In vitro immunotoxicological assays for detection of compounds requiring metabolic activation.

A system for metabolic activation of cyclophosphamide (CP), consisting of a crude microsomal fraction of mouse liver and necessary cofactors (S9 mix), was interfaced with three murine cell culture assays for immunotoxicity. These assays were: the Mishell-Dutton assay for in vitro antibody formation, splenic lymphocyte responsiveness to mitogens and bone marrow cell cultures. There was no effect of CP at doses up to 261 microgram/ml (lmM) on any of the parameters measured unless S9 mix was included. Much greater potency was achieved if the S9 mix was prepared from livers of mice pretreated with phenobarbital. Under these conditions and dose-related inhibition of plaque-forming cells (PFC) in the Mishell-Dutton assay was observed, yielding an ED50 of 6.3 microgram/ml. When splenic lymphocytes were exposed to CP in the presence of induced S9 mix, a dose related inhibition of the response to the B-cell mitogen, lipopolysaccharide (LPS), and to the T-cell mitogen, concanavalin A (Con A), was observed. For the optimum LPS concentration, the ED50 for CP was 8.1 microgram/ml; for the optimum concentration of Con A, the ED50 was 6.7 microgram/ml. DNA synthesis was not inhibited by the doses used. When bone marrow cells were exposed to CP in the presence of induced S9 mix, the stem cell population, enumerated by colonization in semisolid medium, was reduced in a dose-dependent manner, with an ED50 of 5.2 microgram/ml. Again, DNA synthesis was not affected unless higher doses of CP were used.

In vitro assays for defining toxicological effects are becoming increasingly important as the burden ofnecessary investigations increases. Our laboratory has successfully interfaced a system for metabolic activation with several in vitro assays for immunotoxicity, thus widening the scope of these assays to include immunosuppressive agents which are not direct acting. Cyclophosphamide (CP) was chosen as a positive control in these assays since it requires metabolic activation (1) and is immunosuppressive in the mouse (2).

Metabolic Activation System
A crude microsomal fraction was prepared from livers of male BALB/c mice to serve as a source of *Department of Pharmacology, Medical College of Virginia, Richmond, Virginia 23298. February 1982 enzymes for metabolic activation. In some cases, the enzymes were induced with phenobarbital by placing 1 mg/ml in the animals' drinking water for 1 week until 24 hr prior to sacrifice by cervical dislocation. Livers were removed aseptically and homogenized in 3 ml of cold, sterile 1.15% KCl/g liver. This homogenate was centrifuged at 9000g and the supernatant (S9 fraction) removed and frozen in 1 ml aliquots. When used, a tube of S9 was removed from storage at -70°C, thawed and added to a solution of cofactors. This S9 mix contained 3 mg S9 protein, 2 mM NADP, and 35 mM isocitrate in sterile culture medium and was added directly to cell suspensions, 0.1 ml/ml of cell suspension.

Mishell-Dutton Assay
The Mishell-Dutton in vitro antibody-producing assay has been found useful for recognizing immu-  (3). In addition to specificity, the system offers the advantage of complex cellular interactions involved to achieve the ultimate response. The disadvantage of the assay, as pointed out by Kutz et al. (4), is that chemicals which require activation to reactive metabolites are not detectable. With the modification described here, chemicals which require metabolic activation can be tested in the assay.
The modified assay was carried out by first preparing a spleen cell suspension in RPMI 1640 culture medium and adjusting the cell concentration to 1.5 x 107 cells/ml. The suspension was then distributed into 10 x 35 mm plastic petri plates, 1 ml per plate. To each plate was added 0.1 ml of S9 mix and CP (Sigma) as indicated in Table 1. The plates were incubated at 37°C for 1 hr in a culture box gassed with a mixture of 10% Co2, 7%02, 83% N2 on a rocker platform. As described elsewhere (5), the 1 hr preincubation was necessary because of loss of response of the assay when S9 mix was added at the time of antigen and left in for the 5 days of culture. Following this incubation period, the contents of each plate were transfered to a sterile plastic tube and centrifuged at 2000g for 10 min. Each pellet was resuspended in 1 ml RPMI 1640 (1 x 107 cells/ml), 0.1 ml fetal calf serum and antigen (5 x 106 sheep erythrocytes) were added, and 1 ml of the resulting suspension was placed in a fresh petri plate. The plates were then placed in the culture boxes, gassed with the special mixture, and incubated at 37°C with gentle rocking and daily feeding as described by Mishell and Dutton (6). On day 5, the contents of each plate were collected for enumeration of plaqueforming cells (PFC) using Cunningham slides (7). Cell counts were performed by using a Model ZBI Coulter counter.
The effect of CP on the primary antibody response 124 of splenic lymphocytes to sheep erythrocytes using the modified Mishell-Dutton assay is shown in Table  1. The control value (1760 PFC/culture) can be compared to a mean control value from six experiments not involving the modification of 2680 + 273. Cyclophosphamide had no effect up to lmM (261 ,g/ml) when S9 mix was not included, but produced a dose-dependent inhibition in the presence of S9 from livers of either naive or phenobarbital-induced mice. The use of the induced preparation caused a tenfold increase in potency of CP, as would be expected. The preparations used in this experiment had been stored for 3 months at -70°C. Although the fractions were more active when used shortly after preparation, the induced preparation was slightly inhibitory at that time.
There is an effect on cell yield at the higher doses of CP when S9 mix is included which seems to be related to the reduced yield of PFC/culture. For comparison, PFC/106 cells is shown in the last column of Table 1

