Interstrand DNA crosslinks (ICLs) are formed by natural products of metabolism and by chemotherapeutic reagents. Work in E. coli identified a two cycle repair scheme involving incisions on one strand on either side of the ICL (unhooking) producing a gapped intermediate with the incised oligonucleotide attached to the intact strand. The gap is filled by recombinational repair or lesion bypass synthesis. The remaining monoadduct is then removed by Nucleotide Excision Repair (NER). Despite considerable effort, our understanding of each step in mammalian cells is still quite limited. In part this reflects the variety of crosslinking compounds, each with distinct structural features, used by different investigators. Also, multiple repair pathways are involved, variably operative during the cell cycle. G1 phase repair requires functions from NER, although the mechanism of recognition has not been determined. Repair can be initiated by encounters with the transcriptional apparatus, or a replication fork. In the case of the latter, the reconstruction of a replication fork, stalled or broken by collision with an ICL, adds to the complexity of the repair process. The enzymology of unhooking, the identity of the lesion bypass polymerases required to fill the first repair gap, and the functions involved in the second repair cycle are all subjects of active inquiry. Here we will review current understanding of each step in ICL repair in mammalian cells.

The bacteriorhodopsin gene has been identified in a 5.3-kilobase restriction endonuclease fragment isolated from Halobacterium halobium DNA, using a cloned cDNA fragment as the probe. Of the 1229 nucleotides whose sequence was determined in the genomic fragment, 786 correspond to the structural gene of bacteriorhodopsin, 360 are upstream from the initiator methionine codon, and 83 are downstream from the COOH terminus. The bacteriorhodopsin gene codes for a precursor sequence of 13 amino acids at the NH2 terminus, 248 amino acids that are present in the mature protein, and an additional aspartic acid at the COOH terminus. This determination of the DNA sequence for an archaebacterial gene reveals that the standard genetic code is used; however, there is a marked preference for either G or C in the third codon position. The gene does not contain any intervening sequences and no prokaryotic promoter can be identified in the region immediately upstream from the structural gene. The bacteriorhodopsin mRNA contains at the 5' terminus only three nucleotides beyond the initiating AUG codon and this terminus can form a hairpin structure. Immediately downstream from this structure there is a sequence complementary to the 3' terminus of H. halobium 16S rRNA.Bacteriorhodopsin, the only protein in the purple membrane of Halobacterium halobium, catalyzes the light-dependent translocation of protons and thus generates a transmembrane electrochemical gradient (1). The protein consists of a single polypeptide chain of 248 amino acids (2) and contains one molecule of retinaldehyde per protein linked as a Schiff's base to the a-amino group of a lysine residue (1, 3). The amino acid sequence ofthe protein is known (2, 4), and a three dimensional model has been developed (5) that is compatible with this sequence and the diffraction data (6, 7). According to this model the polypeptide chain traverses the membrane seven times in the form of a-helical rods.In view of these and other studies (1)

Triple helix forming oligonucleotides (TFOs) recognize and bind sequences in duplex DNA and have received considerable attention because of their potential for targeting specific genomic sites. TFOs can deliver DNA reactive reagents to specific sequences in purified chromosomal DNA (ref. 4) and nuclei. However, chromosome targeting in viable cells has not been demonstrated, and in vitro experiments indicate that chromatin structure is incompatible with triplex formation. We have prepared modified TFOs, linked to the DNA-crosslinking reagent psoralen, directed at a site in the Hprt gene. We show that stable Hprt-deficient clones can be recovered following introduction of the TFOs into viable cells and photoactivation of the psoralen. Analysis of 282 clones indicated that 85% contained mutations in the triplex target region. We observed mainly deletions and some insertions. These data indicate that appropriately constructed TFOs can find chromosomal targets, and suggest that the chromatin structure in the target region is more dynamic than predicted by the in vitro experiments.

Bloom syndrome and Werner syndrome are genome instability disorders, which result from mutations in two different genes encoding helicases. Both enzymes are members of the RecQ family of helicases, have a 3 3 5 polarity, and require a 3 single strand tail. In addition to their activity in unwinding duplex substrates, recent studies show that the two enzymes are able to unwind G2 and G4 tetraplexes, prompting speculation that failure to resolve these structures in Bloom syndrome and Werner syndrome cells may contribute to genome instability. The triple helix is another alternate DNA structure that can be formed by sequences that are widely distributed throughout the human genome. Here we show that purified Bloom and Werner helicases can unwind a DNA triple helix. The reactions are dependent on nucleoside triphosphate hydrolysis and require a free 3 tail attached to the third strand. The two enzymes unwound triplexes without requirement for a duplex extension that would form a fork at the junction of the tail and the triplex. In contrast, a duplex formed by the third strand and a complement to the triplex region was a poor substrate for both enzymes. However, the same duplex was readily unwound when a noncomplementary 5 tail was added to form a forked structure. It seems likely that structural features of the triplex mimic those of a fork and thus support efficient unwinding by the two helicases.Despite the obvious importance of genetic stability, the mammalian genome has an abundance of sequences that are potentially destabilizing. Elements that can form non-duplex structures, such as DNA triple helices (1, 2), G quartets (3-5), hairpins (6, 7), and cruciforms (8), all have the capacity to interfere with transcription and replication. Moreover, many studies have described the role of these elements in DNA rearrangements such as deletions, sister chromatid exchanges, homologous and illegitimate recombination events, etc. (reviewed in Refs. 9 and 10). With such an array of provocative sequence elements, it seems likely that cells would have developed the capacity for controlling the potential of these sequences for genome destabilization.Some insight into the nature of the enzymology involved in the maintenance of genetic integrity has come from the study of two "genome instability" disorders, Bloom and Werner syndrome. Patients with Werner syndrome (WS) 1 are characterized by premature aging (11) and a high incidence of certain cancers. They have elevated frequencies of spontaneous deletion mutations in the HPRT gene (12) and show a variety of karyotypic abnormalities including inversions, translocations, and chromosomal losses (13,14). The mutant gene in WS has been identified as a member of the RecQ helicase family, and the protein is a 3Ј-5Ј-helicase (15, 16) and a 3Ј-5Ј-exonuclease (17-19). A role in replication is implied by the demonstration that the WRN helicase interacts with and is stimulated by the human replication protein A (RPA) (20, 21). The protein has been recovered from cells in a replication comp...