The replication of E. coli circular DNA begins from the region called oriC. Replication forks run to both directions and they are stoped by the Tus protein (PDB:1ECR) at the region of ter which is opposite to oriC. The dimer of double-stranded DNA (dsDNA) which finished replication is recomposed in the region dif and separates into two monomers. The enzyme recognizing the dif sequence and making recombination is a complex of XerC and XerD. Furthermore, it is noted that XerCD divides the dimer into two monomers with the help of the FtsK protein. FtsK is a membrane-bound DNA translocase found in many eubacteria, such as E. coli.
The DNA translocase is an adenosine triphosphate(ATP)-dependent molecular motor which move DNA rapidly in the case of chromosome division, DNA recombination, and DNA transport. T7gp4, DnaB, SV40(PDB:1SVM), SpoIIIE, etc. belong to this type of protein. The structure of FtsK may be divided into three regions (N, linker, C). The N domain is locates on cell membrane which invaginates in cell division. The length and structure of the linker region varies depending on bacterial species. The C domain transposes DNA to the direction of a daughter cell at the rate of more than 6.7 kbp/s. At the same time, the C domain proceeds to the dif region and separates dsDNA by activating XerCD complex. FtsK expresses the function by forming a hexamer-ring. A large channel is formed in the center of each subunit and dsDNA passes through this channel.
Step A: The replication of circular DNA proceeds with two replication forks from the region of oriC to the region of ter in opposite directions.
(*) Although only two Ftsk is shown in the Fig.1, many FtsK exist on invaginated region of cell membrane.
The structures shown here are a monomer (Fig.1) and a hexamer(Fig.3, Fig.4) of the C domain of the Pseudomonas aeruginosa FtsK. The C domain is further divided into three sub-domains: α, β, and γ. The structure of the α domain is unique to FtsK, whereas the β domain has the archetypal RecA-like fold that is common to many oligomeric ATPases. The role of α domain and β domain is moving DNA by catalyzing ATP. On the other hand, γ domain is involved in the activation of XerCD (since γ domain is a glycine-rich structure and is thus highly flexible, it could not be determined). Six Ftsk monomers formed a hexamer-ring by binding to each other with in the head-head (α-α, β-β) manner. This ring contains a channel twisted clockwise in the center. This hexamer ring is formed only in the presence of dsDNA. Without dsDNA, regardless of existence of ATP and ions, FtsK exists as a monomer. The inside diameter of the channel is about 30Å. This size is larger than that of other proteins forming a hexamer ring (T7, TrwB, etc.) that acts on ssDNA (single-stranded DNA). This is due to the necessity to acommodate dsDNA(double-stranded DNA), like the beta-clamp(PDB:3BEP) and PCNA(PDB:1SXJ).
Based on the this structure, Gatti and others proposed the inchworm model which explains how Ftsk transposes DNA (Fig.2). First, DNA inserts itself from the β side of the channel and is bonded with α domain of one subunit. When β domain of the same subunit hydrolyzes ATP, structural change occurs, and the α domain pulling DNA moves by about 5.5A in the direction opposite to β domain. DNA has moved by 1.6bp the minimum by this reaction. Next, as this rebound, DNA binds with the β domain of the same subunit and separate from the α domain, and binds with the α domain of the next subunit. When each of the six subunits perfoms this step once, DNA presumably moves more than 9.6bp (1.6×6). This value is in good agreement with the pitch of dsDNA (10.5bp/1 turn). The regions which make interactions to DNA are considered to be the five loops of α-domain(300-302, 380-384) and β-domain (606-610, 633-640, 655-673) of each subunit. However, the detail of this mechanism is not yet clearly known.
Protein Data Bank (PDB)
Massey, T.H. Mercogliano, C.P. Yates, J. Sherratt, D.J. Lowe, J.; "Double-Stranded DNA Translocation: Structure and Mechanism of Hexameric Ftsk"; Molecular Cell; (2006) 23:457-469 PubMed:16916635.
author: Jun-ichi Ito