Although RPA is well studied, the functions of its individual subunits and multiple DBDs remain obscure. For example, it is not known what combination of DBDs account for the major ssDNA binding mode. Estimates of the occluded binding-site size of several species of RPA range between 22 and 30 nt (14
), and experiments with both yeast and human RPA indicate that this 30 nt binding mode is achieved by RPA directly interacting with 20-30 nt of ssDNA (12
). It is known that this binding mode involves an initially unstable interaction with 8 nt of ssDNA that resolves into a stable elongated complex covering 30 nt (10
). Despite the consistency of these values, we have considered the possibility that RPA binds ssDNA in a second stable mode. This idea is proposed to reconcile the following facts: the prokaryotic cellular SSB of E. coli
is a homotetramer that binds ssDNA in at least two modes (35 nt and 65 nt) (31
); both RPA and the prokaryotic SSB contain multiple DBDs (19
); RPA2 is an essential ssDNA binding subunit whose function is unexplained (23
); and a 90 nt binding mode was previously reported for yeast RPA (17
Recent structural analysis of RPA has provided sufficient details on the mechanism of ssDNA binding to allow us to test this hypothesis. The crystal structure of domains A and B has been determined in the presence (21
) and absence of ssDNA (28
). These domains, each comprising an OB-fold, reorient upon binding ssDNA and interact with a total of 8 nt. The solution structure of human RPA1N also revealed an OB-fold-like structure (22
), but this domain is not known to bind ssDNA and may be required to mediate interactions with other proteins due to its interaction with other proteins (34
) as it fails to bind ssDNA on its own (19
) and deletion of RPA1N does not affect its activity in vitro
). The C-terminal portion of RPA1 is a third ssDNA binding domain that binds zinc and appears to contain another OB-fold (DBD-C) (19
). Finally, the structure of a sub-complex consisting of the RPA2 core bound to RPA3 revealed OB-folds in each of these domains (32
). However, only the fold in RPA2 (DBD-D) resembles domains A and B and only RPA2 is known to bind ssDNA in vitro
). Thus, RPA consists of six potential ssDNA binding domains of which four are known to bind ssDNA.
To explain the role of the multiple DBDs in the mechanism of ssDNA binding by RPA we considered the following two models. The simplest idea, that all six DBDs are required for stable ssDNA binding, is difficult to support as RPA1N and RPA3 are not known to bind ssDNA. A second model proposes that the four known DBDs are required for the stable 30 nt binding mode while the remaining two domains mediate protein-protein interactions. Bochkarev and colleagues have recently proposed a detailed version of this model (28
). In this model domains A, B, and C, align in a linear fashion and contact 13 - 15 nt. DBD-D is then proposed to align with these domains such that a total of 18 - 20 nt of ssDNA is contacted by RPA. Domains RPA1N and RPA3 are proposed to account for the observed occlusion of 30 nt (28
Some, but not all, of our data are compatible with this model. By inactivating each DBD of scRPA and measuring the apparent association constant of the resulting complex we have determined that DBD-A is essential for RPA to interact with (dT)12 and that DBD-B and -C are required for full binding affinity to this substrate. Mutation of DBD-D had no effect on the affinity of RPA for (dT)12. Consistent with the above view, we interpret this to mean that domains A, B, and C make contact with (dT)12 while DBD-D does not. In contrast to the predictions of this model, mutation of DBD-D had no significant effect on RPA’s binding affinity to 17 or 23 nt substrates. The binding affinity of RPA-D− was significantly impaired only when the substrate size was increased to 40 or 60 nt. These data indicate that DBD-D binds ssDNA in a 40 or 60 nt mode and is unlikely to be involved in the well-characterized 30 nt binding mode.
