Polycyclic aromatic hydrocarbons (PAHs) present in ambient air are considered as potential human carcinogens, but the detailed mechanism of action is still unknown. Our aim was to study the in vitro effect of exposure to dibenzo[a,l]pyrene (DB[a,l]P), the most potent carcinogenic PAH ever tested, and benzo[a]pyrene (B[a]P) in a normal human diploid lung fibroblast cells (HEL) using multiple endpoints. DNA adduct levels were measured by 32P-postlabelling, the expression of p53 and p21(WAF1) proteins by western blotting and the cell cycle distribution by flow cytometry. For both PAHs, the DNA adduct formation was proportional to the time of exposure and dependent on the stage of cell growth in culture. DNA binding was detectable even at the lowest concentration used (24h exposure, 0.01 microM for both PAHs). The highest DNA adduct levels were observed after 24h of exposure in near-confluent cells (>90% of cells at G0/G1 phase), but DNA damage induced by DB[a,l]P was approximately 8-10 times higher at a concentration one order of magnitude lower as compared with B[a]P (for B[a]P at 1 microM and for DB[a,l]P at 0.1 microM: 237+/-107 and 2360+/-798 adducts/10(8) nucleotides, respectively). The induction of p53 and p21(WAF1) protein occurred subsequent to the induction of DNA adducts. The DNA adduct levels correlated with both p53 (R=0.832, P<0.001 and R=0.859, P<0.001, for DB[a,l]P and B[a]P, respectively) and p21(WAF1) levels (R=0.808, P<0.001 and R=0.797, P=0.001, for DB[a,l]P and B[a]P, respectively), regardless of the PAH exposure and the phase of cell growth. The results showed that a detectable increase of p53 and p21(WAF1) proteins (> or = 1.5-fold as compared with controls) requires a minimal DNA adduct level of approximately 200-250 adducts/10(8) nucleotides for both PAHs tested and suggest that the level of adducts rather than their structure triggers the p53 and p21(WAF1) responses. The cell cycle was altered after 12-16h of treatment, and after 24h of exposure to 0.1 microM DB[a,l]P in growing cells, there was approximately 24% increase in S phase cells accompanied by a decrease in G1 and G2/mitosis (G2/M) cells. Cell treatment with 1.0 microM B[a]P resulted in more subtle alterations. We conclude that DB[a,l]P, and to a lesser degree B[a]P, are able to induce DNA adducts as well as p53 and p21(WAF1) without eliciting G1 or G2/M arrests but rather an S phase delay/arrest. Whether the S phase delay observed in our study is beneficial for the survival of the cells remains to be further established.