increased diels-alderase activity through backbone remodeling guided … · increased...

Supporting Material for:

Increased Diels-Alderase activity

through backbone remodeling guided by Foldit players

AUTHORS

Christopher B. Eiben*1, Justin B. Siegel*1, Jacob B. Bale1,2, Seth Cooper3, Firas Khatib1, Betty W. Shen4, Foldit Players, Barry L. Stoddard4, Zoran Popovic3, David Baker1,5,6

AFFILIATIONS *These Authors Contributed Equally 1Department of Biochemistry, University of Washington, Seattle, Washington, USA. 2Graduate Program in Molecular and Cellular Biology, University of Washington, Seattle, Washington, USA. 3Department of Computer Science and Engineering, University of Washington, Seattle, Washington, USA. 4Division of Basic Sciences, Fred Hutchinson Cancer Research Center, Seattle, WA 98109, USA. 5Howard Hughes Medical Institute, University of Washington, Seattle, Washington, USA. 6Correspondence should be addressed to D.B. ([email protected]).

Nature Biotechnology: doi:10.1038/nbt.2109

Table of Contents

I. SUPPLEMENTARY FIGURES AND TABLES

II. METHODS

III. SUPPLEMENTAL DATA

IV. REFERENCES


I. SUPPLEMENTARY FIGURES AND TABLES

Supplementary Figure 1. Workflow of design process for the construction of CE6.

Steps that represent contributions from the online community of Foldit players are

highlighted in green. Fixed portions of DA_20_10 precursor 3I1C are shown in yellow

with modeled transition state as spheres in cyan. (A) DA_20_10 precursor 3I1C with the

four loops players were allowed to design highlighted in red. Here online players were

asked to redesign the loops in order to make additional contacts to the modeled transition

state. (B) DA_20_10 precursor 3I1C versus an example online player design from

“Cover the Ligand” in marine, additional online player designs are shown in

Supplemental Figure 3. (C) An example online player design in marine versus the

advanced Foldit player optimized design in light grey. (D) The optimized design from

the advanced player in light grey versus the online player design from “Back Me Up” in

teal. (E) The online player design in teal versus the posted structure for the “Quick Loop”

puzzle in dark grey. Since the structure for experimentally determined CE6 was

unknown the structure was posted as an unstructured loop and online players were asked

to solve the structure for the given sequence. (F) The posted CE6 quick loop structure in


dark grey versus the online player predicted structure in purple. (G) CE6 online player

predicted in purple versus the experimentally determined crystal structure in blue.


Supplementary Figure 2 The Diels-Alder reaction. Diene (1) and dienophile (2) undergo a pericyclic [4+2] cycloaddition (3) to form a chiral cyclohexene ring (4). Also shown in (3) is a schematic of the design target active site, with hydrogen bond acceptor and donor groups activating the diene and dienophile and a complementary binding pocket holding the two substrates in an orientation optimal for catalysis.


Supplementary Figure 3. The four top ranking online player designs before stability adjustments. Ligand transition state in cyan, DA_20_10 precursor 3I1C in yellow, and player designs in blue, white, black and red. The library which found CE0 was based on design in panel A, but was manipulated to join the two player helices into a longer single helix with a longer linker region and different helix placement (shown in Supplementary Fig. 1C, Library 1).


Supplementary Figure 4. Crystal structure of CE6, with the asymmetric unit comprised of 2 molecules of CE6. In purple, chain A and in blue chain B. Side chains in the hydrophobic groove shown as sticks.


Supplementary Figure 5. CE6 crystal contact knockout activity assessment. Whole cell lysate assay in PBS with 4% DMSO, .1mM diene, 5mM dienophile. Samples were analyzed in quadruplicate after one hour. Error bars represent one standard deviation.

0

1

2

3

4

5

6

CE6 crystal contacts knock out CE6

Rate of P

rodu

ct Forma=

on (a

u/hr)


Supplementary Figure 6. Disembodied view of conserved active site residues between DA_20_10 (gold, from crystal structure of precursor 311C) and CE6 (grey, 3U0S), with the modeled substrates in cyan. No large backbone or side chain changes are observed.


