code G9GG9). The energetic costs of the molecular deformations observed in the disulfide dimers have been estimated by considering for each model compound two different molecular fragments. Fragment 1 X–S–S–H X = tert-butyl or ethyl. Used to study deformation of the covalent geometry of the alkyl group.The molecular geometry of the alkyl group was kept fixed as obtained in the tBSStB and ESSE disulfides with only the ]S]S] ]S]H bond lengths the ]C]S]S] and ]S]S]H bond angles and the ]C]S]S]H torsion angle free to refine. The energies were then compared with those of the fully optimized X]S]S]H structures giving for each the deformation energy D1. Fragment 2 CH3–S–S–CH3 (dimethyl disulfide MSSM). The ]S]S] bond lengths the ]C]S]S] bond angles and the ]C]S]S]C] torsion angles were fixed as in the respective disulfide with other parameters free to refine. The energies were then compared with those of the fully optimised CH3–S–S–CH3 structure giving for each the deformation energy D2. When considering molecular deformation energies we refrained from using further single point calculations since it is our experience that application of other (larger) basis sets than the one used when obtaining the minimum structure invariably leads to an overestimation of the associated energy penalties.In addition to direct structural modifications of the disulfides studied there is also a contribution to the total energy from through-space van der Waals’ contacts between terminal methyl groups in ESSE (G9GG9) and tBSStB. The associated interaction energies were estimated by in each case deleting all other S C and H atoms except the (CH3)C atoms which were transformed into H-atoms. This procedure generated a methane dimer shown for ESSE (G9GG9) in Scheme 1 for which the C C C C C C S S Scheme 1 interaction energy was obtained from single point calculations P26/04691F/A1 at the MP2/6-311++G(2d,p)//MP2/6-31G* level corrected for basis set superposition error by the full counterpoise correction method of Boys and Bernardi.16 A similar method has been used for ESSE (GGG) which has a 2.63 Å H? ? ? H contact between ethylene groups.In this case all atoms other than the ethylene C-atoms were deleted the Catoms being transformed to H-atoms leaving a H2 dimer. This gives a rather gross approximation of the van der Waals’ energy since the H]H bond is not as polar as a C]H bond and no 352 J. Chem. Soc. Perkin Trans. 2 1997 Table 1 Deformation energies and van der Waals’ repulsion (vdW/kJ mol21) in ESSE and tBSStB structures SUMa vdW Molecule (conformation) D1 0.09 0.42 0.52 ESSE(GGG) ESSE(G9GG9) tBSStB 1.38 8.99 12.68 1.06 1.62 2.37 a 2 × D1 + D2 + vdW.Fig. 1 MP2/6-31G* molecular geometry of the C2-symmetric tBSStB energy minimum structure with atomic numbering bond lengths (Å) and bond angles (8). C1 is gauche+ to the SS-bridge C2 is trans and C3 is gauche2. D2 0.14 6.52 9.27 atoms carry partial charges but this should not be a serious problem since this is a weak interaction with a small energy contribution. – 2 – O2C +H3N Results and discussion The complications induced by the Cu+ catalysed pathway can be avoided if we look at the solid state decomposition of S-nitrosothiols.However SNAC cannot be obtained as a solid and so we chose the nearest stable S-nitrosothiol of similar structure which is GSNO. In view of the greater O SNO H N CO N H O GSNO complexity of this molecule results must be used with some caution. P26/04691K/B2 When SNAP was subjected to DSC it was found to decompose very cleanly at 148 8C. Use of TGA showed that the change at 148 8C is consistent with loss of NO. Clearly SNAP is a rather stable substance and the general insistence that it should be stored refrigerated is not well founded. With GSNO the results of DSC were a little more difficult to interpret. There is a general drift in the base line which we now ascribe to loss of moisture but a much sharper heat absorption occurred at 148 8C.Study by TGA confirmed that this change is again consistent with loss of NO. These data suggest that the two gem methyl groups have little or no effect on the strength of the ]S]NO bond and that the enhanced thermal stability of SNAP in solution may reside in radical dimerisation. Results from the ab initio calculations are given in Table 1,† with molecular geometry for tBSStB indicated in Fig. 1. Due to the C2 symmetry of the molecules considered the total energy of deformation Ed is then calculated as Ed = 2 D1 + D2 + vdW where vdW is the van der Waals’ repulsion. The calculated steric hindrance for tBSStB is 12.7 kJ mol21. In comparison the steric hindrance of the ESSE (GGG) minimum is very moderate at 1.4 kJ mol21 which is 7.6 kJ mol21 † Complete listings of molecular geometries and absolute energies of all molecular fragment studied are available from the authors on request.Table 2 MP2/6-31G* molecular geometry (Å,8) for various disulfides C]S]S]C/H C]C]S]S C]S]S S]S Molecule (conformation) 102.1 102.5 102.3 104.5 2.054 2.063 2.063 2.061 85.1 89.2 a 89.2 a 89.7 MSSM ESSH (GG) ESSH (G9G) tBSSH ESSE (GGG) ESSE (G9GG9) tBSStB 87.2 b 111.4 b 113.3 102.3 103.4 105.1 2.056 2.064 2.061 — 65.7 269.4 