Purification of Mitochondria for Enzymes Involved in Nucleic Acid Transactions

Ever since the demonstration that mitochondria contain
DNA
(mtDNA) and a complete apparatus for its replication and expression there has been an interest in identifying the enzymes involved in these processes. This problem is compounded by the fact that similar DNA transactions also occur in the nucleus, where vastly larger quantities of enzymes deal with 1000 times as much DNA. Thus, it has been a consistent challenge to the field to purify a distinct mitochondrial enzyme activity and prove that it is a
bona fide
mitochondrial enzyme. When mtDNA genomes were sequenced in the 1980s and open reading frames assigned to metabolic enzymes, it became clear that
all
proteins involved in mtDNA replication, transcription, and repair are products of nuclear genes that must be imported into mitochondria. Our laboratory has had a long-standing interest in these mitochondrial enzymes. When we first began to approach this problem, the best estimates in the literature were that the mtRNA polymerase had a molecular weight estimated at 66-68 kDa in rat liver (
1
) or 46 kDa in
Xenopus
oocytes (
2
); similarly, some early estimates of the size of mtDNA polymerase suggested that it was a homotetramer of 47-kDa subunits (
3
,
4
). Both the mtRNA polymerase and the catalytic subunit of DNA pol γ are now known to be approximately 125-140 kDa in various organisms (
5
–
12
). The possibility that proteolysis contributed to the early underestimate of the molecular masses of these proteins underscores the need to work quickly and to use a complete set of protease inhibitors during purification of these enzymes. In recent years, the effort to characterize the mitochondrial complement of DNA metabolic enzymes has benefited enormously from the expansion of genetic databases that provide a gold standard in quality control to assure that purified proteins have the properties predicted by primary sequence information.