Splenic Lymphocyte Cultures
We have been using murine splenic lymphocyte cultures to assess immunotoxicity of a variety of chemicals. The parameters measured include effects on DNA synthesis and response of the lymphocytes to mitogens. The use of the S9 mix and effect of CP in this assay will be described.
Cell suspensions were prepared in RPMI 1640 by using spleens from male CD-1 mice, a random-bred albino from Charles River, which were 10-12 weeks of age. This strain is also being used in our laboratory for in vivo toxicological studies. Following was resuspended in RPMI 1640 containing 10% calf serum and adjusted to 5 x 106 cells/ml. Cyclophosphamide and S9 mix were added to triplicate siliconized glass tubes, each containing 5 ml of cells. A 1-hr incubation was performed with the tubes in a roller drum in a 37°C, 5% CO2 incubator. At the end of this incubation, the cells were collected by centrifugation and resuspended in RPMI 1640 containing 10% fetal calf serum for counting. Viability was determined by trypan blue exclusion. For determination of DNA synthesis, a portion of the cell suspension was adjusted to 3 x 106 cells/ml and incorporation of 1251-iododeoxyuridine (IUdR) was measured in the presence of 1 ,uM fluorodeoxyuridine (FUdR). In this assay, 6 replicate 200 p.l aliquots were incubated with label in a microtiter plate for 1 hr, and the cells were collected with a Titertek Cell Harvester. Quantitation of 125I was by gamma counter.
The response of splenic lymphocytes to the B-cell mitogen, lipopolysaccharide (LPS) from Salmonella, and the T-cell mitogen, concanavalin A (Con A), was determined by IUdR incorporation also. Suspensions were adjusted to 5 x 106 cells/ml and six 100 pul replicates incubated with mitogen in microtiter plates in a 37°C, 10% CO2 incubator for 48 hr. Three concentrations of each mitogen were used.
The samples were then pulsed for 18 hr with 1"IUdR/ February 1982 FUdR and the 125I incorporation determined as described for DNA synthesis. Cell recovery was not affected by CP, even when S9 mix was included. This is illustrated graphically in Figure 1. Viability, which was 91% in the control was also not changed by any of the treatments. DNA synthesis was also unaffected, as shown in Table 2 and Figure 1. Data from other experiments which are not shown here indicate that CP in the presence of induced S9 inhibits DNA synthesis at higher doses than used in the experiment described.
Mitogenicity data are shown in Table 3 Table 3 gives the control values, and Figure 1 is a graphic representation of mitogen responsiveness when both CP and S9 are included in the 1-hr preincubation. A dose-dependent inhibition by CP is observed when S9 is included, and this inhibition occurs at lower doses if S9 from phenobarbital induced animals is used as the source of activating enzymes. The ED50 in this case is 8.1 ,ug/ml for LPS (5 ,ug/well) and 6.7 ,ug/ml for Con A (5 ,ug/well).

Bone Marrow Cultures
Immediately after removal of spleens from CD-1 mice, the hind limbs were removed and placed on ice. Using aseptic techniques, the femurs were trimmed and the contents flushed into a plastic centrifuge tube with a MEM (minimum essential medium) containing 10% fetal calfserum and adjusted to 3 x 106 cells/ml. Aliquots of 5 ml were placed in siliconized glass tubes, S9 mix and CP added where indicated, and the tubes incubated 1 hr in the roller drum as described for splenic lymphocytes. Cells were then collected by centrifugation and resuspended in a MEM with 10% fetal calf serum. Cell concentration and viability were determined, then the cells were adjusted to 3 x 106 cells/ml for determination of DNA synthesis as described for lymphocytes.
Viable stem cells were then enumerated by colonization in semisolid medium. For this determination, the cells were adjusted to 105 cells/ml in a MEM-containing 10% fetal calf serum, 5% horse serum, 1.8% methyl cellulose, and 10% L-cell conditioned medium. A 2-ml portion of suspension was placed in each of three wells in a six-well Linbro plate, and the plates incubated at 37°C, 10% C02, 95% humidity for 7-10 days. Colonies, defined as clusters of 25 cells or more, were then counted.
The results of a typical dose response experiment using S9 from both naive and phenobarbital induced animals are given in Table 4. Cell yield was not affected by the treatment; viability was 91-93% in controls as well as all high dose samples. As shown in Table 4, there was no inhibition of DNA synthesis at any dose of CP, even when S9 was included. Colony formation was affected; however, in a dose related manner when S9 from either naive animals or phenobarbital induced animals was included. Approximate ED50 values for the CP effect on stem cells are 20.8 ,ug/ml for uninduced S9 and 5.2 ,ug/ml for induced S9.

Discussion
In vitro activation of CP was first described in Ames' Salmonella microsome test for mutagenicity   (8). Since then, it has been used as a model compound in other, more sophisticated mammalian in vitro assays. Stetka and Wolff were able to induce sister chromatid exchanges in cultured Chinese hamster ovary cells by including rat liver microsomes with CP (9). Recently, an in vitro teratogenic assay was described in which CP induced defects when an hepatic microsomal fraction was included (10). The present report extends use of in vitro activation of CP into the area of immunotoxicology.
A system for metabolic activation of CP in vitro has been successfully interfaced with three cell culture assays for immunotoxicity. A summary of the effects of CP when activated by an S9 mix from phenobarbital induced animals is given in Table 5. The approximate ED50 values for specific end points in the three assays are very similar, ranging from 5.2 to 8.1 p,g/ml or 2.0 to 3.1 x 1OM. It is important to note that DNA synthesis is inhibited only when much higher doses of CP are used. The utility of the described assays is tremendously increased by inclusion of the S9 mix, in that they can now be used for screening for immunotoxicity, as well as elucidation of mechanisms of action. As with all in vitro assays, it is important to keep the information in context with in vivo effects.