Zinc transport in Synechocystis 6803
A number of bacterial metal transporters belong to the ABC transporter family. To better understand the structural determinants of metal selectivity of one such transporter, we have determined the structure of the periplasmic domain of a zinc transporter, ZnuA, from Synechocystis 6803 (blue-green algae) and have investigated structural aspects of the protein that may be involved in import regulation.
ABC Metal Transport
A class of ATP-binding cassette-type (ABC-type) transport systems is involved in the uptake of transition metal ions. This system has several homologues in various Gram positive and Gram-negative bacteria that are responsible for the transport of divalent metal cations such as Mn2+ and Zn2+, particularly at low extracellular levels of these metals. In Synechocystis 6803, the znu operon includes the znuA, znuB, and znuC genes that encode the periplasmic Zn binding protein, the integral membrane protein component, and the cytoplasmic ABC cassette domain, respectively.
Comparison of Mn and Zn binding proteins
A comparison of sequences of periplasmic solute (metal) binding proteins, SBPs classified as belonging to Cluster 9 of the ABC-type binding proteins. Identical residues are highlighted in grey. From this alignment it was not at all clear what defined the Mn vs Zn binding specificity since the proported metal binding residues were identical/similar in this alignment (orange arrows). One major, notable difference was that the Zn transporters all have a long HIS/acid loop in the middle of the protein that is not found in the Mn transporters. Initially, we determined the structure of ZnuA to determine how these proteins could be metal specific in a manner not obvious by genomic information alone.
ZnuA Structure
There are three main features of the ZnuA protein. Unlike many of the other solute binding proteins from ABC transporters, there is a long alpha-helix running along the backside of the protein (A). This likely gives the protein more rigidity and makes it unlikely that the protein undergoes large conformational changes upon solute binding and release. The bound zinc atom is clearly visible in the structure (C) and was found to be chelated by 3 HIS residues rather than the 4-residue metal binding site observed in the Mn transporters. Finally, the HIS/acid loop was mainly disordered and is immediately adjacent to the zinc binding site.
Zinc binding environment
Shown here is the electron density map at 1.9Å of the bound zinc (mauve sphere) in the ZnuA structure. While the Mn transporters had three HIS residues and an acid, ZnuA has a bound water molecule (red sphere) associated with the bound zinc instead. This is likely to be key for some of the selectivity in that Zn can bind extremely well to 3 HIS residues but Mn requires an additional set of interactions. Therefore, this is a good zinc binding site but a very poor Mn binding site. While the sequence alignment suggested that there was an acid at that water position, the structure clearly showed that it pointed out of the metal binding pocket instead.
The role of the HIS/acid loop
The structure clearly showed the high affinity zinc binding structure, however, the role of the HIS/acid loop immediately adjacent to this site was not clear. To this end, we make two varients of ZnuA; one where the entire loop was deleted and one where half the loop was deleted. We then used isothermal titration calorimetry to examine the binding properties of wt ZnuA and the loop variants. As shown below, the wt protein has two clear classes of binding sites - a high and low affinity class of sites with about 100x difference in binding affinity. From the ITC data, there are apparently 3 weakly bound Zn atoms per 1 tightly bound zinc. In contrast, there is only the high affinity site left when the loop has been deleted and about 2/3 of the weakly bound zinc atoms are lost in the half deleted mutant. This clearly showed that the loop is indeed binding zinc but does not seem to have a significant effect on the high affinity zinc binding site.
Structural changes upon zinc binding and release
We were then able to determine the structure of the deletion mutant in the presence and absence of bound zinc. It was immediately apparent that the backbone of the protein does not change significantly upon the binding and release of zinc. This is quite different from, for example, the sugar transport proteins that open and close like a large 'C clamp'. While the protein does not undergo large conformational changes, the residues ligating the Zn atom do. Two of the three binding HIS residues move into and out of the binding site during binding and release. Again, the deletion of the HIS/acid loop did not affect the Zn binding site at all.
Possible regulation of the ZnuABC complex.
The HIS/acid loop clearly does not affect the intrinsic binding of Zn to the high affinity site. What then is a possible role for this unusual feature. Shown here is one possibility. At low concentrations of zinc, the HIS/acid loop does not bind zinc and transport can occur. However, at high concentrations of zinc, it may bind to the HIS/acid loop and perhaps this interferes with the ability of ZnuA to associate with the Zn transport pore. In this way, the cell can control the intracellular concentration of zinc and not waste ATP in importing unecessary and potentially toxic levels of zinc.
Our relevant publications: