Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)


Pharmaceutical Sciences

First Advisor

Bongsup Cho


Cancer is the second deadliest disease in the United States. Over 100 different types of cancers exist, among which lung, breast and prostate cancers are those most frequently diagnosed. Genetic factors are important. However, exposures to tobacco smoke and environmental pollutants are considered to be responsible for 75%–80% of cancer. About 6% of cancer deaths every year in the US are reportedly to be directly linked to known carcinogen exposures. Therefore, it is important to study the mechanisms of how the environmental carcinogens trigger cancer initiation. Most chemical carcinogens are metabolized into reactive species in vivo to interact with DNA, consequently producing covalent DNA adducts. These harmful lesions can be removed by various repair systems including base excision and nucleotide excision repair machinery in the cell. However, unrepaired lesions can enter into cell’s DNA replication cycle and generate various point and frameshift mutations. In particular, the latter represents a gain or loss of base pairs, which alters the genome information. As an example, mutations on the specific genes such as the tumor suppressor p53 may trigger cancer initiation.

Arylamine is known as an important group of environmental chemical carcinogens. Some members of this group, such as 4-aminobiphenyl (ABP), benzidine and 2-naphthylamine, are classified as human bladder carcinogens. These chemicals are found commonly in cigarette smoke, incomplete diesel exhausts, and hair dye products. 2-Aminofluorene is a prototype animal carcinogen that undergoes metabolic activation by liver enzymes to form electrophilic nitrenium ion to form two major C8 substituted DNA-adducts: N-(2'-deoxyguanosin-8-yl)-2-aminofluorene (dG-C8-AF) and N-(2’- deoxyguanosin-8-yl)-2-acetylaminofluorene (dG-C8-AAF). Similarly, the human carcinogen ABP produces N-(2’-deoxyguanosin-8-yl)-4-aminobiphenyl (dG-C8-ABP). Encountering these lesions in a replicative or a bypass polymerase will result in different types of biological outcomes, such as error-free, error-prone, or frameshifts.

Manuscript I (published in Chemical Research in Toxicology, 2012) is a rapid report. In this communication, we used a real-time, label-free chip-based technique named surface plasmon resonance (SPR) to determine the binding interaction between the DNA replicative polymerase exonuclease-free Klenow fragment and three arylamine DNA lesions (FAF/FAAF/FABP). We designed biotin labeled DNA hairpin construct with modified lesions and immobilized the DNA on the streptavidin coated chip. The analyte Kf-exo- was added over the DNA surface in the presence or absence of dNTP. The results showed a tight binding between the enzyme and unmodified DNA with great dNTP selectivity. In contrast, the dNTP selectivity was minimal in adduct modified DNA. Moreover, lesion included DNA tended to have better and stronger binding than unmodified DNA.

Manuscript II (published in Chemical Research in Toxicology, 2014) contains the full details of Manuscript I. The full paper involves two 5’-flanking sequence (CG*A and TG*A), two adducts (FAAF and FABP), and two different polymerases (E.coli replicative polymerase Kf-exo- and human repair polymerase B). We employed the same SPR methodology to study the binding interaction and complementary 19F NMR and primer steady-state kinetics. Results showed significant substrate specificity for Kfexo- and polymerase B, which are double-stranded/single-stranded junction and a doublestranded DNA with a nucleotide gap structure, respectively. Tight binding with nativeDNA was observed, as well as the high nucleotide selectivity. However, Kf-exo- binds tightly to lesion DNA, but not for polymerase B. A minimal nucleotide selectivity for modified was observed with both enzymes. Moreover, the dynamic 19F NMR and primer steady-state kinetics results indicated the importance of lesion-induced conformational heterogeneity in polymerase binding.

In Manuscript III (to be submitted to Journal of Molecular Biology), we conducted a series of systematic studies to probe the conformational mechanisms of arylamine-induced -2 base deletion mutations frequently observed in the NarI mutational hot sequence (5’---TCGGCG*CN---3’; N= dC and dT) of E. coli during translesion synthesis (TLS). We employed two well-characterized fluorinated bulky DNA lesions FAAF and FABP that were derived from the environmental carcinogens 2-aminofluorene and 4-aminbiphenyl. Our work focused primarily on elucidating the effects of lesion size, bulkiness, and overall topology and the 3’-next flanking base N in producing the bulge structure responsible for -2 frameshift mutations. Two chemical simulated TLS models were examined, in which the FAAF/FABP lesion is positioned at G3 position of two 16-mer NarI sequences, which were annealed systematically with increasing primer lengths in the full length and -2 deletion pathways. Their thermodynamic, conformational, and binding profiles at each elongation step were measured by various biophysical techniques including spectroscopic (dynamic 19F NMR/CD), thermodynamic (UV-melting/DSC), and affinity binding (SPR). Results showed two different -2 bulge formations, which are triggered by the conformational stability of the G3*: C base pair at the replication fork, as well as the nature of base sequences surrounding the lesion site. Each bulge structure exists in a mixture of “external solvent exposed” B-type (B-SMI) and “inserted solvent protected “stacked” S-type (S-SMI), and their conformational rigidity increases as a function of primer lengths. The results indicate the importance of conformational stability, heterogeneity, and flexibility in the mechanisms of bulky arylamine-induced frameshift mutagenesis.