The application of discriminant analysis and Machine Learning methods as tools to identify and classify compounds with potential as transdermal enhancers.

Discriminant analysis (DA) has previously been shown to allow the proposal of simple guidelines for the classification of 73 chemical enhancers of percutaneous absorption. Pugh et al. employed DA to classify such enhancers into simple categories, based on the physicochemical properties of the enhancer molecules (Pugh et al., 2005). While this approach provided a reasonable accuracy of classification it was unable to provide a consistently reliable estimate of enhancement ratio (ER, defined as the amount of hydrocortisone transferred after 24h, relative to control). Machine Learning methods, including Gaussian process (GP) regression, have recently been employed in the prediction of percutaneous absorption of exogenous chemicals (Moss et al., 2009; Lam et al., 2010; Sun et al., 2011). They have shown that they provide more accurate predictions of these phenomena. In this study several Machine Learning methods, including the K-nearest-neighbour (KNN) regression, single layer networks, radial basis function networks and the SVM classifier were applied to an enhancer dataset reported previously. The SMOTE sampling method was used to oversample chemical compounds with ER>10 in each training set in order to improve estimation of GP and KNN. Results show that models using five physicochemical descriptors exhibit better performance than those with three features. The best classification result was obtained by using the SVM method without dealing with imbalanced data. Following over-sampling, GP gives the best result. It correctly assigned 8 of the 12 "good" (ER>10) enhancers and 56 of the 59 "poor" enhancers (ER<10). Overall success rates were similar. However, the pharmaceutical advantages of the Machine Learning methods are that they can provide more accurate classification of enhancer type with fewer false-positive results and that, unlike discriminant analysis, they are able to make predictions of enhancer ability.

Gastroretentive dosage forms: overview and special case of Helicobacter pylori.

The challenge to develop efficient gastroretentive dosage forms began about 20 years ago, following the discovery of Helicobacter pylori by Warren and Marshall. In order to understand the real difficulty of increasing the gastric residence time of a dosage form, we have first summarized the important physiologic parameters, which act upon the gastric residence time. Afterwards, we have reviewed the different drug delivery systems designed until now, i.e. high-density, intragastric floating, expandable, superporous hydrogel, mucoadhesive and magnetic systems. Finally, we have focused on gastroretentive dosage forms especially designed against H. pylori, including specific targeting systems against this bacterium.

Discriminant analysis as a tool to identify compounds with potential as transdermal enhancers.

Structure-activity relationships were sought for 73 enhancers of hydrocortisone permeation from propylene glycol across hairless mouse skin. Enhancers had chain lengths (CC) from 0 to 16 carbon atoms, 1 to 8 H-bonding atoms (HB), molecular weight 60 to 450, log P (calculated) -1.7 to 9.7 and log S (calculated) -7.8 to 0.7. These predictive properties were chosen because of their ready availability. Enhancement ratio (ER) was defined as hydrocortisone transferred after 24 h relative to control. Values for the ER ranged from 0.2 to 25.3. Multiple regression analysis failed to predict activity; ER values for the 'good' enhancers (ER > 10) were underestimated. Simple guidelines suggested that high ER was associated with CC > 12 and HB 2-5. This was refined by multivariate analysis to identify significant predictors. Discriminant analysis using CC, HB, and molecular weight correctly assigned 11 of the 12 'good' enhancers (92%). The incorrectly assigned compound was a known, idiosyncratic Br compound. Seventeen of the 61 'poor' enhancers (28%) were incorrectly assigned but four could be considered marginal (ER > 8). The success of this simple approach in identifying potent enhancers suggested its potential in predicting novel enhancer activity.

Topical delivery of retinyl ascorbate co-drug. 2. Comparative skin tissue and keratin binding studies.

Retinyl ascorbate (RA-AsA), an ester co-drug of vitamins A (RA) and C (AsA), is proposed as a topical antioxidant/cell division regulator for reducing UV-induced generation of free radicals and disrupted dermal cell growth. The efficacy of dermatological agents is influenced by their retention within the skin, which is increased by the interaction with skin components. Keratin is the major protein (approximately 95%) in the skin, and this paper reports the binding of RA-AsA, RA, AsA, retinol, ascorbic acid palmitate and retinol palmitate to three tissues-human callus, pig ear skin and bovine horn keratin. Tissue samples were incubated with solutions of compounds and the uptake measured as the ratio of bound/free compound at equilibrium. Binding to keratin was assessed using delipidised tissue, and was much higher for the polar compounds, suggesting dipolar/H-bonding interaction. Binding strength was ranked as human > porcine > bovine, but there was no distinction for highly lipophilic compounds. The binding characteristic of native tissues was complicated by lipid content of the tissues. There seemed to be a dual effect. The binding of very lipophilic materials increased with lipid content, implying that a substantial amount is dissolved in the lipid matrix. For highly polar AsA, lipid content decreased the binding, suggesting that the lipid reduced the strong polar interactions with skin protein/keratin.