Supplementary Figure 7. Substrate specificity. Reactions were carried out with 0.1 mM diene and 5mM of appropriate dieneophile depicted above in PBS with 4% DMSO at 298K with or without 5uM of CE6 or DA_20_10. Samples were taken in triplicate. Colored bars represent the mean of product formation per hour (arbitrary units), and capped bars represent standard deviation from three independent measurements.


Supplementary figure 8. Gel of purified protein used for kinetic analysis. First lane is kaleidoscope gel ladder (Bio-Rad, Catalogue #161-0375). Lane 2-8 are CE0, CE1, CE5, CE6, DA_20_10, DA_20_10H287N respectively.


Supplementary Table 1. Data collection and refinement statistics for 3U0S. ________________________________________________________________________ Data set ID CE6-1 Wavelength (Å) 0.997 ________________________________________________________________________ Data collection Space group p41 Cell dimensions: a (Å) 87.002 b (Å) 87.002 c (Å) 163.330 a, b, g (°) 90 Resolution (Å) 46.15 – 2.60 Rmergea 0.070(0.416)* I/sigI 20.5(6.01)* Completeness (%) 100(100)* Redundancy 12.8(9.9)* Refinement: Resolution (Å) 46.15 – 2.6 (2.96-2.6)* R-factor/R-freeb 0.171/0.198 No. atoms Protein 5090 Ligands/Ions 98 Water 240 B factors (Å2) Protein 44.94 Ligands/Ions 106.5 Water 46.29 R.m.s deviations: Bond length (Å) 0.012 Bond angles (o) 1.348 Ramachandran statistics: Core (%) 83.3 Allowed (%) 15.9 Generously allowed (%) 0.70 Disallowed (%) 0.00 * Highest resolution shell values in parenthesis. a Rmerge = S|Ihi - <Ih>|/SIh, where Ihi is the ith measurement of reflection h, and <Ih> is the average measured intensity of reflection h. b R-factor/R-free = Sh|Fh(o) - Fh(c)|/Sh|Fh(o)|. Where R-free was calculated with 5% of the data excluded from refinement.


II. EXPERIMENTAL CHARACTERIZATION

LIBRARY WHOLE CELL LYSATE ASSAY Libraries were screened using a whole cell lysate assay. Single colonies were

isolated from petri dishes and used to start cultures in 0.5 mL Lysogeny Broth (LB) in 2 mL 96 deep well plates with 50 µg/mL kanamycin. Plates were grown with shaking at 1250 rmp in a plate shaker at 37oC for 16-24 hours. 20 µL of overnight was then used to inoculate 1 mL of Terrific Broth (TB) with 50 µg/mL Kanamycin and allowed to grow for 2.5 hours at 37oC shaking at 1250 rmp. After 2.5 hrs, cells were induced with 50 µL of 10 mM Isopropyl β-D-1-thiogalactopyranoside (IPTG) and transferred to an 18 oC shaker for 24 hours. Cells were harvested by centrifugation at 2900 g for 20 min and the supernatant decanted. Cells were resuspended in 250 µl of lysis buffer consisting of 2 mg/ml lysozyme isolated from egg white, 0.2 mg/ml DNAse, 1 mM phenylmethanesulfonylfluoride (PMSF) and 1x PBS. Lysis was performed either by three cycles of freeze thaw (from -80 oC to room temp) or via bath sonication (Ultrasonic Processor Gex750) at 70% power for 1 minute, pulsing 20 s on, 40 s off. Whole cell lysate was then spun down at 2900 g for 20 min, and supernatant was used for product formation assay. DIELS-ALDER PRODUCT FORMATION ASSAY The Diels-Alder reactions were performed in PBS, at 25˚C, with 4% DMSO. The DMSO came from the diene stock, for which the stock solution (100 mM) was made in DMSO and always diluted such that the final reaction would contain 4% DMSO. After incubation of the desired amount of diene and dienophile in the presence or absence of protein, 5 µL aliquots were taken at various time points and quenched in a solution of 80:20 acetonitrile:water with 0.1% formic acid and 1 mM Benzoic Acid (95 µL). This solution was incubated for 5 minutes and then the precipitated protein was filtered using a Millipore multiscreen solvinert filter plate (Product Number: MSRLN0450). 20 µL of the filtered quench solution was subsequently injected and analyzed using a liquid chromatography-tandem mass spectrometry (LC-MSMS) assay. The column used for the chromatography run was a Hypersil Gold C18 (Dimensions 100 x 2.1, 1.9 micron particle size, Thermo). The following gradient was preformed for the chromatography run at a flow of 500 µL/min: 95:5 Water:ACN (0.1% formic acid) for 30 seconds, followed by a gradient over 4.5 minutes ending at 5:95 Water:ACN (0.1% formic acid), ending by switching back to 95:5 Water:ACN (0.1% formic acid) and letting the column re- equilibrate for 1 min before the next injection. The Diels-Alder reaction product and benzoic acid (used as an internal standard for injection differences and sample evaporation) was detected using a TSQ Quantum Access (Thermo Fisher Scientific) with an ESI probe to ionize the parent ion followed by fragmentation using Argon gas with the following parameters (tuned on the diene product standard, synthesized a previously described (2): Spray Voltage: -4000 Sheath Gas Pressure = 20


