Modeling bovine serum albumin binding of flavor compounds (alcohols, aldehydes, esters, and ketones) as a function of molecular properties.

Interactions between bovine serum albumin (BSA) and 4 classes of flavor compounds (alcohols, esters, aldehydes, and ketones) in aqueous solution were investigated using solid phase microextraction (SPME) of sample headspace and gas chromatography. Alcohols did not bind significantly to BSA. The binding of fixed amounts of individual flavor compounds with each functional group to increasing amounts of BSA was modeled as a function of several descriptors using partial least squares regression (PLSR). Ester binding was modeled as a function of the number of carbon atoms in a molecule and its boiling point (R(2)= 0.954). Aldehyde binding was modeled as a function of the number of hydrogen atoms and boiling point (R(2)= 0.922). Ketone binding was modeled as a function of the numbers of carbon atoms, length of the longer hydrocarbon chain, and degree of branching (R(2)= 0.961).

Chemometric analysis of proteolysis during ripening of Ragusano cheese.

Chemometric modeling of peptide and free amino acid data was used to study proteolysis in Protected Denomination of Origin Ragusano cheese. Twelve cheeses ripened 3 to 7 mo were selected from local farmers and were analyzed in 4 layers: rind, external, middle, and internal. Proteolysis was significantly affected by cheese layer and age. Significant increases in nitrogen soluble in pH 4.6 acetate buffer and 12% trichloroacetic acid were found from rind to core and throughout ripening. Patterns of proteolysis by urea-PAGE showed that rind-to-core and age-related gradients of moisture and salt contents influenced coagulant and plasmin activities, as reflected in varying rates of hydrolysis of the caseins. Analysis of significant intercorrelations among chemical parameters revealed that moisture, more than salt content, had the largest single influence on rates of proteolysis. Lower levels of 70% ethanol-insoluble peptides coupled to higher levels of 70% ethanol-soluble peptides were found by reversed phase-HPLC in the innermost cheese layers and as the cheeses aged. Non-significant increases of individual free amino acids were found with cheese age and layer. Total free amino acids ranged from 14.3 mg/g (6.2% of total protein) at 3 mo to 22.0 mg/g (8.4% of total protein) after 7 mo. Glutamic acid had the largest concentration in all samples at each time and, jointly with lysine and leucine, accounted for 48% of total free amino acids. Principal components analysis and hierarchical cluster analysis of the data from reversed phase-HPLC chromatograms and free amino acids analysis showed that the peptide profiles were more useful in differentiating Ragusano cheese by age and farm origin than the amino acid data. Combining free amino acid and peptide data resulted in the best partial least squares regression model (R(2) = 0.976; Q(2) = 0.952) predicting cheese age, even though the peptide data alone led to a similarly precise prediction (R(2) = 0.961; Q(2) = 0.923). The most important predictors of age were soluble and insoluble peptides with medium hydrophobicity. The combined peptide data set also resulted in a 100% correct classification by partial least squares discriminant analysis of cheeses according to age and farm origin. Hydrophobic peptides were again discriminatory for distinguishing among sample classes in both cases.

Chemometric analysis of Ragusano cheese flavor.

Ragusano cheeses were produced in duplicate from milk collected from pasture-fed and total mixed ration (TMR)-fed cattle at four time intervals. The cheeses were subjected to chemical analysis, conventional sensory testing, and gas chromatography-olfactometry (GCO). Data from each type of analysis were examined by principal component and factor analysis and by pattern recognition (SIMCA) to see if sufficient information for classification into pasture-fed and TMR-fed groups was contained therein. The results clearly indicate that there are significant differences in sensory panel and chemical analysis results between the two cheeses. The data were also examined to see if models of sensory responses as a function of analytical or GCO results or both could be constructed with the modeling technique partial least-squares regression (PLS). Strong PLS models of some sensory responses (green and toasted odor; salt, pungent, bitter, and butyric sensations; and smooth consistency) were obtained.

Comparison of pattern recognition techniques for the identification of lactic acid bacteria.

AIMS: The goal of this study was to evaluate three pattern recognition methods for use in the identification of lactic acid bacteria. METHODS AND RESULTS: Lactic acid bacteria (21 unknown isolates and 30 well-characterized strains), including the Lactobacillus, Lactococcus, Streptococcus, Pediococcus and Oenococcus genera, were tested for 49 phenotypic responses (acid production on carbon sources). The results were scored in several ways. Three procedures, k-nearest neighbour analysis (KNN), k-means clustering and fuzzy c-means clustering (FCM), were applied to the data. CONCLUSION: k-Nearest neighbour analysis performed better with five-point-scaled than with binary data, indicating that intermediate values are helpful to classification. k-Means clustering performed slightly better than KNN and was best with fuzzified data. The best overall results were obtained with FCM. Genus level classification was best with FCM using an exponent of 1.25. SIGNIFICANCE AND IMPACT OF THE STUDY: The three pattern recognition methods offer some advantages over other approaches to organism classification.

