1. Three-Part Small-scale Screening Platform for the Masses Ronnie Frederick Center for Eukaryotic Structural Genomics Department of Biochemistry University of WisconsinMadison NIGMS U54 GM074901, JL Markley, PI, GN Phillips, & Brian G. Fox, Co-Directors Promega/UW 133-GT34, BG Fox, PI http://www.uwstructuralgenomics.org

2. CESG Protein Production Unified cell-based and cell-free expression pipelines Expression screening in E. coli and wheat germ extract Data captured into Sesame at each step

3. Modular Vector Design Essential modules bounded by unique restriction sites –Promoter, affinity tags, solubility tags, linker combinations –In vitro TEV proteolysis of target from other tags –Over 70 vector variants

4. Small-Scale Expression Testing Small-scale expression evaluations (Studier media) –SDS-PAGE in slab gels (~2 days processing and analysis) –Tested for total expression, fraction soluble and fraction cleaved –~1.5 weeks to evaluate, labor intensive, expensive ($450/ 96 targets) –No reliable assessment of target status after TEV proteolysis –Predicting large-scale results ~83% accurate (10% false negative and 6% false positive) 96-well plasmid workgroup96-plate transformation96-well growth

5. Expression and Solubility of MBP-Fusion Proteins Expression, Solubility, and TEV cleavage Solubility and TEV Cleavage of MBP-Fusion Proteins MBP 46 kDa Tev protease Target ENLYFQ ^ S Target ENLYFQS MBP 46 kDa Solubility unknown Expression & Solubility His n(6-8)

6. Results from 66 Targets Passed On to Large-Scale and Grown in 5SM SS suitable in 5SM & TB+g LSPP suitable 22 4 2 IPTG+ 5 SS unsuitable LSPP suitable SS suitable in TB+g LSPP suitable 25 SS suitable in 5SM LSPP suitable 2 Positive in Small Scale but Failed in Large Scale SS unsuitable in both 5SM & TB+g LSPP suitable SS unsuitable in both 5SM & TB+g (IPTG+) LSPP suitable 6 SS unsuitable in 5SM LSPP unsuitable Prediction of Large-Scale Results (Analysis of Medical Relevance Target Workgroup, WG1323) Total LSPP growths = 66 Total LSPP failures = 10 Unsuitable in LSPP = 6 LSPP Suitable (small-scale error) = 4 Small-scale correctly predicted 49 (suitable) + 6 (LSPP unsuitable) (True positive + true negative = success rate) 49+6/66 = 55/66 = 83.33% False negative rate = 7/66 = 10.6% False Positive rate = 4/66 = 6.1% Source of Attrition Within Small-Scale For a given vector (VP56K), attrition was due to: poor expression, low solubility and failure in TEV protease cleavage Solubility of liberated target is not assessed 1803 Galdieria 1888 Cyanidioscyzon merolae (C. merolae) 2015 Medical Relevance

7. Part 1: A New Modular Vector Design Essential elements bounded by unique restriction sites –Promoter, solubility tags, affinity tags, linker combinations –Over 70 expression vector variants Linker module altered to contain a TVMV-His8-TEV context Constitutive expression of TVMV protease from vector Altered expression of the lac repressor (LacI) In vivo proteolysis of MBP by TVMV protease (self-cleaving VP62K). Auto-induction medium

8. First pass Se-Met medium in the production pipeline Factorial design identified favored compositions that were different from original PASM (Studiers media) Expression performance is strongly linked to the level of lac repressor (LacI) expressed from the backbone Part 2: Medium Evolution Reference: Enhanced Bacterial Protein Expression During Auto-induction Obtained by Alteration of Lac Repressor Dosage and Medium Composition Paul G. Blommel, Katie J. Becker, Petar Duvnjak, and Brian G. Fox (in press).

9. Part 3: Simple robotics for the masses Automated Purification System Can the simple Maxwell robotics contribute to our pipeline? Couple the designed vector and the improved expression medium with the new robotics E. coli pipeline: Auto-induction in vivo cleavage automated Maxwell purification Use it to obtain predictive information of protein behavior in purification, NMR and X-ray structure determination www.promega.com/paguide/images

10. Automated Analysis System Caliper Lab Chip-90, 1 chip and ~4 hours of screening ($225/96 targets) Accurate determination of the amounts of purified protein targets –Preliminary applications –Maxwell 16 purified targets –MBP-fusions (Se-labeled targets) –TEV-treated 15 N-labeled targets http://www.caliperls.com/downloads/LC90-BR-01Single.pdf

11. LC90 Analysis of Maxwell Screen Various structural genomics targets or controls –A mix of the good, the bad and the ugly Target protein detected for all 24 targets In vivo cleavage with GFP as a control worked

12. Extending Small-scale screening: Human embryonic stem cell proteins 24152.2214NPM2C1 13459.6114tcl1A8 31035.3258C10 Orf 6A5 87639.8786CCNFA4 Mol. wt.ResiduesProteinCoord. Human embryonic stem cell protein A8 was expressed in 15 N auto-induction medium for an overnight culture from a colony pick Automated purification with the Maxwell 16 system gave 3 mg from 12 lanes Multiple-well purification of 15N labeled Human Embryonic Stem protein (A8) New Path to Structure Screening: Micro and Meso-scale protein purification

13. 15 N HSQC NMR of His 8 -GFP expressed in self- cleaving vector pVP62K at 35 ˚C. The NMR time required was 1 h Cost is $50 (labeled medium and purification cartridges). 40 minute run (1.5 mg from 8 mL). NMR measurements by Dr. Jikui Song 15 N HSQC NMR of human embryonic stem cell protein A8 expressed in self-cleaving vector pVP62K. The NMR time required was 8 h Crystals of SeMet-labeled GFP Meso-Scale Protein Production and Purification for HSQC NMR Analysis

14. Acknowledgements Professor Brian G. Fox (section supervisor) Vector Design and FlexiVector Cloning Paul Blommel, Michael Goren, Peter Martin, Kory Seder, Eric Steffan and Russell Wrobel Factorial Evolution of Auto-Induction Media Paul Blommel, Katy Becker and Petar Duvnak CESG Small-scale screening team Lai Bergeman, Mike Cassidy, Ah-Young Lim and Jung (James) Whan Yoon, Jason Bunge (High school student) & Dr Jikui Song (NMR analysis) Past members John Kunert, Megan Riters, Nick Dillion, Rachel Schiesher & Jay Juhjung Chin