Enzyme immunoassay (EIA) was introduced in by Engvall and Perlmann, 1972, for applications where RIA would also be an option. This general class of assays can involve the detection of antibody or antigen of interest. The basic idea is to bind an enzyme-conjugated anti-Ig to an antigen-antibody complex. Adding the substrate for the enzyme can produce a color change or fluorescence in the product which is easily quantified -- the magnitude of the color change or the fluorescence related to the amount of the complexes present, although difficulties can be present in interpretation (reviewed in Pesce and Michael, 1992, also Francois-Gerard, et al., 1988; Beatty, et al., 1987; Pick and Mizel, D., 1981; Stemshorn, et al., 1983; Ehle et al., 1989; and Herraez, A., 1993). Enzyme-linked immunosorbant assays (ELISA) refers to the assay involving a solid-phase antigen. As this assay is often done in microtiter plates, much of the work involving computational models involves the methodology in dealing with the data in that form, and with enhancements to this assay Examples are Canellas and Karu, 1981; Richardson, et al., 1983; Slezak, et al., 1983; Caulfield and Shaffer, 1984; Franco, et al., 1984; Slade, et al., 1986; Zrein, et al., 1986; Mixter et al., 1986; Karpinski et al., 1987; La Belle, 1987; Gigase et al., 1988; Huet et al., 1990; Raghava, et al., 1992; Glaser, 1993a; Glaser, 1993b; Tremain, 1993; Reiken, et al., 1994; Fuchs, et al., 1995; Reizenstein, et al., 1995; Stevens and Kelso, 1995; and Iznaga Escobar, et al., 1996. Problems with affinity measurements in ELISA are discussed in Underwood, 1993; a discussion of the statistical framework for ELISA assays can be found in Bunch, et al., 1990, Sittampalam, et al., 1996, and McGuinness et al., 1997.
The oldest form of enzyme immunoassay is hemolysis, the complement-mediated lysis of red blood cells discussed previously. Lysis of the cells indicates the presence of complement fixing antibody which can bind to antigen on the surface of the red cells. Using erythrocytes as indicators in assays such as the Jerne plaque assay (Jerne and Nordin, 1963), allows both an indication of the number of antibody producing cells and an estimate of the rate of antibody production and affinity. These last results require computational models of the assay. DeLisi and Bell, 1974 and Jerne, et al., 1974 present general theoretical results. Sette et al., 1988 presents software for analysis of modulation of cellular responses through the use of plaques. Esrig, et al., 1977 presents a computational approach suitable for the laboratory.
Computational modeled applied to cellular assays both developed later, and generally has a different nature than that discussed thus far. More of the work involves statistical models, and mathematical models and simulations that do appear tend to be more sophisticated. Given the nature of this review, the discussion area of each will be brief, with one exception.
There are several approaches to quantitatively describing the action of a cytotoxic cell population on a "target" cell -- usually a susceptible cell line. One which has turned out to be applicable to many cell types (cytotoxic T lymphocytes and NK cells in particular) is to use an analogy with classical enzyme kinetics (Michaelis and Menten, 1913) where the cytotoxic cells play the role of an enzyme, the product being eventual target cell lysis. Lysis can be noted either through the release of a radioactive label or through non isotopic markers (such as in a colorimetric assay). The "substrate" in the enzyme-kinetic analogy is the target cell -- usually supplied in excess in the assay (Miller and Dunkley, 1974; Dunkley, et al., 1974; Thorn and Henney, 1976; Zeijlemaker, et al., 1977; Callewaert, et al., 1978). The advantage of using this analogy is that familiar transformations, such as the double reciprocal Lineweaver-Burk plot, can be directly applied, along with extensions to this formalism to include inhibition and other factors, and improvements in parameter estimation (Callewaert, et al., 1982; Merrill, 1982; Merrill and Sathananthan, 1986; Garcia-Penarrubia and Bankhurst, 1989a; Garcia-Penarrubia, P., et al., 1989b; Garcia-Penarrubia, et al., 1992; Galvez, et al., 1994; Garcia-Penarrubia, et al., 1995). Other approaches to quantifying cytotoxicity can be found in Chalmers et al., 1982; Crispe and Gascoigne, 1983; Hogan and Evans, 1984; Bol et al., 1986; Sheeran et al., 1988; Hudig et al., 1988. Rodgers et al., 1992 used a colorimetric assay to identify CTL determinants.
.gif)
.gif)
.gif)