Journal of General Virology

Fig. 2. The kinetics of neutralization. (a) The kinetic neutralization plot shows the logarithm of the fraction of residual infectivity (I/I₀) as a function of the time of Ab–virus incubation. The study of neutralization kinetics requires a method for quenching the reaction before addition of the Ab–virus mixture to susceptible cells. If the Ab–virus reaction is only negligibly reversible, or if Ab dissociation does not reverse neutralization, then dilution of the Ab–virus mixture can be such a method (Dulbecco et al., 1956; Jerne & Avegno, 1956; Brioen et al., 1985a; McDougal et al., 1996). Linear curves for different Ab concentrations, increasing from [Ab]₁ to [Ab]₄, are shown. The rate of neutralization, dln[I/I₀]/dt, can thus be calculated for each Ab concentration. (b) When the rates of neutralization obtained from (a) are plotted as a function of Ab concentration and a linear relationship is obtained (as illustrated), this has been taken as evidence that the reaction is of first order in Ab concentration. But the molar excess of Ab over virus means that it is merely of pseudo-first order: determining the real order of the neutralization reaction would involve measurement of the concentration of uncomplexed Ab. Not even the true kinetics of the reaction would demonstrate the number of molecules involved, i.e. the molecularity of the reaction. (c) The common term 'single-hit kinetics' (not justified as a label for first-order, let alone pseudo-first-order, kinetics of the neutralization reaction) is sometimes applied to the lack of an initial shoulder or lag on the curve describing neutralization over time (Della-Porta & Westaway, 1978). The argument is that if the first Ab molecule knocks out the infectivity of the virion that it ligates, then there should be no lag, even at low temperatures and low Ab concentrations. Evidence for such lags has sometimes been overlooked in papers arguing for a single-hit mechanism (e.g. Dulbecco et al., 1956). But it is uncertain whether absence of a lag could legitimately be used to support a single-hit hypothesis because of the comparatively rapid Ab binding, the difficulty in stopping and recording neutralization after a few seconds and, at least in the case of enveloped viruses, possible heterogeneity among virions, some of which may be neutralized more quickly than others (Klasse & Moore, 1996). A recurrent phenomenon is the levelling off of neutralization after a sharp drop in infectivity. The remaining non-neutralized virus is called the persistent fraction. Persistent infectivity, which can vary between different target cells (Kjellen, 1985), has been attributed to Ab dissociation and virus aggregation (reviewed in Burton et al., 2001; Klasse & Sattentau, 2001). (d) The molecularity of virus neutralization by Ab corresponds to the minimum occupancy required for neutralization of the individual virion but does not follow directly from stoichiometrically determined average occupancy: when a subset of virions in a population mixed with Ab is neutralized, some of them will have greater than necessary occupancy. Occupancy can be analysed under the premise of Poisson-distributed Ab binding, as shown in the stoichiometric plot of the logarithm of relative infectivity as a function of the average number of Ab molecules per virion. If single-hit neutralization is assumed, then only virions with no Ab bound would be infectious. Single-hit neutralization corresponds to the linear curve that goes through the point (1, –1), i.e. 37 % relative infectivity at the average of one Ab molecule per virion. That value, or preferably the whole curve, should be used to test the correspondence between the actual relative infectivity and the predicted one for single-hit neutralization. When a requirement for several Abs bound per virion has been shown (Flamand et al., 1993; Icenogle et al., 1983), the figure 37 % is arbitrary. Each hypothetical multiplicity of hits should be tested by comparing its proper theoretical curve with the actual one (Klasse & Moore, 1996). Three non-linear, few-hit curves are plotted as examples.

Back to article

0001-8242 © 2002 SGM