Thesis (Ph. D.)--Massachusetts Institute of Technology, Division of Bioengineering and Environmental Health, 2002.Includes bibliographical references (leaves 176-200).The study of human mutagenesis requires methods of measuring somatic mutations in normal human tissues and inherited mutations in human populations. Such methods should permit measurement of rare mutations in the presence of abundant wild-type copies and should be general to the human genome. A sensitivity of 2 x 10-6 for point mutations was recently achieved in human cells using a novel method of target isolation, constant denaturant capillary electrophoresis (CDCE), and high-fidelity polymerase chain reaction (hifi-PCR) (Li-Sucholeiki and Thilly, 2000). This method is applicable to 100-base pair (bp) DNA domains juxtaposed with a naturally occurring domain of a higher melting temperature, or a natural clamp. Such sequence domains represent about 9% of the human genome. To permit analysis of rare point mutations in the human genome more generally, this thesis developed a procedure in which a clamp can be ligated to any 100-bp sequence of interest. This procedure was combined with the previous method to create a new method of point mutational analysis that is not dependent on a naturally occurring clamp. To demonstrate the new method, a sequence with a natural clamp, a part of the human hypoxanthine-guanine phosphoribosyl transferase (HPRT) gene (cDNA-bp 223-318), was analyzed using both the natural and ligated clamps. A sensitivity of 2 x 10-5 in human cells was demonstrated using the ligated clamp as opposed to 5 x 10-6 using the natural clamp.(cont.) The sensitivity of the new method using the ligated clamp was demonstrated to be limited by the fidelity of Pfu DNA polymerase used for PCR. The sequence of the ligated clamp accounted for the differences in sensitivity as a result of causing a decreased efficiency of mutant enrichment by CDCE. The new method can be applied to measure somatic mutations in normal human tissues, such as lung tissues, in which point mutations at fractions above 10-5 have been observed. This method can also detect predominant inherited mutations even for genes carrying recessive deleterious alleles in pooled samples derived from a large number of individuals.by Andrea Seungsun Kim.Ph.D.