Receptor regulatory properties evident in the molecular similarity of serotonin receptor ligands and purine nucleotides.

Previous computational studies have explored the relative molecular similarity inherent in the ligands of neurotransmitter-regulated cell receptors and purine nucleotides. This study presents the results of an investigation of the major serotonin (5-HT) receptor classes, using molecular superimposition and fitting data. Ligands for 5HT(1B/C/D) and 5HT(4/7) receptors identified pharmacophores in the adenine ring of ATP. 5-HT(2) and 5-HT(3) receptor ligands identified pharmacophores in the guanosine nucleotide and cyclic nucleotide, respectively. The described molecular similarity is consistent with the cyclic nucleotide responses observed during signal transduction events initiated by 5-HT, and the reported similarity between ligands of the 5-HT(1B) and 5-HT(1D), 5-HT(1A) and 5-HT(7), and 5-HT(4) and 5-HT(3) receptors. The results are discussed in terms of current pharmacophoric models and signal transduction events involving interaction between G-protein receptors and catalytic sites.

Receptor regulatory properties evident in the molecular similarity of dopamine receptor ligands and purine nucleotides.

Computational studies have revealed similarities in the relative configurations of purine nucleotides and ligands for histamine, acetylcholine and adrenergic receptors. In common with other G-protein-regulated receptors, dopamine receptors are associated with specific changes in nucleotide levels during signal transduction processes. The purpose of this study was to investigate molecular similarity in dopamine receptor ligands and purine nucleotides. Molecular superimposition and fitting data for D1-like receptor ligands identified a pharmacophore in the adenine and ribose rings of ATP. D2-like agonists and antagonists related to a pharmacophore in the guanine and ribose rings of GTP. The results are consistent with the hypothesis that the dopamine receptor family may have evolved from receptors for the ATP and GTP nucleotides.

Probing the effect of vehicles on topical delivery: understanding the basic relationship between solvent and solute penetration using silicone membranes.

PURPOSE: In the present study we examined the relationship between solvent uptake into a model membrane (silicone) with the physical properties of the solvents (e.g., solubility parameter, melting point, molecular weight) and its potential predictability. We then assessed the subsequent topical penetration and retention kinetics of hydrocortisone from various solvents to define whether modifications to either solute diffusivity or partitioning were dominant in increasing permeability through solvent-modified membranes. METHODS: Membrane sorption of solvents was determined from weight differences following immersion in individual solvents, corrected for differences in density. Permeability and retention kinetics of 3H-hydrocortisone, applied as saturated solutions in the various solvents, were determined over 48 h in horizontal Franz-type glass diffusion cells. RESULTS: Solvent sorption into the membrane could be related to differences in solubility parameters, MW and hydrogen bonding (r2=0.76). The actual and predicted volume of solvent sorbed into the membrane was also found to be linearly related to Log hydrocortisone flux, with changes in both diffusivity and partitioning of hydrocortisone observed for the different solvent vehicles. CONCLUSIONS: A simple structure-based predictive model can be applied to the sorption of solvents into silicone membranes. Changes in solute diffusivity and partitioning appeared to contribute to the increased hydrocortisone flux observed with the various solvent vehicles. The application of this predictive model to the more complex skin membrane remains to be determined.

The effect of hydrogen bonding on diffusion across model membranes: consideration of the number of H-bonding groups.

The diffusion of a series of phenols across simple silicone membranes impregnated with either octanol or toluene was studied. These solvents are taken up and saturate the membrane. The presence of the solvents in a solid membrane allows them to interact with any permeant that cross the membrane. This membrane was used to simulate a bio-membrane, e.g. the skin, capable of hydrogen bonding with the permeant. As the number of H-bonding groups was increased the flux across both the octanol and toluene impregnated membranes decreased. However, deconvolution of the data showed that for the octanol impregnated membrane the diffusion coefficient (Dm) decreased significantly with the number of H-bonding groups. This was not the case for the toluene impregnated membrane. Furthermore the spatial configuration of the -OH groups around the aromatic ring had a significant effect on the decrease in Dm. These findings have considerable implications in understanding the absorption of permeants across bio-membranes capable of H-bonding.

