Trends in Cell Biol. 10, 159-162.
Great leaps forward: translesion synthesis gets unstalled.
This year's American Society of Microbiology meeting on "DNA Repair and Mutagenesis*" was one of a series of DNA Repair meetings (held every four years) that started 25 years ago. These specialized meetings have been seminal to the development of this important field. In his opening remarks, conference co-organizer Graham Walker (Cambridge, USA), expressed concern that learning about DNA repair at this meeting would be like "getting a drink of water from a fire hydrant." Walker's analogy could not have been more apt. This exciting meeting lasted 6 full days, included twelve plenary sessions and 4 poster sessions and was attended by almost 600 scientists. Six full days of intense scientific exchange was barely sufficient to cover the many specialized areas of DNA repair and mutagenesis, which include Base Excision Repair (BER), Nucleotide Excision Repair (NER), Transcription-coupled NER and BER (TC-NER and TC-BER), Translesion Synthesis (TLS), Mismatch Repair (MMR), Double-Strand Break Repair, Recombinational Repair........ and more. This meeting report describes just a few highlights of a meeting that was replete with superb presentations of important, interesting and novel research (see Key Conference Results).
A Profusion of Polymerases: How Many Polymerases Does It Take to Bypass a Lesion?
Replicative DNA polymerases must copy DNA templates with extraordinary fidelity, otherwise an organism's genetic information rapidly degenerates. However, such evolutionary fine-tuning to achieve exquisite DNA replication fidelity in turn prevents these polymerases from incorporating nucleotides opposite template DNA lesions, and from elongating from primers that are not properly paired with the template. As pointed out in the Keynote Address by Bryn Bridges (Sussex, UK), early speculation about UV-induced mutagenesis in Escherichia coli involving replication of a damaged template, led to a simple model: the E. coli replicative polymerase (pol III) synthesizes DNA with high fidelity on undamaged DNA, but mutagenic synthesis past a replication blocking lesion was thought to require a modified pol III, or perhaps an alternative polymerase. For many years the prevailing model has been that SOS-induced factors (RecA, LexA, UmuC, UmuD) interact with pol III to lower its fidelity, thus allowing it to bypass lesions, albeit in an error-prone, mutagenic fashion. The molecular nature of such translesion synthesis (TLS) remained strangely elusive, and now with the recent identification of a new polymerase superfamily, specialized for DNA synthesis on damaged or unusual DNA templates, the prevailing model has been replaced with one in which TLS involves recruitment of an alternative polymerase to the replication fork (Figure 1). However, it is important to note that TLS could still involve a modified pol III in certain situations. The identification and characterization of the "TLS polymerases" represents a major advance in the field of DNA repair and mutagenesis, and it was discussed extensively at this meeting.The known TLS polymerases in E. coli are pol IV (dinB) and pol V (umuCD'2); those in Saccharomyces cerevisiae are pol n(rad30) and pol z (Rev3/Rev7) and Rev1, which may or may not end up being classified as a bone fide polymerase; those in humans are pol h (XPV protein), pol z (hRev3) and hRev1. However, sequence database searches indicate that the TLS polymerase superfamily includes 36 members and extends from bacteria to yeasts, plants, worms, mice and humans1. The superfamily can be divided into four distinct subgroups related to the E. coli prototypes DinB and UmuC, or the S. cerevisiae prototypes Rad30 and Rev1. Robert Fuchs (Strasbourg, France), Zvi Livneh (Rehovat, Israel) and Myron Goodman (Los Angeles, USA) presented in vitro evidence that the E. coli DinB and UmuC proteins have intrinsic polymerase activity. Satya Prakash (Galveston, USA) and Christopher Lawrence (Rochester, USA) described the Rad30- and the Rev3/Rev7/Rev1-dependent TLS pathways in yeast. The human TLS enzymes, described by Lawrence, Prakash, Roger Woodgate (Bethesda, USA), and Valerie Gerlach (Dallas, USA), include at least three distinct human gene products homologous to S. cerevisiae Rad30, Rev3, Rev1, and one homologous to E. coli DinB. The TLS polymerases bypass abasic sites and DNA base lesions induced by UV, acetylaminofluorene, Benzo(a)pyrene, and probably many more DNA damaging agents. But which polymerases are recruited to bypass which lesions, and how, is just beginning to be explored. Importantly, the TLS polymerases so far characterized (biochemically) are all highly distributive: that is, they only synthesize extremely short patches (as short as one or two nucleotides) before falling off the DNA template. In addition, Prakash, Goodman and Thomas Kunkel (Research Triangle Park, USA) showed that even on an undamaged DNA template, the TLS polymerases have lower fidelity than even the most inaccurate polymerases previously characterized (HIV Reverse Transcriptase and DNA polymerase b), with error rates anywhere from 1 in 30 to 1 in 1000. This new family of TLS polymerases "appear" at a time when structural biologists such as Kunkel, Samuel Wilson (Research Triangle Park, USA), Lorena Beese (Durham, USA), and Tom Ellenberger (Boston, USA) are providing exquisitely detailed pictures of how "normal" polymerases achieve high fidelity, and it won't be long until the crystallographers determine how the TLS polymerases relax that fidelity.
The profusion of "new" polymerases for TLS raises a number of important questions. Just how is the DNA polymerase traffic controlled? Is the switching of polymerases a passive, stochastic process, or is it an orchestrated process analogous to the control of cell cycle checkpoints? What determines the access of different TLS enzymes to different lesions? Are different classes of DNA lesions normally bypassed by different TLS polymerases? How tightly is the number of nucleotides incorporated opposite DNA lesions controlled in vivo? Given the observation by Fuchs that inappropriately high DinB expression can dramatically increase spontaneous mutation in E. coli, the answers to these questions have important implications for genomic stability, and hence for susceptibility to cancer, aging and other degenerative diseases.
