Research Article: Manipulation of an existing crystal form unexpectedly results in interwoven packing networks with pseudo-translational symmetry

Date Published: October 01, 2016

Publisher: International Union of Crystallography

Author(s): Janice M. Reimer, Martin N. Aloise, Harold R. Powell, T. Martin Schmeing.


A nonribosomal peptide synthetase di-domain construct was produced using known crystal packing as a guide, and the resulting crystal has an unanticipated packing.

Partial Text

Nonribosomal peptides (NRPs), which are small molecules that are produced by nonribosomal peptide synthetases (NRPSs), cover an enormous expanse of chemical space and have uses ranging from green chemicals to important antibiotics (Walsh, 2004 ▸). The large diversity in NRPs comes from the ability of NRPSs to use >500 monomers as substrates and to co-synthetically introduce chemical modifications into these substrates.

We recently solved the structure of the F-A didomain construct to 2.5 Å resolution in a crystal belonging to space group P41212 with an apparent solvent content of 82% (Reimer et al., 2016 ▸). No electron density is visible for residues C-terminal to the four-residue hinge region connecting the Acore subdomain to the Asub subdomain (Supplementary Fig. S1) and adjacent, symmety-related molecules blocked previously observed positions of the Asub subdomain (Reger et al., 2008 ▸; Conti et al., 1997 ▸; Yonus et al., 2008 ▸; Tanovic et al., 2008 ▸; Gulick, 2009 ▸; Mitchell et al., 2012 ▸). We had also crystallized F-A-PCP, but initially were only able to obtain poorly diffracting (>9 Å resolution) crystals. Our principal interest in the F-A-PCP construct was to observe the interactions between the F and PCP domains that enable the valine substrate attached to the PCP domain to be formylated. Because the Asub subdomain (∼11 kDa) is not involved in the crystal contacts and was not known to be required for the F domain–PCP domain interaction, we reasoned that removal of the Asub subdomain could lead to a crystal that had large solvent channels, unoccupied by disordered protein regions, that would be suitable for soaking experiments using an entire purified PCP domain (∼8 kDa), for co-crystallization or as the basis for crystals of a future F-AΔsub–linker–PCP construct.




Leave a Reply

Your email address will not be published.