REPLY:

We are grateful to D. Scott Wilbur for his précis regarding the nature of the dye–protein bond and accept that covalent bonding is not proven but proposed. The following response is offered nevertheless, as a rebuttal to his critique. Sulfonation or acylation reactions are chemical transformations that occur by an identical mechanism, that being the nucleophilic attack of nitrogen, oxygen, carbon, or sulfur atoms onto electrophilic sulfonyl or carbonyl centers, respectively. Such reactions are successfully performed in vitro under a variety of conditions, which usually include the use of activated substrates (sulfonyl- or acyl-chlorides, esters, anhydrides, etc.) in organic or even aqueous solvents. In a simple system in which a sulfonic acid–containing molecule is united with an amine such as lysine or arginine under neutral aqueous conditions, a water-solvated ion pair is expected to ensue (similar to the product depicted in the reaction sequence of the Wilbur letter). When the reaction medium is blood serum, however, this complex milieu comprises molecules with the potential to promote alternative interactions.

Different enzymes have frequently been used for coupling organic molecules in vitro. In particular, good yields of oligopeptides can be synthesized from the reaction of (nonactivated) carboxylic acids and amines in organic-aqueous solvents with subtilisin or endopeptidase/chymopapain (1) and in aqueous solvents with thermolysin/α-chymotrypsin, papain, or penicillin acylase (2). Amide bonds formed in these enzyme-catalyzed acylation reactions result in an overall loss of water. Similarly, a phosphorylation reaction between 2 phosphate-containing molecules can be achieved in an aqueous buffer (3), where a phosphorus–oxygen bond is created at the expense of H₂O. The statement that a dehydration reaction would be difficult to perform in an aqueous environment has been otherwise shown in the literature.

In lymph, serum, or extracellular tissue, proteins are known to participate in covalent bond formation either with themselves or with other molecules. Examples of the “chemisorption” reaction include cross-linking of fibrin during blood clot formation, cross-linking of collagen to produce connective tissue, reaction with other proteins as a consequence of aging, disulfide bond formation of fibrillins during assembly of microfibrils via sulfhydryl oxidases, retinoylation of proteins, and substrate phosphorylation. Cytosolic enzymes/proteins are routinely involved in covalent bond-forming reactions with sulfotransferase in sulfonation/sulfation (4) or the ribosome 50S subunit, nonribosomal synthetases, or polyketide synthases in amino acid acylation (peptide synthesis).

Coupling reactions between biologic molecules and xenobiotic small molecules ex vivo have been shown to form covalent-bonded adducts at temperatures up to 37°C, including carboxylated dyes with horseradish peroxidase (5), amine- or sulfhydryl-specific dyes and cytochrome c (6), malachite green cation and chicken egg albumin (7), thiazole orange derivatives with oligonucleotides (8), 4-nitrobenyl-³⁵S-mercaptan S-sulfonic acid with rat cytosolic proteins, a leukotriene-tetraenoic acid with 15-lipooxygenase, and a ribozyme-catalyzed formation of dipeptides. In view of the evidence supporting the formation of amide and other covalent bonds between proteins and endogenous/xenobiotic molecules, one cannot preclude the possibility of an enzyme-mediated sulfonamide bond formation between proteins and sulfonic acid-dyes as reported earlier (9).

The literature also provides evidence of a “physisorption” interaction of protein hydrophobic groups and small aromatic water-soluble molecules, where for example, naphthalene sulfonic acid groups are particularly associated with arginine groups on the exterior of the protein. Preliminary observations in this laboratory thus far, after use of instant thin-layer chromatography (ITLC), suggest that a simple ion pair does not define the dye–protein interaction, nor does a hydrophobic interaction. At the time of publication (9), it was known that naphthol blue black (NBB) and Evans blue (EB), in separate experiments, migrated on ITLC-silica gel/glass paper in saline (0.9%) with migration coefficients of R_f = 0.8 in the presence or absence of excess arginine or of lysine, phenylalanine, or polylysine. Mixing any of these dyes with plasma for a brief time (∼20 s) at room temperature, and then performing ITLC on the dye–protein mixture, resulted in a colored spot at R_f = 0.8, indicating no chemisorption or physisorption. This was confirmed by size-exclusion chromatography, a technique that serves to separate products from starting materials, where there was 100% recovery of the initial dye. However, altering the incubation conditions to 37°C for 10 min found, for NBB exclusively and EB predominantly, that the colored spots were visible at the baseline and near the baseline, respectively. Likewise for ^99mTc-EB, the location of radioactivity on the ITLC strip correlated with blue color. These observations indicate that a strong affinity exists between dye and protein and that a simple ion pair association is absent. Furthermore, although Coomassie blue G (a sulfonic acid reported to bind to proteins by physisorption) has shown anomalous behavior in its interaction with polylysine (10), both Evans blue and NBB, with R_f = 0.8, have not. In an ongoing study, this laboratory is gathering additional evidence to elucidate the bond between serum proteins and naphthalene-sulfonic acid dyes.