Epidemiological evidence indicated that residents, especially cigarette smokers, in arseniasis areas had significantly higher lung cancer risk than those living in non-arseniasis areas. Thus, an interaction between arsenic and cigarette smoking in lung carcinogenesis was suspected. p53 dysfunction or mutation in lung epithelial cells was frequently observed in cigarette smokers. Our present study was to explore the differential effects by arsenic on H1355 cells (human lung adenocarcinoma cell line with mutation in p53), BEAS-2B (immortalized lung epithelial cell with functional p53) and pifithrin-alpha-treated BEAS-2B cells (p53-inhibited cells). These cells were treated with different doses of sodium arsenite (0, 0.1, 1, 5 and 10 microM) for 48 h. A greater reduction in cell viability was observed in the BEAS-2B cells vs. p53 compromised cells (H1355 or p53-inhibited BEAS-2B). Similar observation was also made on 7-day cell survival (growth) study. TUNEL analysis confirmed that there was indeed a significantly reduced arsenite-induced apoptosis found in p53-compromised cells. Centrosomal abnormality has been attributed to eventual chromosomal missegregation, aneuploidy and tumorigenesis. In our present study, reduced p21 and Gadd45a expressions and increased centrosomal abnormality (atopic and multiple centrosomes) were observed in both arsenite-treated H1355 and p53-inhibited BEAS-2B cells as compared with similarly treated BEAS-2B cells. Increased anchorage-independent growth (colony formation) of BEAS-2B cells co-treated with pifithrin-alpha and 5 microM sodium arsenite was also observed in soft agar. Our present investigation demonstrated that arsenic would act specifically on p53 compromised cells (either with p53 dysfunction or inhibited) to induce centrosomal abnormality and colony formation. These findings provided strong evidence on the carcinogenic promotional role of arsenic, especially under the condition of p53 dysfunction.