Epidermal permeability-penetrant structure relationships: 4, QSAR of permeant diffusion across human stratum corneum in terms of molecular weight, H-bonding and electronic charge.

Principal components analysis (PCA) and multivariate regression analysis (MRA) are used to assess the predictors of permeant diffusion across human stratum corneum. Log(D/h), was estimated from logk(p)+0.024-0.59 logK(oct), where D=diffusion coefficient (cm(2)/h), h=path length (cm), k(p) permeability coefficient (cm/h), K(oct)=partition coefficient (octanol/water). Molecular weight (MW) with (1) scaled H-bonding parameters alpha and beta, or (2) summed modulus of partial charge from molecular modelling were tested as predictors of (D/h). Charge may be computed for any molecule, whilst alpha and beta values are generally unavailable for molecules of biological interest. PCA suggests a dominant permeation pathway since 93% of data variation is in PC1 of log(D/h), MW and charge and 82% in PC1 of log(D/h), MW, alpha and beta. MRA using MW, alpha and beta is unsatisfactory because of collinearity amongst predictors. The best predictor was the product MW*charge. Similarity of the eigenvectors in PCA and normalised coefficients in MRA indicates that charge and MW are equally important predictors of diffusion.

The selection and design of topical and transdermal agents: a review.

One of the major problems in topical and transdermal drug delivery is the efficiency of the barrier property of the stratum corneum. Topical and transdermal agents were often originally designed as drugs to be given orally. In this paper a mechanistic evaluation of skin penetration shows that topical and transdermal drugs should be designed using different strategies. The relative effects of basic physicochemical parameters are examined. An understanding, at a molecular level, of the permeation process will enable us to produce more effective topical agents and to extend the repertoire of transdermal drugs.

Interaction of silver nitrate with readily identifiable groups: relationship to the antibacterial action of silver ions.

Microbiologically it was demonstrated that amino acids, e.g. cysteine (CySH), and other compounds, e.g. sodium thioglycollate, containing thiol groups neutralized the activity of silver nitrate against Pseudomonas aeruginosa PAO1. Amino acids with disulphide bonds were inactive, with the exception of L-cystine dimethyl ester, as were all amino acids with no sulphur groups. Iodoacetamide reacted with CySH to produce a CyS-acetamide complex that was unable to quench the activity of Ag+. Chemical analyses using cyclic voltammetry demonstrated that high coordination numbers (3.1) were obtained with thiol-containing amino acids and low numbers (0.28-0.4) with other amino acids. Both microbiologically and chemically, the results imply that interaction of Ag+ with thiol groups plays an essential role in bacterial inactivation.

Prostaglandin synthetase inhibitors in feverfew.

The IC50 values for the in-vitro inhibition of the prostaglandin synthetase (bovine seminal vesicle mitochondrial fraction) mediated PGE2 production from arachidonic acid by parthenolide, michefuscalide and chrysanthenyl acetate were 11.0 +/- 0.44, 12.1 +/- 0.51 and 14.2 +/- 0.58 microM (mean +/- 95% confidence limits), respectively.

Preparation and characterisation of the C(10) to C(18) even-numbered triethanolamine alkyl sulphates.

Synopsis The preparation and characterisation of purified C(10) to C(18) even-numbered triethano-lamine alkyl sulphates are described. Critical micelle concentrations were determined at 25 degrees C, using both conductance and drop volume methods. The authenticity and purity of the products are established, and evidence presented for amending the accepted melting points for two of the compounds.

A relationship between foaming and H.L.B. value.

Synopsis The foaming properties of solutions of purified samples of the C(10) to C(18) even-numbered sodium alkyl sulphates in water have been studied using three methods of foam generation. Foam volumes increased with increase in surfactant concentration, and reached a limiting value at the C.M.C. At post-C.M.C. concentrations, foam volumes increased to a maximum with temperature, and then decreased. The point at which the maximum occurred, the maximum foam temperature (M.F.T.), varied from one surfactant to another, but was independent of the experimental method. M.F.T. was inversely proportional to the hydrophile-lipophile balance (H.L.B.) value. The corresponding triethanolamine salts behaved in the same way. Similar results were observed with non-ionics, but the plots of H.L.B. value against M.F.T. were positive.