Guest-protein incorporation into solvent channels of a protein host crystal (hostal)

This study reports the incorporation of small guest proteins by diffusion (soaking) into the solvent channels of a host protein crystal, visualized in time and space by microscopic techniques. The results represent a first step towards the realization of a protein-based crystal host (‘hostal’) system that may be used in the future as a carrier/entrainment system for protein guests and/or for guest-protein structure determination.


S2.2. Estimation of guest content for soaking with high concentrations of CaM
The guest content estimation described here corresponds with the CaM soaking results described in the main text result section 3.3.3. The Alexa532-CaM fluorescence intensity within host crystals soaked for 2 weeks is 240 AU, which is above the detection range. This result suggests an enrichment of the guest within the crystals that was not observed at low concentrations of CaM. As the reservoir solution with 110 mg/ml CaM with 0.1% Alexa532-CaM gives a fluorescence of 60 AU (Fig. 5d) and background fluorescence of CaM without dye is neglectable, the fluorescence from the crystal corresponds to a CaM concentration > 400 mg/ml. Assuming this relationship between fluorescence and concentration is correct also in the crystal as no spectral change is observed, the minimum concentration of 400 mg/ml CaM in the crystal, as derived from to equation (2), results in:

S2.3. Attempts to solve guest structures (molecular replacement)
The data set for the CytC-soaked crystal processed in P6 1 22 and in P1 used for the reported structure determinations was also used to attempt molecular replacement using the structure of CytC (PDB ID: 1HRC) as search model including ds-TrpR as fixed model, but no reasonable solution could be found. The same was the case with the data set for the CaM-soaked crystals using the calcium-free structure of CaM (PDB ID: 1CFD) or the separate N-or C-terminal domains from PDB 1CFD as search models. Although the guest structures could not be resolved, differences were observed in the data processing, structure refinement statistics, and channel electron density compared with the non-soaked structure. The raw diffraction data for soaking with both guests will be made publicly available to the community.

Table S6
Variations in cell dimension and merging statistics between data sets after processing in P1 and

Figure S1
Soaking procedures for guest protein incorporation and imaging. Crystals grown in hanging drops were soaked in two different ways. 1) For soaking without controlling the concentration of the guest protein, the guest was added as lyophilized powder directly to the crystal drop on the cover slip that was transferred back to the hanging drop set-up to prevent evaporation (not shown). 2) Soaking stock solution with controlled guest protein concentration was added to a cover slip and crystals were transferred into the guest-protein drop using crystal loops. The cover slip was back-transferred to the hanging drop set-up to prevent evaporation (not shown). Imaging of the crystals using a brightfield microscope was done on drops in their well setups. For imaging using CLSM the crystals were transferred either by pipette or crystal loops into the well of a microscope slide pre-filled with reservoir or soaking solution and covered by a cover slip.

Figure S6
END maps and 2Fo-Fc +/-bulk solvent correction maps. END maps (upper three panels) were calculated as described in Methods for the indicated data sets (full processing statistics in SI, Tables S7-9) and the channel density is presented at an absolute contour level of 0.45 electrons per unit area (e -/Å 3 ). 2Fo-Fc maps calculated using phenix.refine with a low-resolution cut-off at 25 Å are shown in the middle three panels including bulk solvent correction (BSC) and for the lower three panels without BSC.

Figure S7
Resolution-dependent effect of bulk solvent correction on refinement parameters. Rfactors (R free in red and R work in blue) are presented as function of resolution for the indicated data sets after 30 cycles of restrained refinement with refmac in space group P6 1 22. The refinement was done including either the entire resolution range (upper row; data sets reported in SI, Tables S7-9) or including only reflections between 2.8 -8 Å (lower row). Bulk solvent correction (BSC) was applied as indicated by the legend in the upper left panel. The overall R work /R free values are given in each panel.

Figure S8
Guest electron density in solvent channels for 4 Å truncated data. Crystals like those in Figure 6d after 15 days of soaking with Alex-CaM or CytC were diffracted at BioMAX, Lund. The data sets were processed in P6122 (a) or P1 (b). Left, protein assembly cartoons (gray ribbons) were generated from the molecule(s) (blue ribbon) in the asymmetric unit surrounding each channel area marked with a red square. Right, electron density within the channel areas corresponding to each red square on the left for ds-TrpR crystals alone or soaked with Alexa-CaM or CytC as indicated. Each structure was solved at 4 Å resolution. Electron density is represented as 2Fo-Fc maps at 0.6 rmsd contour level (upper panels), or at 0.5 rmsd contour level with blurring to 150 A 2 (lower panels). The images were generated using Coot.