During several DNA metabolic reactions, double-stranded DNA is converted to single-stranded DNA temporarily by the actions of DNA helicase or exonucleases. The exposure of a single-stranded DNA region is necessary for enzymes or the complementary DNA strand to interact with the region, but this also produces regions that form unfavorable secondary structures and are susceptible to attack by nucleases. To avoid such problems and make the enzyme reactions efficient, all organisms have single-stranded DNA binding proteins (SSBs) that preferentially bind only single-stranded DNA. Due to their involvement in vast tasks, they appear to be multifunctional and one of the major players in various cellular processes.
The first discovery of SSB was bacteriophage T4 gene32 protein. Affinity chromatography of T4 phage-infected cell lysate on single-stranded DNA cellulose successfully identified a protein that bound strongly and preferentially to single-stranded DNA (1). Genetic studies had shown that the gene product was essential for recombination and DNA replication, indicating the importance of this type of DNA-binding protein in these processes. Following this discovery, many SSBs were identified from various systems, eg, from Escherichia coli, bacteriophages, and viruses. Although there are almost no similarities in their protein structures, they commonly bind to single-stranded DNA with more than 10 times greater affinity than to duplex DNA and have a crucial rule in DNA metabolism. In general, their binding to single-stranded DNA is cooperative; this means that each molecule prefers to bind adjacent to another molecule along the DNA strand. As a result, many molecules of SSB coat DNA and make it an active target for many enzymes (2). This aspect of SSBs sometimes induces the denaturation of duplex DNA under physiological conditions, so that they were initially called unwinding or helix-destabilizing proteins. This function is distinguishable from DNA helicase action, since the denaturation of duplex DNA by SSB does not require ATP hydrolysis. SSB interacts with the backbone of single-stranded DNA rather than the bases. Thus, the binding is almost nonspecific for the DNA sequence, and the bound DNA is extended. The extended conformation facilitates base pairing of complementary strands and increases the rate of DNA polymerase migration; this function explains the requirement of SSB for replication and recombination.
SSBs have been sought from eukaryotic cells, and several single-stranded DNA-specific binding proteins were identified, but none of them were obviously demonstrated to be involved in DNA metabolism. The essential cellular SSB was first identified from human cell extracts as a component essential for in vitro SV40 virus DNA replication and named RPA (replication protein A, also called RFA or HSSB; for a review, see (3)). During the replication reaction, RPA interacts with the viral-coded initiator protein, large T Antigen, and in collaboration with its helicase activity forms a replication origin unwound complex (Fig. 1). Another interaction between RPA and DNA polymerase a/primase facilitates the recruitment of this polymerase on the unwound DNA and initiates DNA synthesis. In the following DNA elongation step, RPA stimulates DNA polymerases through either specific or nonspecific interactions (4). Therefore, RPA has critical roles throughout this viral DNA replication and will have the same function during chromosomal DNA replication. Related to this point, studies by immunostaining with anti-RPA antibodies have revealed the tight interaction of RPA with the replication apparatus. RPA is localized in the nucleus, exists in the replication foci prior to the initiation of DNA synthesis, and remains there during S phase (5).
Figure 1. Unwinding of the SV40 virus replication origin by SV40 T-antigen and RPA.
Unlike prokaryotic SSBs, which are monomers or oligomers of identical subunits, human RPA has three heterogeneous subunits of 70, 32, and 14 kDa. This oligomeric structure is conserved among eukaryotic SSBs, so homologous SSBs in heterotrimeric complexes have been identified in all tested eukaryotes from yeast to human (3). Since eukaryotes have several specific cellular processes, such as the mitotic cell cycle, DNA damage signaling, chromatin formation, and transcription activation, this complex structure of RPA might reflect its extra roles in these processes, in addition to DNA metabolism. For example, RPA interacts directly with nucleotide excision repair proteins XPA, XPG, and XPF (xeroderma pigmentosa group A, G, and F proteins) and plays crucial roles in damage recognition and cleavage (6). Another type of RPA-interacting protein includes transcription activators such as GAL4, VP16 (7), and tumor suppressor p53 protein (8), suggesting the involvement of RPA in transcription regulation. Indeed, RPA has been isolated as a transcription factor for several genes in yeast (9). In addition to protein-protein interactions, the 32k-Da subunit of RPA is phosphorylated in a cell-cycle-dependent manner and by DNA damage. An up-shift of the mobility of the 32-kDa subunit in gel electrophoresis was observed upon the phosphorylation that takes place efficiently in the presence of single-stranded DNA (10). Although the phosphorylation of RPA has been studied intensively, no direct evidence to connect it with the modulation of RPA functions could be obtained. However, if we consider this specific phosphorylation, the localization of RPA in nuclei, and its interaction with several important factors for various cellular processes, we see that eukaryotic single-stranded DNA-binding proteins are not merely DNA metabolic proteins, but will also play a role in coordinating DNA metabolism with several cellular processes.
