) Despite limited efficacy as a single agent (20% response rate), fluorouracil (5-FU) is the most commonly prescribed agent for colorectal cancer. Tumor resistance to 5-FU is the principal cause of 5-FU failure in most cancer patients. Several mechanisms for 5-FU resistance have been demonstrated in vitro and include high levels of dUMP and UTP which compete with 5-FU metabolites for target sites (thymidylate synthase or TS, RNA) and increased levels of TS. The long-term objective of this project is to improve the antitumor response and survival of colorectal cancer patients treated with 5-FU, PALA and leucovorin (LV), two agents with no significant anticancer activity, can biochemically modulate the cytotoxicity of 5-FU. These agents may also circumvent resistance to 5-FU. PALA, an inhibitor of aspartate transcarbamylase (ATCase), depletes intracellular pools UTP and CTP thus allowing enhanced binding of 5-FU metabolites to TS. LV increases pools of reduced folates leading to enhanced binding of 5-FU metabolites to TS. To date, much controversy exists concerning the dose of PALA or LV when given with 5-FU. These proposed studies will determine the dose-response relationship of PALA in inhibiting ATCase and indirectly inhibiting TS catalytic activity in normal tissue and tumor. The effect of PALA on TS activity will also be examined. In addition, the dose-effect of LV in inhibiting TS activity with 5-FU will be examined. In this novel translational study, biopsies of tumor tissue and normal bowel mucosa will be obtained from patients with suspected colorectal cancer during sigmoidoscopy. Patients with histologically documented, resectable colorectal cancer will be eligible to enter the treatment protocol. A 3 x 3 factorial design will be used to randomize patients to three PALA doses (0, 250, or 500 mg/m2 on day 1) and LV (0, 20, or 500 mg/m2 on days 2-3). All patients will receive 5-FU 450 mg/m2 on days 2-3. On day 4, the patients will undergo resection of their tumor and tumor and normal bowel mucosa samples will be obtained. These samples, along with the samples obtained at pretreatment will be assayed for ATCase activity and total, free, and bound TS catalytic activity. These studies offer the rare opportunity to evaluate the biochemical efficacy of these agents in humans and will lead to the rational selection of PALA and LV doses in future trials of 5-FU in colorectal cancer.