Aux Gas Pressure = 5 Capillary Temperature = 300 Chrom Filter = 10 Collision Pressure = 1.5 Skimmer Offset = 10 Scan Width = 0.01 Scanned in Negative Centroid Mode Benzoic Acid Specific Parameters Parent Mass = 120.994 Product Mass = 120.994 Scan Time = 0.08 Collision Energy = 5 Tube Lens = 112 Diels-Alder Product Specific Parameters Parent Mass = 345.150 Product Mass = 211.033 Scan Time = 0.2 Collision Energy = 21 Tube Lens = 140 MUTAGENSIS All mutagenesis was performed following the Kunkel protocol (1), using oligonucleotides designed with Stratagene’s primer design online tool found here: https://www.genomics.agilent.com/CollectionSubpage.aspx?PageType=Tool&SubPageType=ToolQCPD&PageID=15. To design degenerative primers, Andrew E. Firth and Wayne M. Patrick’s AA-Calculator (2005, Statistics of protein library construction, Bioinformatics, 21, 3314-3315) was used, which can be found here: http://guinevere.otago.ac.nz/cgi-bin/aef/AA-Calculator.pl Kunkel mutagenesis reactions were transformed directly into Stratagene BL21DE3*, and colonies were picked and screened for Diels-Alderase activity using the whole cell lysate assay described above. For each library we screened enough colonies to cover the theoretical library by four-fold. KINETIC CHARACTERIZATION The most promising designs (CE0, CE4, CE6 and DA_20_10) were grown up at 0.5 L scale and purified as previously described (2). Protein concentrations were determined by measuring their absorbance at 280 nm, and using their calculated extinction coefficients to convert the absorbance into molarity. After normalization the proteins concentrations were verified to be equivalent by running them on an SDS-PAGE gel (Supplemental Figure 7). Kinetics were measured by holding either the diene or dienophile constant and varying the other in the presence or absence of enzyme. Reactions were incubated at 298K in PBS, 4% DMSO, pH 7.4, 298K, 10 µM protein. Triplicate samples were taken after three hours points from which the amount of product


formation was determined using the LC-MSMS assay described earlier. For obtaining kinetic constants we used the standard Michaelis-Menten model to fit each curve, with the exception of CE0, which was fit to a linear slope as no saturation was observed. (Figure 2). The kcat was determined from the calculated maximum product production from the Dienophile curve, and dividing by the amount of enzyme and hours the sample was incubated before quenching. A linear product production rate was assumed since the substrate depletion was less than 5% total substrate. Errors for kcat/(KM-diene * KM-dienophile) were calculated using the standard error propagation formula: δkcat/(KM-diene * KM-dienophile) = |kcat/(KM-diene*KM-dienophile)|*√((δkcat/kcat)2 + (δKM-diene/KM-diene)2 + (δKM-dienophile/KM-

dienophile)2). SUBSTRATE SPECIFICITY To experimentally determine substrate specificities each dienophile from Supplemental Figure 6 was either purchased or synthesized, as previously described (2). The dienophiles were assayed in the presence or absence of 5 µM CE6 or DA_20_10 with 0.1 mM diene and 5 mM dienophile at 298K in a PBS solution with 4% DMSO. Samples were taken in triplicate at 3 hours, quenched, and filtered as previously described. To detect product, a liquid chromatography-mass spectrometry assay was used. The mass spectrometer was used in a selective-ion-monitoring mode and set to specifically detect the expected product from each reaction. The same tuning parameters were used as previously described. The average area of each detected peak formed per hour was used to determine the relative amount of product formed with and without enzyme. The relative peak areas formed per hour are reported as Supplemental Figure 6. CRYSTALLOGRAPHIC DATA COLLECTION AND REFINEMENT

CE6 was expressed in E. coli BL21DE3* cells (Stratagene) and purified as previously described (2). Purified protein was dialyzed into crystallization buffer containing 25 mM HEPES pH 7.25, 100 mM NaCl, and 5% glycerol and concentrated to approximately 30 mg/mL.

