Protein determination by microchip capillary electrophoresis using an asymmetric squarylium dye: noncovalent labeling and nonequilibrium measurement of association constants.

In response to a growing interest in the use of smaller, faster microchip (mu-chip) methods for the separation of proteins, advancements are proposed that employ the asymmetric squarylium dye Red-1c as a noncovalent label in mu-chip CE separations. This work compares on-column and precolumn labeling methods for the proteins BSA, beta-lactoglobulin B (beta-LB), and alpha-lactalbumin (alpha-LA). Nonequilibrium CE of equilibrium mixtures (NECEEM) represents an efficient method to determine equilibrium parameters associated with the formation of intermolecular complexes, such as those formed between the dye and proteins in this work, and it allows for the use of weak affinity probes in protein quantitation. In particular, nonequilibrium methods employing both mu-chip and conventional CE systems were implemented to determine association constants governing the formation of noncovalent complexes of the red luminescent squarylium dye Red-1c with BSA and beta-LB. By our mu-chip NECEEM method, the association constants K(assoc) for beta-LB and BSA complexes with Red-1c were found to be 3.53 x 10(3) and 1.65 x 10(5) M(-1), respectively, whereas association constants found by our conventional CE-LIF NECEEM method for these same protein-dye systems were some ten times higher. Despite discrepancies between the two methods, both confirmed the preferential interaction of Red-1c with BSA. In addition, the effect of protein concentration on measured association constant was assessed by conventional CE methods. Although a small decrease in K(assoc) was observed with the increase in protein concentration, our studies indicate that absolute protein concentration may affect the equilibrium determination less than the relative concentration of protein-to-dye.

Protein labeling with red squarylium dyes for analysis by capillary electrophoresis with laser-induced fluorescence detection.

Two new red luminescent asymmetric squarylium dyes (designated "Red-1c and Red-3") have been shown to exhibit absorbance shifts to longer wavelengths upon the addition of protein, along with a concomitant increase in fluorescence emission. Specifically, the absorbance maxima for Red-1c and Red-3 dyes are 607 and 622 nm, respectively, in the absence of HSA, and 642 and 640 nm in the presence of HSA, making the excitation of their protein complexes feasible with inexpensive and robust diode lasers. Fluorescence emission maxima, in the presence of HSA, are 656 and 644 nm for Red-1c and Red-3, respectively. Because of the inherently low fluorescence of the dyes in their free state, Red-1c and Red-3 were used as on-column labels (that is, with the dye incorporated into the separation buffer), thus eliminating the need for sample derivatization prior to injection and separation. A comparison of precolumn and on-column labeling of proteins with these squarylium dyes revealed higher efficiencies and greater sensitivities for on-column labeling, which, when conducted with a basic, high-salt content buffer, permitted baseline resolution of a mixture of five model proteins. LOD for model proteins, such as transferrin, alpha-lactalbumin, BSA, and beta-lactoglobulin A and B, labeled with these dyes and analyzed by CE with LIF detection (CE-LIF) were found to be dependent upon dye concentration and solution pH, and are as low as 5 nM for BSA. Satisfactory linear relationships between peak height (or peak area) and protein concentration were obtained by CE-LIF for this on-column labeling method with Red-3 and Red-1c.

Role of phenoxyl radicals in DNA adduction by chlorophenol xenobiotics following peroxidase activation.

Chlorophenol (CP) toxins are classified as probable human carcinogens and are known to undergo bioactivation to generate benzoquinone (BQ) electrophiles that react covalently with biopolymers. Recently, we characterized the ability of pentachlorophenol (PCP) to react covalently with deoxyguanosine (dG) following treatment with horseradish peroxidase (HRP)/H2O2 or myeloperoxidase to yield a C8-dG oxygen (O)-adduct that suggested the intermediacy of the pentachlorophenoxyl radical in covalent bond formation [Dai, J., Wright, M. W., and Manderville, R. A. (2003) Chem. Res. Toxicol. 16, 817-821]. Investigations currently focus on a wider range of CP substrates (PCP, 2,4,6-trichlorophenol (2,4,6-TCP), 2,4,5-TCP, and 2,4-dichlorophenol (2,4-DCP)) to establish their reactivity toward dG and duplex DNA (calf thymus (CT)) following activation by HRP/H2O2, as a representative peroxidase system. Our data show that chlorophenoxyl radicals may either react directly with dG and CT-DNA to form C8-dG O-adducts in an irreversible process or couple to yield 1,4-BQ electrophiles that react with dG to afford adducts of the benzetheno variety. These results are the first to establish the in vitro relevance of C8-dG O-adducts of phenolic toxins. The 1H NMR chemical shifts and reactivity of the benzetheno adducts favor 4' '-hydroxy-1,N2-benzetheno-dG adduct assignment, which is in contrast to other literature which has assigned the 1,4-BQ-dG adduct as 3' '-hydroxy-1,N2-benzetheno-dG. Overall, the results from this current study have provided new insights into peroxidase-mediated activation of CP substrates and have strengthened the hypothesis that direct reactions of phenoxyl radicals with DNA contribute to peroxidase-driven toxic effects of phenolic xenobiotics.