Our results, as well as the earlier report of a 90 nt binding mode (17
), were obtained with scRPA not hsRPA. We do not believe, however, that this larger binding mode is unique to scRPA. The contribution of RPA2 to ssDNA binding may have gone undetected because it interacts only with long substrates. As a result, more sensitive assays are required to detect RPA2 binding as it does not significantly affect the overall binding affinity of RPA to these substrates. This idea can explain the discrepancy in ssDNA binding site size and earlier evidence that RPA1 possesses all the ssDNA binding activity of the trimer. In this model, the occlusion of 30 nt arises from the interaction of DBDs A, B, and C with 23 nt. Binding by RPA2 (DBD-D) does not occur until the substrate is 40 nt or more which could account for the occlusion of 60 nt or more. The binding by RPA1 alone is expected to be sufficient for the stable 30 nt binding mode. Indeed, the affinity of RPA-D−
for a 23 nt substrate (Ka
= 1.7 × 10E9) is in the same order of magnitude as that obtained with wt RPA and a 30 nt substrate (Ka
= 4.6 × 10E9) (13
We previously described a simple UV crosslinking assay to identify interactions between ssDNA and the RPA subunits. This assay suggested that the binding of ssDNA by RPA2 occurred with low efficiency and that it could be stimulated by increased concentrations of NaCl (23
). By including an immunoprecipitation step in the experiments described here we have found that the interaction between ssDNA and RPA2 is more efficient than originally thought. This result is consistent with the fact that dimeric (DBD-D/RPA3) or trimeric (DBD-C/-D/RPA3) subcomplexes of hsRPA bind ssDNA with relatively high affinity (25
). We suggest that it is inherently difficult to identify an interaction between RPA2 and ssDNA in the context of wt RPA because binding by RPA2 requires prior binding by the potent RPA1 subunit. This idea is confirmed by crosslinking studies; immunoprecipitated RPA2 was associated with a significant amount of RPA1 that was itself bound to labeled ssDNA. As above, a second difficulty in identifying this interaction is that substrate must be at least 40 nt in length before DBD-D is able to contact it. If the stable 30 nt binding mode required RPA2, then significant crosslinking would be expected with this substrate. However, crosslinking was not obtained with 30 or 35 nt substrates. We observed only weak interactions with the 40 nt substrate and strong interactions with the 60 nt substrate. These data support the model in which RPA1 is exclusively responsible for the 30 nt mode and that DBD-D promotes a second, larger binding mode. An alternative explanation for the improved crosslinking of RPA2 to the 60 nt substrate is that two RPA trimers bind this substrate and undergo a conformational change that results in the interaction between ssDNA and RPA2. However, this explanation can be excluded as we fail to see doubly-occupied 60-mers under the stoichiometric conditions used in the crosslinking experiment.
A mutational approach has previously been used to study the role of DBDs A and B in hsRPA (39
). Walther and colleagues concluded that mutating a single aromatic residue of DBD-A or B had a minimal effect on the overall ssDNA binding affinity of hsRPA to (dT)30 (39
). Our results are in partial agreement with this study as we found that RPA-A−
were equally defective in binding (dT)23. In addition, both studies revealed a synergistic effect in simultaneously mutating domains A and B. In contrast, we found that the double aromatic mutation had a more profound effect than the single point mutation. The affinity of RPA-A−
for (dT)23 or (dT)40 was 1/20 that of wt RPA () while the affinity of RPA containing a single aromatic mutation in domain A (F238A) for (dT)30 was 2/3 that of wt RPA (39
). This defect was amplified when binding to smaller substrates was examined; binding of RPA-A−
to substrates such as (dT)12 was not detectable (). We conclude that mutating both aromatic residues significantly reduces the activity of a single DBD. In addition, it is important to consider substrate size when determining the effects of these mutations as some effects are masked by the activity of additional DBDs within the RPA complex.
Crosslinking of labeled ssDNA to RPA results in a 97 kDa complex that represents RPA2 crosslinked to RPA1 (). This species is under-represented when using (dT)12 - (dT)35, even though these oligos bind well to RPA1. Thus, RPA1 and RPA2 are poorly crosslinked to each other under these conditions. The dramatic increase in signal that occurs with substrates of 40 to 60 nt suggests that the interaction of RPA2 with ssDNA increases the probability of a crosslink between RPA1 and RPA2. This is consistent with previously observed rearrangements that occur upon ssDNA binding. The binding of ssDNA to RPA has previously been shown to result in the alignment of domains A and B (28
), increased proteolysis of RPA2 (40
), and increased phosphorylation of RPA2 by DNA-PK (11
). Interestingly, Blackwell and colleagues observed that while phosphorylation of RPA by DNA-PK was dependent on ssDNA binding, there was a significant increase in modification as the ssDNA template was increased from 30 to 45 nt (11
). We suggest that this modification, like the crosslinking of RPA1 and RPA2, is due to the interaction of RPA2 with ssDNA and rearrangement of the heterotrimer.
The crosslinking assay described here will allow us to further examine the role of NaCl and other factors in modulating ssDNA binding by RPA2. The function of this binding is still unclear. Unlike domain A, which has a significant effect on ssDNA binding affinity, RPA2 can contribute only a small amount to the overall binding affinity of RPA. One possibility is that this small degree of binding affinity is significant in vivo
given that RPA2 is essential for viability in yeast. On the other hand, ssDNA binding by RPA2 might control RPA’s cooperativity or its interaction with other proteins. This function may in turn be regulated by the cell-cycle and DNA damage-dependent phosphorylation of RPA2 (41
). In light of the present results, it is not surprising that phosphorylation of RPA2 did not significantly affect the ssDNA binding activity of RPA in vitro
). Changes in this activity would be expected to have a small affect with large substrates and no effect with small substrates. An alternative role for ssDNA binding by RPA2 could be to mediate the compaction of RPA-ssDNA complexes that has been observed by electron microscopy at high salt (44
). Further experimentation will be required to determine whether RPA2 or its modifications affect these activities.