Quantum theory of atoms in molecules (QT-AIM) allows detailed insight into the electronic structure of molecules by analysis of the gradient vector field of the electronic density distribution function. First results of a QT-AIM analysis of neutral pterin as well as its anionic and cationic forms in aqueous solution are reported based on density functional theory using B3LYP/6–311+G(2d,p)//6–31G(d) level of theory. Besides reporting QT-AIM results of the atomic partial charges and bond orders of the molecules, their electron density functions, Laplacians and electrostatic potential functions are also shown. The results demonstrate the rather extensive delocalization of both negative and positive extra charges predominantly over the pyrimidine moiety of the pterin ring system and allow a precise quantitation of the effects of addition or elimination of a proton on the bonding structure. Quantum chemical techniques in combination with QT-AIM procedures are thus well-suited to describing the bonding structure as well as the major atomic and bonding features of neutral pterin and its cationic and anionic forms. Moreover, the data demonstrate that the chosen level of theory yields chemically reasonable results while not being computationally too expensive, thus providing a sound basis for further theoretical chemical work on pteridines.
The electronic structures of the five radicals resulting from homolytic elimination of one of the hydrogen atoms from the most stable tautomeric form of neutral pterin were investigated in gas phase as well as in aqueous solution. Molecular wave functions obtained by density functional theory were analysed by quantum theory of atoms in molecules and electron localisation functions (ELF). Spin densities of the radicals as well as electrostatic potential functions were analysed. Radicals resulting from elimination of N-bonded hydrogen atoms are more stable in comparison with radicals obtained after abstraction of C-bonded hydrogen atoms. N-centred radicals show strong delocalisation of spin density over both heteroaromatic rings; in C-centred radicals delocalisation does not occur. ELF analyses showed that in N-derived radicals particularly the lone electron pair at N2′ is strongly involved into the bicyclic heteroaromatic π-electron system. Thereby, bonding geometry at N2′ in these radicals changes from pyramidal to planar. Transition from gas phase to solution phase (water) generally leads to increased polarity of the structures. Pterin-derived free radicals have been implicated in several biologically important reactions; so this investigation provides first insights into the detailed electronic structures of such molecular systems.
The influence of dihydroneopterin on tyrosine hydroxylase activity in PC 12 cells was investigated under normoxic and hyperoxic conditions, and with or without iron ions in the incubation medium. Low dihydroneopterin concentrations as well as hyperoxia increase tyrosine hydroxylase activity. Due to a too vigorous radical formation, high dihydroneopterin concentrations lead to a decrease of tyrosine hydroxylation in normoxic state and even more strongly in hyperoxic state. Similar depression of tyrosine hydroxylase activity was seen after addition of iron ions, which can form hydroxyl radicals by a Fenton type reaction. As higher iron concentrations in combination with dihydroneopterin do not suppress tyrosine hydroxylase activity completely, we conclude that dihydroneopterin and iron ions react with each other resulting in neutralisation of their effects. In conclusion, 7,8 dihydroneopterin can modulate tyrosine hydroxylase activity and its effects are dependent on concentration of oxygen as well as presence or absence of iron ions.
Since the early nineties the number of scientific papers reporting on artificial neural network (ANN) applications in medicine has been quickly increasing. In the present paper, we describe in some detail the architecture of network types used most frequently in ANN applications in the broad field of laboratory medicine and clinical chemistry, present a technique-structured review about the recent ANN applications in the field, and give information about the improvements of available ANN software packages.
ANN applications are divided into two main classes: supervised and unsupervised methods. Most of the described supervised applications belong to the fields of medical diagnosis (n=7) and outcome prediction (n=9). Laboratory and clinical data are presented to multilayer feed-forward ANNs which are trained by the back propagation algorithm. Results are often better than those of traditional techniques such as linear discriminant analysis, classification and regression trees (CART), Cox regression analysis, logistic regression, clinical judgement or expert systems. Unsupervised ANN applications provide the ability of reducing the dimensionality of a dataset. Low-dimensional plots can be generated and visually understood and compared. Results are very similar to that of cluster analysis and factor analysis. The ability of Kohonen's self-organizing maps to generate 2D maps of molecule surface properties was successfully applied in drug design.
The quality of wavefunctions for complex systems derived by either “full” ab initio MO-SCF calculations or within the MESQUAC-MO framework  is investigated comparing chemically interesting quantities as atomic and overlap populations and the character of the chemical bond. A direct relation between the results of both methods is shown to exist, allowing an extrapolation of the much less time consuming MESQUAC computations to the “full” MO SCF level. Hydrate complexes of main group and transition metal ions have been chosen for some practical applications
In order to investigate the stabilities of different tautomeric forms of pteridine derivatives with a phenacetyl side chain in the 6- or 7-position, we have performed ab initio quantum chemical geometry optimizations of both the keto form and the vinylogous amide form of 6- and 7-phenacetyl pterin. The results are in accordance with experimental expectations: the keto form is slightly more stable for the 6-substituted derivative (1 .8 kcal per mol). while the vinylogous amide form is substantially more stable for the 7-substituted compound (8 .7 kcal per mol). The vinylogous amide forms are planar molecules, and this geometry enables the interaction between the hydrogen atom at N-5 (or N-8) and the carbonyl function of the side chain. Inspection of the calculated electron densities (population analysis) underlines the suggestion that the vinylogous amide form of 7-phenacetyl pterin is stabilized by a shift of electron density from the nitrogen bonded to C-2 to the carbonyl oxygen atom of the side chain (resonance stabilization via an additional highly conjugated vinylogous amide structure).
Recently; the 4-amino analogue of tetrahydrobiopterin was found to be a strong inhibitor of nitric oxide synthase while being bound to the enzyme in a manner similar to the natural cofactor tetrahydrobiopterin. We were interested in the electronic properties of these and similar compounds and studied therefore the following model tetrahydropteridine structures: tetrahydrolumazine, tetrahydropterin, 4-amino-analogue of tetrahydropterin and N5-methyl-tetrahydropterin. Ab initio quantum chemical computations used the Hartree- Fode method with basis set 6-31G** after geometry optimization with basis set 3-21G*. Results reveal dramatic differences in distribution of electronic charge and all the molecular properties derived thereof~ between a) the lumazine system, b) the normal pterin system, and c) the 4-amino analogue. In contrast, differences of electronic properties between tetrahydropterin and its N5-methyl-derivative are negligible. Our results are compatible with recent speculations that the striking differences between the effects of the tetrahydropterin struculrc and its 4-amino analogue on cnznllatic activity may be due to electronic interaction between the pyrimidine moiety of the ptcrin ring systcm and the heme group.