Cycle number, immunosorbent

For B (1 mg ml" ), cycled 50 times over a 195 day period, n]/2 = 53 cycles. Figure 2 shows the capacity of immunosorbent B above cycled 40 times over a period of 3 days. In this case no change in the capacity was observed as a function of cycle number. This model showed that antibody loading and time between cycles had a greater effect on capacity as a function of cycle number than process stream or elution solvent. However, every system is not expected to parallel this example. 

Figure 1. Capacity of model immunosorbent systems as a function of cycle number. (A) Goat anti-human IgC Fab fragment immobilized at 7 g/L. Column volume = 6 ml. (B) Same as above except that the initial antibody loading was 1 gL- and the immunosorbent volume was 4.2 ml.

Figure 2. Capacity of a model immunosorbent system as a function of cycle number.

As the capacity of the immunosorbent column decreases with increasing cycle number the concentration of the purified protein decreases as the volume in which it is eluted is a function of the column volume. If the concentration of eluted protein is 1 in the first cycle, the average concentration eluted in n = n/2 cycles = 0.75. The average concentration in the second cycle half-life = 0.375 and in the third cycle half-life - 0.1875. As a general rule if the cost of the final isolation is 1 for the first cycle half-life, it will be 1.33 for the second cycle half-life and 1.71 for the third cycle half-life. Regardless of the final process step, the decrease in concentration of eluted protein with cycle number will increase the final isolation costs and must be weighed against the cost of antibody needed to increase column volume and decrease the number of cycles. 

Allowing for the decrease in activity with cycle number the total amount of immunosorbent required is 21.4 L. At 2 gL-l loading and 200 g l for the monoclonal antibody and 200 IT for the matrix and immobilization, the total cost of the immunosorbent is 12,840. This amounts to 4.90 g l of urokinase. In this example the low costs are obviously attractive. 

After the appropriate monoclonal antibody, immobilization method and matrix have been chosen according to the criteria discussed above and methods previously described (7,8) the major factor in determining the cost of this purification method is the amount of antibody required. The amount of antibody required is determined by the capacity per cycle of the immunosorbent and the number of cycles that can be utilized in a given process. The capacity per cycle for the immunosorbent is given by Equation 1. 

As shown in Equation 1, the capacity per cycle is directly proportional to the amount of antibody immobilized, the immobilization yield, the M.W. of the protein and the column volume and an exponential function of the number of cycles. The amount of antibody immobilized will usually be less than 10 gL l. Higher activation of the matrix required for greater than 10 gL l loading results in a decrease in the immobilization yield. The maximum immobilization yield is 1.0 (100%) while 0.8 (80%) is not difficult to obtain. The M.W. of the protein to be isolated is fixed. The only way to increase the capacity per cycle significantly is to increase the volume of the immunosorbent or increase the number of cycles prior to reaching 50% of initial capacity (cycle half-life). Increasing the volume of immunosorbent increases the amount of monoclonal antibody required. 

Since the cost of the antibody is the major cost, increasing the volume of immunosorbent is the most expensive way to increase capacity per cycle. The least expensive way to increase the capacity of the isolation system is to increase the number of cycles. The number of cycles that can be obtained in any given purification is dependent on the cycle half-life and time-volume constraints. The total amount of protein that can be isolated in a given number of cycles is given in Equation 2. 

Ultimately, the cost of immunosorbent isolation will depend on the entire process and must be evaluated against alternative processes. Consider, as an example, the costs and decisions involved in the purification of urokinase. One course of drug therapy consists of 33 mg of urokinase (4,000,000 CTA units). At the hospital pharmacy the drug costs for one course of treatment are currently 3,000 (9), or 91,000/gram. There are approximately 76,000 patients in the U.S. that could be treated with urokinase therapy each year requiring an annual production of approximately 2,500 g. We have selected a monoclonal antibody that has allowed the purification of urokinase from urine, tissue culture media, and bacterial culture media in a single step with 85% retention of urokinase activity (6). This monoclonal antibody was immobilized at 2 gL l with an immobilization yield of 0.8 and a cycle half number of 300 cycles. The urokinase capacity for the first cycle would be 1.2 gL l of immunosorbent. 

---