The objective of a single fragment screening (FS) experiment is to individually soak 96 different chemical fragments into crystals of the desired protein target and to analyze the respective complex structures by X-ray crystallography. In order to maximize the chances for a successful FS experiment, the following requirements should be met at the outset.
- Reliable and reproducible crystallization conditions should already be established for the desired target. At least 200 crystals are needed for a single FS campaign. Ideally, the crystals should be at least 50 μm in size in each direction.
- Soaking and cryo-conditions that yield diffracting crystals should be established. Ideally, the crystals should tolerate DMSO to a concentration of 5 or 10% (v/v). The diffraction properties after the desired soaking time (ideally o/n, but 1h - 48h possible) should be better than 2.0 Å, but certainly not worse than 2.5 Å. If a known substrate or inhibitor is availabe, a positive control binding experiment should be performed prior to attempting a FS experiment. The experiment becomes less cumbersome, if soaking and cryo-protection can be combined into one step.
- A proposal for beam time for the FS experiment should be submitted in advance through HZB's electronic user office GATE in order to be able to book beam time for the experiment. Typically, six shifts should suffice. GATE accounts and valid online-safety training is also necessary to access the ring and thus the prep lab, in case soaking experiments are carried out on site. If you need access to our BioLab as well, registration is necessary at least one week prior to your visit.
Please go through the following list of questions carefully.
- Can >200 protein crystals be produced in one crystallization condition?
- Are your crystals larger than 50 μm in each direction?
- Do the crystals tolerate 5-10% (v/v) DMSO in the soaking/cryo-condition?
- Is the resolution of the crystals after soaking and cryo-cooling typically better than 2.0 Å?
- Is your cryo-protectant included in the soaking condition?
- Was the positive control bound to the protein after soaking and cryo-cooling?
- In case you want to bring crystals, do they still diffract after e.g. several weeks?
If you can answer all questions with YES you are perfectly prepared and ready to start your FS campaign. If you can answer most questions with YES you are still well prepared and almost ready. Also, not all campaigns necessarily need to fulfill all criteria. Please do not hesitate to contact us if you have any questions.
Preparation for the visit:
- Loops, vials, bases and pucks for sample mounting and storage need to be provided by you. Please bring a sufficient number of vials and bases as well as sufficient number of loops of appropriate size for your crystals (we recommend the MiTeGen MicroLoops). If you would like to store your samples for data collection after your initial visit, please bring an appropriate number of SPINE pucks. Alternatively, we have aluminum canes available for storing samples.
- Minimize storage time of the frozen crystals. This is recommended in general but especially if you perform the crystal treatment on site, let the beam time then follow as soon as possible.
During the visit:
- F2X-Entry screen or HZB library will be provided by us. However, make sure to use of it in the most economical way. Ideally, plates should be unsealed and used for soaking experiments in one day.
- Accommodation is available for BESSY II users at the guesthouse.
- Naming of Datasets. To facilitate data processing, refinement and analysis via XDSAPP, fspipeline and PanDDA, please stick to the following naming conventions:
F2X-Entry Screen and FragXtal screen according to the well identifyers:
HZB-library and other libraries according to 3-digit numbering:
(the letter in the end of the name identifies measurements of the same Fragment, e.g. duplicates)
After the visit:
Data treatment and hit finding
To facilitate the analysis of the hundreds of data sets measured in a FS campaign, we provide a completely automated solutions for data processing, refinement and hit identification.
Following data processing with XDSAPP (Sparta et al., 2016), all created MTZ files are loaded into a refinement and hit identification pipeline based on PHENIX (Adams et al., 2010) and COOT (Emsley et al., 2010) that was developed together with the Drug Design Group in Marburg (Schiebel et al., 2016). Given an input PDB model, all data sets are refined and their final electron density maps are screened for unexplained blobs. The use of the pipeline is available to users at our beamlines who have signed a collaboration contract for a FS experiment.
Additionally, the results of the refinement pipeline are automatically arranged and submitted for PanDDA (Pan-Dataset-Density-Analysis), in order to efficiently visualize low-occupancy binders (Pearce et al., 2017, PanDDA-website).