Retention properties of hydrophobically end-capped poly(ethylene glycol)s on a beta-cyclodextrin support.

High-performance liquid chromatography (HPLC) was used to examine the retention behavior of monomethoxypoly(ethylene glycol)s bearing one hydrophobic naphthyl end group (Nap-MPEG) on beta-cyclodextrin polymer (poly-beta-CD) immobilized on a silica support, under isocratic elution conditions and using water as mobile phase. Studies of retentions and theoretical plate heights H were conducted at infinite dilution by comparing the behavior of Nap-MPEGs having different molecular weight (750, 1000 and 5000 g/mol). The larger is its molecular size, the lower is the retention of the polymer. The linear increase of H with mobile phase velocity reveals slow mass-transfer kinetics arising from the restricted diffusion into the pores of the support. The complexation constants between the Nap-MPEGs and beta-CD in solution (around 500M(-1)) were determined from the decrease of retention observed by adding increasing concentrations of hydroxypropyl beta-CD into the eluent. The peak profiles in mass-overload conditions were studied by fitting a model based upon bi-Langmuir kinetics which assumes a non-uniform support having two types of binding sites and apparent adsorption rate constants are used to describe mass-transfer kinetics. A three-parameter adsorption equilibrium isotherm was sufficient to account for the modifications of peak shapes observed when increasing amounts of polymer were injected. This result indicates an interaction with a heterogeneous poly-beta-CD support mainly composed of low affinity groups, non-saturable in the range of polymer concentration studied. An upper limit was estimated for the equilibrium constant (<1000 M(-1)) characterizing the affinity of Nap-MPEG for the non-saturable sites of the poly-beta-CD support. Large affinity constants (8-9 x 10(4)M(-1)) were found for the interaction of Nap-MPEGs with a small percentage of active sites.

Frontal analysis for characterizing the adsorption-desorption behavior of beta-lactoglobulin on immunoadsorbents.

High-performance frontal affinity chromatography was employed to study the adsorption-desorption kinetics characterizing the retention of beta-lactoglobulin (beta-LG) onto polyclonal anti-beta-lactoglobulin (anti-beta-LG) chromatographic supports. The adsorption and desorption processes were studied by analyzing two different elution fronts separated by a relatively long rinsing step. The method consists in performing two successive frontal injections of the protein. In between, the column was rinsed with a given volume of mobile phase (buffer alone). During this rinsing stage, a partial desorption may occur and a novel amount of protein could be adsorbed in the second frontal injection step. The whole process (first adsorption, possible desorption, and second adsorption) was simulated by a numerical procedure, in which the column was divided into a large number of slices. A model based on bi-Langmuir type kinetics was used to describe the adsorption of the protein on the support. The model assumes a non-uniform adsorbent with two types of binding sites. At equilibrium the adsorption isotherm is of the bi-Langmuir type. A global adsorption effect was considered which includes the effective binding process and the mass transfer resistances due to the transport to the binding site. Therefore, the column capacity and the kinetic parameters of the model (apparent adsorption and desorption rate constants) were determined from the best fit of the first and second adsorption fronts to the experimental ones. The other parameters of the model are the saturation capacities for the adsorption on each type of sites. The equilibrium affinity constants were estimated in a single experiment from the ratio of the apparent adsorption and desorption rate constants. The high values found (around 10(8) M(-1)) reveal a strong interaction of beta-LG with the immunoadsorbent. Kinetic measurements were carried out at different flow rates. Both the apparent adsorption and desorption kinetics were faster at larger flow rates, indicating an important contribution of the mass transfer resistance in the stagnant fluid at the particle boundary. However, as expected, close values were found for the resulting equilibrium constants calculated from the ratio of the apparent adsorption and desorption rate constant determined at various flow rates.

Numerical simulation of the chromatographic process for direct ligand-macromolecule binding studies.