Quantitative structure-activity relationship modeling of peptide and protein behavior as a function of amino acid composition.

A quantitative structure-activity relationship (QSAR) modeling approach based on the location of each amino acid along three axes obtained by principal component analysis (called z scores) was extended to physical and functional properties of proteins, where the proportion of particular amino acids rather than a precise sequence is the determining factor. Coomassie Brilliant Blue spectral responses to amino acid homopolymers (R = 0.926) and proteins, either as a function of their contents of six basic and aromatic amino acids (R = 0.976) or as a function of the contributions of these amino acids to the three z scores (R = 0.935), were modeled. The ultraviolet absorbance of proteins was modeled in terms of the z score contributions of tyrosine, tryptophan, and cysteine (R = 0.995). Modeling many protein functional properties in this manner appears to be possible. An approach to modeling peptide behaviors that depend on short sequences of amino acids was also considered.

Effects of protein-polyphenol interactions on beverage haze, stabilization, and analysis.

The haze-forming activity of a polypeptide depends greatly on its proline content. Haze-forming polyphenols have at least two binding groups, each of which has at least two hydroxy groups on an aromatic ring. The protein/polyphenol ratio has a strong influence on the amount of haze formed; the largest amount occurs when the numbers of polyphenol binding ends and protein binding sites are nearly equal. This has important consequences for turbidimetric methods used to measure haze-active proteins and polyphenols in beverages. The ratio also influences the effectiveness of a number of stabilization procedures.

Modeling the inhibitory effects of organic acids on bacteria.

The inhibitory effect of acids on microbial growth has long been used to preserve foods from spoilage. While much of the effect can be accounted for by pH, it is well known that different organic acids vary considerably in their inhibitory effects. Because organic acids are not members of a homologous series, but vary in the numbers of carboxy groups, hydroxy groups and carbon-carbon double bonds in the molecule, it has typically not been possible to predict the magnitude, or in some cases even the direction, of the change in inhibitory effect upon substituting one acid for another or to predict the net result in food systems containing more than one acid. The objective of this investigation was to attempt to construct a mathematical model that would enable such prediction as a function of the physical and chemical properties of organic acids. Principal Components Analysis (PCA) was applied to 11 properties for each of 17 acids commonly found in food systems; this resulted in four significant principal components (PCs), presumably representing fundamental properties of the acids and indicating each acid's location along each of these four scales. These properties correspond to polar groups, the number of double bonds, molecular size, and solubility in non-polar solvents. Minimum inhibitory concentrations (MICs) for each of eight acids for six test microorganisms were determined at pH 5.25. The MICs for each organism were modeled as a function of the four PCs using partial least squares (PLS) regression. This produced models with high correlations for five of the bacteria (R2 = 0.856, 0.941, 0.968, 0.968 and 0.970) and one with a slightly lower value (R2 = 0.785). Acid susceptible organisms (Bacillus cereus, Bacillus subtilis, and Alicyclobacillus) exhibited a similar response pattern. There appeared to be two separate response patterns for acid resistant organisms; one was exhibited by the two lactobacilli studied and the other by E. coli. Predicting the inhibitory effects of the organic acids as a function of their chemical and physical properties is clearly possible.

All-trans retinoic acid stimulates growth and extracellular matrix production in growth-inhibited cultured human skin fibroblasts.

All-trans retinoic acid was examined for effects on human dermal fibroblast proliferation and for effects on fibroblast production and expression of non-collagenous and collagenous components of the extracellular matrix in vitro. Fibroblast proliferation was blocked when the cells were cultured in the presence of a serum-free culture medium containing epidermal growth factor, hydrocortisone, insulin, ethanolamine, phosphoethanolamine, and bovine pituitary extract as growth supplements and 0.15 mM Ca++. This level of extracellular Ca++ is lower than that needed to support fibroblast growth. Under these conditions, growth was stimulated by all-trans retinoic acid. Proliferation was also stimulated in the same basal medium without the growth supplements. Growth-promoting concentrations of all-trans retinoic acid ranged from 0.5-2.0 micrograms/ml (1.7-6.6 X 10(-6) M). Stimulation of proliferation was not seen at higher or lower concentrations. Concentrations of all-trans retinoic acid that stimulated proliferation also induced increased production of fibronectin as indicated by biosynthetic labeling/immunoprecipitation and by enzyme-linked immunosorbent assay. Increased production was associated with increased staining for fibronectin in the extracellular matrix. Increased production of two other non-collagenous extracellular matrix component, i.e., thrombospondin and laminin, also occurred in all-trans retinoic acid-treated cells. At 0.5 micrograms/ml, all-trans retinoic acid also stimulated production of type I collagen by the dermal fibroblasts, but at higher concentrations (2.5 micrograms/ml) production of type I collagen was inhibited. These data indicate that all-trans retinoic acid can induce changes in dermal fibroblasts in vitro (i.e., increased proliferation and extracellular matrix production) that mimic the major changes seen in the dermis after topical treatment with this agent.