Crystal interactions

Here I study interactions of one macromolecular unit with other units in crystal structure of XRC experiment.

MerA 1ZK7 structure

Crystal resume

MerA is a homodimeric protein. 1ZK7 entry contains only one chain A (fig. 1), while the second part of biological assembly is generated by the two fold axis: -Y, -X, -Z+1/2 (fig. 2). However, there wasn't found any PyMol function to restore only biological assembly and apply symmetry to the dimeric unit as a whole. Unit cell is a parallelepiped with sides: 86.788 Å, 86.788 Å, 136.823 Å.

Fig. 1. Surface representation of 1ZK7 MerA monomer.

Fig. 2. Surface representation of 1ZK7 MerA homodimer, biological assembly.

Unit interactions

5 Å symmetric neighbours were reconstructed with the following PyMol command: symexp sym_5_, 1zk7, 1zk7, 5;. Due to indistinguishible homodimers in symmetry it is almost impossible to inspect macromolecular structure by sight (fig. 3).

Fig. 3. Surface representation of 5 Å neighbourhood. Green — original unit, blue — restored symmetric units. Unit cell is drawn as green wireframe parallelepiped.

Deep inspection of the pdb file has revealed 8 symmetric operators for 1ZK7 that represent 4 operators for one subunit and 4 operators for the other subunit of the dimeric structure. After visual inspection, symmetric unit sym_5_07000000 was defined as the dimeric twin of the origin subunit.

Interacting residues of the original subunit and its neighbourhood were defined to be placed within 3.5 Å from the other subunit with the following set of commands:

select dimer1, 1zk7;
select side1, byres (dimer1 and (sym_5_* around 3.5));
select side2, byres (sym_5_* and (dimer1 around 3.5));

See fig. 4 for visualization.

Fig. 4. Surface representation of 5 Å neighbourhood. Green — original unit, magenta — homodimeric twin, blue — other restored symmetric units. Putatively interacting residues (within 3.5 Å from the original unit and vice versa) are colored as follows: blue — oxygen, red — nitrogen, unit color — carbon.

Visually, original subunit interacts most with its dimeric counterpart. To prove this, I calculated number of atoms in restored units involved in interacting residues with the following commands:

syms = cmd.get_object_list('sym_5_*');
for sym in syms: print(sym, cmd.count_atoms('%s and side2' % sym));

Excecution has proven that original unit interacts most with its dimeric counterpart (table 1).

Dimer-comprising interactions in detail

Interacting residues in the dimer are shown in fig. 5.

Fig. 5. Surface representation of MerA homodimer with putatively interacting residues. Color scheme as in fig. 4.

To define the type of interaction between dimeric subunits I sought to find putative hydrogen bonds. They were found as donor and acceptor atoms from different subunits at the distance not greater than 3.2 Å with the following set of commands:

select dimer2, sym_5_07000000;
select side3, byres (dimer1 and (dimer2 around 3.5));
select side4, side2 and dimer2;
dist HBA, (side3 and acc and pol), (side4 and don and pol), 3.2;
dist HBD, (side3 and don and pol), (side4 and acc and pol), 3.2;

See fig. 6 for visualization of discovered hydrogen bonds.

Fig. 6. Putative hydrogen bonds supporting homodimerization of MerA.

The number of the hydrogen bonds was calculated as a sum of donor atoms of the one unit within 3.2 Å from acceptor atoms of the other unit and vice versa:

count_atoms (side3 and don and pol) and ((side4 and acc and pol) around 3.2); # 12 atoms
count_atoms (side3 and acc and pol) and ((side4 and don and pol) around 3.2); # 13 atoms

It totally yielded 25 hydrogen bonds connecting subunits into dimer. Thus I claim those interactions in crystal as biologically relevant.

Odd elements of 3HDD structure

Several terminal protein residues are oddly located in 3HDD structure (fig. 7).

Fig. 7. An odd protein chain at the left bottom corner of the image.

The crystal was recovered with asymmetric units at 10Å proximity (fig. 8).

Fig. 8. 3HDD crystal structure. Authentic asymmetric unit is coloured, restored asymmetric units are achromatic.

Close investigation of that odd terminus (chain B, residues 1–10) indicated proximity of the terminus to the protein part of one asymmetric unit and to the DNA part of another asymmetric unit (fig. 9). The protein counterpart consists of residues 22, 23 and 30, chain B. The DNA counterpart consist of nucleotide 342, chain B, and nucleotides 203 and 204, chain A.

Fig. 9. Interactions between macromolecular complexes in 3HDD crystal are formed by "odd" protein chain, protein loop of one asymmetric unit and DNA nucleotides of another asymmetric unit. Interacting residues are shown as spheres.