1. Expression and purification of bacteria membrane proteins for structural studies

2. To solve a membrane protein structure is very hard for the following reasons: Hard to over-express large quantity, functional protein; To get the protein into solution for purification it needs to be solublized with detergent, which might destroy the protein, we need protein that is functional, folded and monodisperse, many detergents need to be screened; Hard to crystallize: with detergent around the molecule it’s much harder to make contacts for crystal packing; Crystals often don’t diffract well because of high solvent contents.

3. Start with multiple homologues Multiple Homologues Plasmid Construction Expression Screening Detergent Screening Crystallization It is considered a good starting point if we could get 1mg of protein after passing through one column from one liter cell culture. To find the right detergent/detergent mix to keep the protein stable and homogenous.

4. Expression Assay 21.5 14.4 6.0 kD Eleven out of fifty targets overexpressed using arabinose induction. P a ra 6-aa Pml I 6-his CAC GTG TM gene TTA ATT AA Pac I Pme I Nco I RBS Sma I PT7PT7 Expression Vector 96 x 65 mL Fermentor From Linda Columbus

5. Special considerations when over- expressing membrane proteins for crystallographic studies DifficultiesSolutions Over-expressed protein often toxic to host cells Choose a tight promoter to prevent leaky expression. Induce at mid-log phase so the cell will keep growing after induction, monitor the cell growth after induction. Problems with misfolded protein in inclusion bodies Lower growth temperature, avoid over-loading translocon and chaparones.

6. First purification step: isolate the membrane fraction Disrupt the harvested cells by mechanical force (high pressure) using equipment such as French Press or Xpress; Remove unbroken cells with a “low speed” centrifugation (10k xg). This step removes cell debris that are not membrane, as well as some cell organelles such as inclusion bodies; Collect the membrane fraction by centrifuge the supernatant from the last step with a “high speed” ultra-centrifugation (35-40k xg). This step removes all the soluble proteins.

7. Properties of detergent molecules From Anatrace.com detergent = very expensive soap

8. Critical micelle concentration (CMC) From Anatrace.com

9. Solublize membrane protein with detergent From Anatrace.com

10. Lipopeptide detergents McGregor CL et. al. Nature Biotechnology, 2003 detergent LPD phospholipid

11. Special considerations when purifying membrane proteins DifficultiesSolutions Extracting protein from its lipid bilayer often result in unstable/inactive proteins Test a few detergent chosen by their reported success and price. Consider detergent exchange. The amount of de-lipidation is often important for the protein’s stability and homogeneity Keep in mind every column the protein passes is a de-lipidation process as well. Sometimes it’s important to add back lipid to keep the protein happy. Because detergent forms micelles, it often concentrates with the protein sample, when the concentration is too high, the empty detergent micelles pack into crystals and making it impossible to get well-diffracting protein crystals. Use 2-3 times CMC in all the buffers used for purification; control the protein : detergent ratio in the last purification step; Dialyze the sample to lower detergent concentration if necessary.

12. An example: Major facilitator superfamily (MFS) Largest family of secondary active transporters, > 5,000 members  Including: - Glucose-6-phosphate transporter from E. coli - UhpT - Glucose-6-phosphate transporter from human - G6PT - Glucose transporters in erythrocyte and muscle - Glut1, Glut4 - Vesicular glutamate transporters in presynapse - VGlut1, VGlut2 TransportersFamiliesMFSABC E. coli297596667 S. cerevisiae258427822 Human1,2476910494

13. Fifteen MSF proteins are expressed in E. coli Searching crystallization space by cloning homologues pBAD-MycHis MCS-Myc-His pBAD term AMPr AraC pUC ori SDS-PAGE Western blot (INDIA anti-His)

14. Flexible tail is identified and removed by proteolysis and mass spectrometry C N N C Wile-type: 1-452 Final construct: 2-452 G2->L2, C-terminal RNGG->LVPR (thrombin site) R449

15. GlpT is a monomer in detergent solution as measured by size-exclusion column on HPLC In DDM Void

16. Commercially-available MP screens MembFac (Hampton) MemStart & MembSys, MemGold (Molecular Dimensions) JBScreen Membrane 1-3 (Jena Biosciences) The MbClass (Nextal) Statistics show most membrane protein crystalizes in PEG based Conditions, a through PEG/pH screen should be performed as well. Crystallization

17. Web resources Membrane protein structure databases: Topology database http://blanco.biomol.uci.edu/mptopo/ Known structures http://blanco.biomol.uci.edu/Membrane_Proteins_xtal.html Known structures with statistics of crystallization conditions http://www.mpibp- frankfurt.mpg.de/michel/public/memprotstruct.html PDB database for membrane proteins http://pdbtm.enzim.hu/

18. Lemieux MJ et. al. Protein Science, 2003

21. Structure determination to 3.3 Å P3 2 21, a = b = 97.6 Å, c = 175.2 Å, 72 % sol W 15 SIRAS phasing R = 29.65 %, R free = 32.53 %

22. Another example: bacteria amino acid transporter It took 3 years before a well-behaving target was identified in the pipeline: Leucine transporter from the thermophilic bacterium Aquifex aeolicus

23. By varying the cell culture temperature, the protein became very homogeneous shown by size exclusion chromatography

24. Yamashita A et. al. Nature, 2005

25. It often takes a long time before a suitable target is identified, Can we speed up this process?

26. Green Fluorescent Protein (GFP) has existed for more than one hundred and sixty million years in one species of jellyfish, Aequorea victoria. The protein is found in the photoorgans of Aequorea. Green Fluorescent Protein

27. Kawate & Gouaux Structure. 2006 Apr;14(4):673-81