br Structural and biochemical features
Structural and biochemical features of pol X family polymerases The X-family DNA polymerase (polX) is specialized in DNA repair. This family is composed of four different DNA polymerases: pol β, pol λ, pol μ and TdT. Only three members of the polX family possess an N-terminal BRCT (BRCA1 carboxy-terminal) domain (Figure 1a) that is essential for NHEJ activity both in vitro and in vivo [29•, 30]. Pol β participates only in Swainsonine excision repair (BER) [31, 32] and is devoid of this small domain (around 11kDa), which mediates protein–protein or protein–DNA interactions . Three individual structures of a BRCT domain from the polX family were solved by NMR [34, 35]. BRCT domain is typically an αβα sandwich made up of a central antiparallel β-sheet flanked by three α-helices [36, 37, 38]. Although the structures of all known BRCT domains are highly conserved, their interaction mode with ligands is greatly variable and their role in the sequential recruitment of the different proteins during NHEJ repair is not completely understood . Ligase IV contains 2 BRCT domains whose peptide junction interacts with XRCC4, which otherwise makes filaments with XLF [40, 41] and interacts with Ku heterodimer . The BRCT domain of pol μ binds DNA  but the precise mode of binding of the BRCT domain of either pol μ or TdT to Ku heterodimer remains to be elucidated, as well as its orientation with respect to the catalytic polymerase domain. Sequence analysis of the polymerase domain of polX polymerases allows to divide this family into two subgroups . Pol β is closer to pol λ (34% of pairwise sequence identity) and TdT is closer to pol μ (44% of pairwise sequence identity). The structures of all four polymerases belonging to X-family have been solved by X-ray crystallography. Pol β was the first polX to be solved, alone and with different binary and ternary complexes [45•, 46•, 47••]. The overall structure of the catalytic domain shares the same general architecture (but not the topology) of all DNA polymerases, namely a finger domain, a palm domain and a thumb domain (Figure 1). The additional 8-kDa domain in pol β and pol λ contains a deoxyribose phosphate (dRP) lyase activity required in base excision repair (BER) of oxidative DNA damage (Figure 1). The amino acids necessary for dRP lyase activity are not conserved in TdT and pol μ, which do not participate in BER. One striking feature common to all polX is the high degree of conservation of the catalytic site, with three strictly conserved Aspartates (Figure 1b) that coordinate two essential metal ions, involved in the so-called two-metal ions mechanism, first described in  and later shown to be present also in DNA polymerases . Metal A activates the 3′OH of the last nucleotide to allow the attack of the alpha phosphate while Metal B that comes in with the incoming nucleotide triphosphate stabilizes the leaving group (PPi) (Figure 1b). In TdT, the coordination geometry of the divalent metal ions was studied in atomic detail during a full catalytic site, including transition metal ions such as Mn++, Co++ and Zn++ which are known to be more efficient than Mg++ for the nucleotidyltransferase activity. It was concluded that Metal A has to leave and be replaced by Na+ in order to allow translocation of the newly extended primer strand into a catalytically competent position for a new addition . A movie of the reaction cycle, based on thirteen different structures was built . This scenario was also described in even greater details for pol β where it was also found that binding of Na+ in Metal A binding site, after nucleotide incorporation, is a key step for DNA translocation . In addition, time-lapse crystallography showed that in pol β, there is an additional divalent ion (a third Mg++, Metal C) that comes during the reaction to counter-balance the apparition of a charge on the beta phosphate, and then leaves before the end-state is reached [50••, 51•]. This situation was also observed for pol μ, in the case of Mn++ ions .