61.0 179.3 262.4 68.3 270.9 61.0 178.6 262.4 a From ref. 15. b From ref. 12. less than the G9GG9. From calculations on ESSH it has been shown that with positive disulfide chirality a gauche2 rotamer for the C]C]S]S torsion is inherently more stable than the gauche+ rotamer by 0.34 kJ mol21.With two such torsions in ESSE the G9GG9 conformation is tentatively 0.68 kJ mol21 more stable than the GGG conformation in the absence of steric conflict. Combining these figures one arrives at a 6.9 kJ mol21 estimate for the conformational energy difference between ESSE in GGG and G9GG9 conformations. The actual energy difference was calculated to 7.3 kJ mol21 at this level of theory. These values are surely sufficiently close to lend credit to the procedure used for calculating steric conflict and confidence to the 12.7 kJ mol21 estimate for tBSStB. It is obvious from Table 1 that the strain in tBSStB and ESSE (G9GG9) is relieved mainly at the central disulfide bond.Data in Table 2 show that while the S]S bond lengths and C]S]S bond angles change little with dimerisation the C]S]S]C torsion angles undergo major shifts from 87.28 in ESSE (GGG) to 111.48 in ESSE (G9GG9). In ESSE the C]C]S]S torsion angles deviate slightly from the ideal staggered positions but the tertbutyl groups in tBSStB are almost perfectly staggered forcing a further opening of the C]S]S]C torsion angle to 113.38. This theoretical value is close to the 113.28 mean value for the C]S]S]C torsion angles of penicillamine disulfide bridges in three crystal structures meso-penicillamine disulfide dihydrate 17 = 119.48 D-penicillamine disulfide dihydrochloride 18 = 114.78 and [D-Pen2 D-Pen5]enkephalin 19 = 105.68 (average of three molecules with closely related conformations in the asymmetric unit).Conclusions When a disulfide bridge is formed from cysteine residues it can adopt a number of conformations. For some of these steric conflict from close H? ? ? H contacts is negligible. A disulfide bridge formed from penicillamine on the other hand is inevitably forced into a high energy sterically congested conformation in which the C]S]S]C torsion angle has been opened ca. 258 from the values observed in sterically unstrained molecules. This observation may be important if in the solution decomposition of an S-nitrosothiol step (1) above is an equilibrium (19) RS? + NO RSNO (19) and the rate-determining process is thiyl radical dimerisation (2).In solution in contrast to the situation in the solid (2) 2RS? æÆ RS]SR state NO and the thiyl radical may recombine and decomposition occurs only with thiyl radical dimerisation. This study gives insight into one factor to be taken into account in designing S-nitrosothiols as NO-donor drugs when thermal stability is a matter of importance. Already GSNO has been used clinically to inhibit platelet aggregation during coronary angioplasty,20 and other S-nitrosothiols are currently under scrutiny. Paper 6/04148H References 1 A. R. Butler and D. L. H. Williams Chem. Soc. Rev. 1993 22 233. 2 D. L. H. Williams Chem. Soc. Rev. 1985 14 171. 3 P. R. Myers R. L. Minor R. Guerre J. N. Bates and D. G. Harrison Nature 1990 345 161. 4 R.F. Furchgott and J. V. Zawadzki Nature 1980 288 373. 5 M. Feelisch M. te Poel R. Zamora A. Deussen and S. Moncada Nature 1994 368 62. 6 B. Roy A. du Moulinet d9Hardemare and M. Fontecave J. Org. Chem 1994 59 7019. 7 L. Field R. V. Dilts R. Ravichandran P. G. Lenhert and G. E. Carnahan J. Chem. Soc. Chem. Commun. 1978 249. 8 S. C. Askew D. J. Barnett J. McAninly and D. L. H. Williams J. Chem. Soc. Perkin Trans. 2 1995 741. 9 J. McAninly D. L. H. Williams S. C. Askew A. R. Butler and C. Russell J. Chem. Soc. Chem. Commun. 1993 1758. 10 A. P. Dicks H. R. Swift D. L. H. Williams A. R. Butler H. H. Al- Sa9doni and B. G. Cox J. Chem. Soc. Perkin. Trans. 2 1996 481. 11 P. D. Josephy D. Rehorek and E. G. Janzen Tetrahedron Lett. 1984 25 1685. 12 C.H. Görbitz J. Phys. Org. Chem. 1994 7 259. 13 T. W. Hart Tetrahedron Lett. 1985 26 2013. 14 GAUSSIAN94 Revision B.3 M. J. Frisch G. W. Trucks H. B. Schlegel P. M. W. Gill B. G. Johnson M. A. Robb J. R. Cheeseman T. Keith G. A. Petersson J. A. Montgomery K. Raghavachari M. A. Al-Laham V. G. Zakrzewski J. V. Ortiz J. B. Foresman C. Y. Peng P. Y. Ayala W. Chen M. W. Wong J. L. Andres E. S. Replogle R. Gomperts R. L. Martin D. J. Fox J. S. Binkley D. J. Defrees J. Baker J. P. Stewart M. Head-Gordon C. Gonzalez and J. A. Pople Gaussian Inc. Pittsburgh PA 1995. 15 C.H. Görbitz J. Phys. Org. Chem. 1993 6 615. 16 S. F. Boys and F. Bernardi Mol. Phys. 1970 19 553. 17 L. G. Warner T. Ottersen and K. Seff Acta Crystallogr. Sect. B 1974 30 1077. 18 R. E. Rosenfield Jr and R. Parthasarathy Acta Crystallogr. Sect. B 1975 31 462. 19 J. L. Flippen-Anderson V. J. Hruby N. Collins C. George and B. Cudney J. Am. Chem. Soc. 1994 116 7523. 20 E. J. Langford A. S. Brown R. J. Wainwright A. J. de Belder M. R. Thomas R. E. A. Smith W. R. Radomski J. F. Martin and S. Moncada Lancet 1994 344 1458. Received 13th June 1996 Accepted 24th September 1996 J. Chem. Soc. Perkin Trans. 2 1997 353

ISSN:1472-779X    

DOI:10.1039/a604148e

出版商:RSC    

年代:1997

数据来源: RSC