An improvement of the Ames test using a modified human liver S9 preparation

In preliminary studies on the Ames test using human liver S9 fractions, we found that the crude human S9 fractions, obtained following centrifugation of the tissue homogenate for 20 min at 9000 x g, were not always sterile. When this was the case, the S9 fractions were often accompanied by an increased number of colonies above the normal range on plates in the solvent control used in the Ames test. In addition, we also sometimes identified the incorporation of a small amount of fat in the crude human liver S9 fractions. We have therefore obtained a purified fat-free S9 fraction by a simple modification to the crude S9 preparation; fat was completely removed by centrifugation of the crude S9 fraction.

Using the purified and crude human S9 fractions (two lots each), both the sterility and the number of bacterial colonies produced on a plate with five bacterial tester strains by solvent controls (purified water and dimethyl sulfoxide) were examined. The findings were then compared to those observed with phosphate buffer or S9 fraction from rats pretreated with phenobarbital/5,6-benzoflavone.

The data show that each of the crude human S9 fractions was not sterile and produced an increasing number of colonies with each solvent control, almost equal to the sum of the numbers of contaminating bacterial colonies and spontaneous revertant colonies observed with phosphate buffer or the rat S9 fraction. On the other hand, both the purified human S9 fractions were sterile, and the number of colonies that appeared in each solvent control was similar to that of spontaneous revertant colonies observed with phosphate buffer or the rat S9 fraction.

These results indicate that this new procedure of S9 preparation, modified with an additional recentrifugation step, may provide a high quality of purified fat- and bacteria-free S9 fraction for use in the Ames test.