A numerical simulation of the direct zonal liquid chromatographic method is described for studying the binding of a ligand to a macromolecule by quantification of the interacting species present in a sample at equilibrium. The algorithm accounts for both the kinetic exchanges in solution and the dispersion effects depicted by the Fick law. Dimensionless variables are used for the concentrations which are expressed as a function of the equilibrium constant, KD. The free ligand concentration was varied in the injected samples from 0.1 to 20 KD, while that of the macromolecule was kept constant. An apparent binding isotherm was obtained from the total ligand chromatogram generated by the simulation run, when the amount emerging at almost column dead volume is plotted against that eluting at the free ligand retention time. As a continuous dissociation of the complex may occur during its migration, the apparent binding curve and the theoretical binding isotherm coincide at extremely low dissociating rates. At larger dissociation rates (0.001 s(-1) < kd <0.1 s(-1), for a first peak eluting in 1 min) the simulations were used to test various chromatographic conditions. The flow rate (or column volume) is the major effect which influences the on-column dissociation process as an exponential decay was found when the apparently bound fraction is plotted against the analysis time. The apparent equilibrium coefficient is close to the theoretical one for a binding curve generated with an initial solution containing a relatively low total concentration of binding sites (< or = KD). The apparent stoichiometric term is largely underestimated as its value decreases exponentially at increasing dissociation rates. An extrapolation at extremely short analysis times could be used to determine the stoichiometric coefficient characterizing the binding interaction.

Direct zonal liquid chromatographic method for the kinetic study of actinomycin-DNA binding.

The binding of an anticancer drug (actinomycin D or ACTD) to double-stranded DNA (dsDNA) was studied by means of high-performance liquid chromatography (HPLC). ACTD is an antitumor antibiotic containing one chromophore group and two pentapeptidic lactone cycles that binds dsDNA. Incubations of ACTD with DNA were performed at physiological pH. The complexed and free ligand concentrations of the mixture were quantified at 440 nm from their separation on a size-exclusion chromatographic (SEC) column using the same buffer for the elution and the sample incubation. The DNA and the ACTD-DNA complexes were eluted at the column exclusion volume while the ligand was retained on the support. An apparent binding curve was obtained by plotting the amount emerging at the exclusion column volume against that eluted at free ACTD retention volume. A dissociating effect was evidenced and the binding parameters were significantly different from those obtained at equilibrium by visible absorbance titration. The equilibrium binding parameters determined by absorption spectroscopy were used as starting data in the numerical simulations of the chromatographic process. The results showed a strong dependency of the apparent binding parameters on the reaction kinetics. Finally the comparison of the apparent binding curve obtained from the HPLC experiments and from the numerical simulations permitted an evaluation of the dissociation rate constant (kd = 0.004 s(-1)).

Adsorption kinetics of beta-lactoglobulin on a polyclonal immunochromatographic support.

Beta-Lactoglobulin is one of the main components of whey proteins. Among other reasons, its allergenicity makes its determination in hypoallergenic foods and bio-pharmaceutical products necessary. Immunoaffinity chromatography is a widely accepted technique for purification and analysis of proteins. Knowledge of the apparent kinetics of the adsorption of beta-lactoglobulin onto the anti-beta-lactoglobulin immunochromatographic column is important to optimize the analytical process. High-performance frontal affinity chromatography was used to study the apparent kinetics of the adsorption process. Langmuir and bi-Langmuir kinetic models, assuming one and two kinds of binding sites, respectively, were used to characterize the adsorption kinetics of beta-lactoglobulin B on a polyclonal immunoadsorbent. Very good fits were obtained with the bi-Langmuir model for two different concentrations of beta-lactoglobulin and this allowed us to calculate the apparent adsorption rate constants and the column capacities for both kinds of sites. Experimental results indicate the possibility that the adsorption process is not irreversible. The values of the apparent dissociation rate constants leading to the best fit were estimated and the affinity constants were calculated.

Monte Carlo model of nonlinear chromatography: correspondence between the microscopic stochastic model and the macroscopic Thomas kinetic model.

The Monte Carlo model of chromatography is a description of the chromatographic process from a molecular (microscopic) point of view and it is intrinsically based on the stochastic theory of chromatography originally proposed by Giddings and Eyring. The program was previously validated at infinite dilution (i.e., in linear conditions) by some of the authors of the present paper. In this work, it has been further validated under nonlinear conditions. The correspondence between the Monte Carlo model and the well-known Thomas kinetic model (macroscopic model), for which closed-form solutions are available, is demonstrated by comparing Monte Carlo simulations, performed at different loading factors, with the numerical solutions of the Thomas model calculated under the same conditions. In all the cases investigated, the agreement between Monte Carlo simulations and Thomas model results is very satisfactory. Additionally, the exact correspondence between the Thomas kinetic model and Giddings model, when near-infinite dilution conditions are approached, has been demonstrated by calculating the limit of the Thomas model when the loading factor goes to zero. The model was also validated under limit conditions, corresponding to cases of very slow adsorption-desorption kinetics or very short columns. Different hypotheses about the statistical distributions of the random variables "residence time spent by the molecule in mobile and stationary phase' are investigated with the aim to explain their effect on the peak shape and on the efficiency of the separation.