The root of the solution.

It has been known for over a decade that the alkylating agent Cyclophosphamide (CY) has a differential effect on the immune system, exerting greater toxicity against B than T lymphocytes CY selectively depletes the B-cell rich peripheral splenic pulp and exerts a greater inhibition of stimulation of B-cells by pokeweed mitogen than T cells by phytohaemagglutinin (PHA; Stockman et al., 1973). This differential effect is found even at the earliest time of life; CY given to newly hatched chicks leads to a severe depression of antibody synthesis in adult chickens while T cell functions are virtually unimpaired (Lerman & Weidanz, 1970). Recent experiments have indicated that part of these effects are due to inhibition of T suppressor cell activity (Turk & Poulter, 1982). The exact mechanism by which this selectivity is exerted could not previously be studied in vitro since CY requires activation in the liver (Hill, 1975). We show here that the newly available active form of CY, phosphoramide mustard (PM, Fenselau et al., 1977) is virtually inert against colony-forming T cells in tissue culture at levels highly cytotoxic to other cells. Cultured human T lymphoma cells retain a significant amount of this resistance; recent work in other laboratories suggests that the same is true for important human marrow-repopulating stem cells (see below). To determine the effect of PM on T lymphocytes we used the quantitative colony assay technique we developed for examining the proliferation-dependency of cytotoxicity in vitro. In this assay (Byfield & Calabro-Jones 1981) purified preparations of mononuclear cells from human blood are treated with PHA and the number of colonies (T-CFUCS, T cell colony-forming units in culture) are subsequently counted. The addition of active cytotoxic drugs inhibits this colony formation and yields survival curves similar to those found with permanently explanted lines (Byfield & Calabro-Jones, 1981). In the experiments reported here we studied the effect of PM exposure both prior to and after the induction of proliferation by PHA, using an identical protocol employed by us with other alkylating agents (Byfield & Calabro-Jones, 1981). We have previously shown that the sequence of exposure to PHA and drug influences the shape of the resultant survival curves, the effect being a function of solubility characteristics of each agent (Byfield & Calabro-Jones, 1981). The survival curves for log phase WI-L2 cells (a malignant human B cell line) and CEM cells (a human T lymphoma line) exposed to PM were …