CE6 crystals were grown at 18oC by the hanging drop vapor diffusion method. A 15 mg/mL solution was prepared of CE6 in crystallization buffer with Trypsin (Sigma T1426) added in a 1 to 10,000 molar ratio as compared to CE6. Approximately 20 minutes later, drops were set containing 1 μl of this protein solution and 1 μL of mother liquor (100 mM HEPES pH 7.5, 2% v/v PEG-400, and 2.0 M (NH4)2SO4) over a reservoir volume of 0.3 mL. Crystal growth was observed within one week. Crystals were cyroprotected in a mixture of 75% mother liquor with 25% ethylene glycol and flash frozen in liquid nitrogen.

A dataset was collected at the Advanced Light Source on BL5.0.1 (at a temperature of 100 K and a wavelength of 0.97740 Å) and processed using HKL2000 (3) (See Supplemental Table 1 for statistics). Phases were determined by molecular replacement (MR) using the program PHASER (4) provided in the Collaborative Computational Project No. 4 software suite (CCP4) (5). The MR search model was derived from the wild-type 1E1A structure (6) by removing all water molecules and ligands, 6 residues of the unstructured C-terminus, and residues 36-44 of the redesigned active site loop. Following MR, the missing residues and ligands of CE6 were built into the model using the molecular modeling software Coot (7,8) and refinement was


performed using the CCP4 provided version of REFMAC5 (9,10,11). The C-terminal HIS-tag and linker region were not visible in the electron density and thus were not modeled. Alternate conformations were modeled for some surface exposed residues, including LYS44 and ARG56 of the redesigned loop. Also of note, a HEPES molecule was modeled into the active site in both chains of the asymmetric unit. Ramachandran statistics (Supplemental Table 1) were generated using the online version of PROCHECK (12) provided by the JCSG server found here: http://www.jcsg.org/scripts/prod/validation/sv_final.cgi. COORDINATE AND DATA DEPOSITION Structure factors and coordinates have been deposited in the RCSB database under PDB ID code 3U0S.


LIBRARY DESIGN

LIBRARY 1. The below sequences depict the protein loops tested starting at position 36. The most active variant is denoted with a bullet point (�) and the corresponding sequence is highlighted in red.


LIBRARY 2. The two libraries below (2A and 2B), starting at position 36, were tested sequentially. The best sequence of the 2A was used as the input for 2B. All combinatorial sequences indicated were tested most active variant is denoted with a bullet point (�) and the corresponding sequence is highlighted in red. * Denotes all 20 amino acids tried at the position.


LIBRARY 3. The two libraries below (3A and 3B), starting at position 36, were tested sequentially. The best sequence of the 3A was used as the input for 3B. All combinatorial sequences indicated were tested. The fastest variant from library 3A and 3B is denoted with a bullet point (�) and corresponding sequence is colored in red.


III. SUPPLEMENTAL DATA

DESCRIPTIONS OF PUZZLES POSTED

Name: 192: Cover the Ligand Status: Closed

Created: 08/21/2009 Points: 100

Expired: 08/28/2009 – 23:00 Difficulty: Advanced

Description:

Our first ligand competition puzzle! The problem: With the current backbone the protein can only touch half of the ligand. The goal: Rebuild and insert and delete backbone segments in combination with mutating amino acids to shelter the ligand. There are four editable regions so focus on those!

Significant contributions

made by: g_s

Name: 195: Cover the Ligand 2 Status: Closed


Difficulty: Advanced

Description:

We've got a slighly different setup this time, now the ligand isn't movable! The problem: With the current backbone the protein can only touch half of the ligand. The goal: Rebuild and insert and delete backbone segments in combination with mutating amino acids to shelter the ligand. There are four editable regions so focus on those!


made by: aap, gdnskye, and anne romaine

Name: 366: Back Me Up 1 Status: Closed


Expired: 10/03/2010 - 23:00 Difficulty: Intermediate

Description:

We designed a helix to contact our substrate, but the helix needs back up to stay in place. The goal of this puzzle is to design a loop and mutate residues to create a more stable configuration, making a more active catalyst (ie create a Helix-Loop-Helix). See the blogpost for more details: http://fold.it/portal/node/988587


made by: CharlieFortsConscience, infjamc, Mark-


Name: 401: Quick Loop Puzzle Status: Closed


Expired: 02/27/2011 - 23:00 Difficulty: Intermediate

Description: This puzzle will only be up until this Sunday. Can you make the loop better before then? Significant

contributions made by:

CharlieFortsConscience

Name: 494: Quick Fix the Loop Puzzle Status: Closed


Expired: 12/16/2011 - 17:00 Difficulty: Beginner

Description: Can you make the loop better before this puzzle closes on Friday? Significant

contributions made by:

BootsMcGraw, Bletchley Park, and Mark-


FOLDIT PLAYER PREDICTION OF APO CE6.

The “unfolded” apo version of CE6 (the apo form equivalent of Supplementary Figure 1e, grey) was given to the Foldit players to predict the structure in the absence of the modeled transition state ligand (Puzzle #494, “Quick Fix the Loop Puzzle”). The top Foldit model (orange) forms a helix, but is shifted towards the active site relative to the crystal structure (purple), which would sterically prevent the transition state (cyan) from binding as modeled. This suggests that the helix may shift into a functional form upon ligand binding.


PROTEIN SEQUENCES

>CE6 Amino Acid Sequence MEIPVIEPLFTKVTEDIPGATGPVFDKNGDFYIVASPLSEALTKANSPAEAYKASRGAGEILRIDLKTGKKTVICKPEVNGYGGSPIGCQCDRDANQLFVADMRLGLLVVQTDGTFEEIAKKDSEGRRMQGCAYCAFDYEGNLWITAPAGEVAPADFTISLREKFGSIYCFTTDGQMIQVDTAFQCPAGIAVRHMNDGRPYQLIVAEQPTKKLWSYDIKGPAKIENKKVWGHIPGTHKGGAAGMDFDEDNNLLVANWGSSHIEVFGPDGGQPKMRIRCPFEKPANLHFKPQTKTIFVTEHENNAVWKFEWQRNGKKQYCETLKFGIFGSLEHHHHHH >CE5 Amino Acid Sequence MEIPVIEPLFTKVTEDIPGATGPVFDKNGDFYIVASPLSEALTKANSPAEAYKASRGGSGGEILRIDLKTGKKTVICKPEVNGYGGSPIGCQCDRDANQLFVADMRLGLLVVQTDGTFEEIAKKDSEGRRMQGCAYCAFDYEGNLWITAPAGEVAPADFTISLREKFGSIYCFTTDGQMIQVDTAFQCPAGIAVRHMNDGRPYQLIVAEQPTKKLWSYDIKGPAKIENKKVWGHIPGTHKGGAAGMDFDEDNNLLVANWGSSHIEVFGPDGGQPKMRIRCPFEKPANLHFKPQTKTIFVTEHENNAVWKFEWQRNGKKQYCETLKFGIFGSLEHHHHHH >CE4 Amino Acid Sequence MEIPVIEPLFTKVTEDIPGATGPVFDKNGDFYIVASPLSEALTKGSGANNTNGAGEILRIDLKTGKKTVICKPEVNGYGGSPIGCQCDRDANQLFVADMRLGLLVVQTDGTFEEIAKKDSEGRRMQGCAYCAFDYEGNLWITAPAGEVAPADFTISLREKFGSIYCFTTDGQMIQVDTAFQCPAGIAVRHMNDGRPYQLIVAEQPTKKLWSYDIKGPAKIENKKVWGHIPGTHKGGAAGMDFDEDNNLLVANWGSSHIEVFGPDGGQPKMRIRCPFEKPANLHFKPQTKTIFVTEHENNAVWKFEWQRNGKKQYCETLKFGIFGSLEHHHHHH >CE1 Amino Acid Sequence MEIPVIEPLFTKVTEDIPGATGPVFDKNGDFYIVASPLSEALRKGSGANNTNGAGEILRIDLKTGKKTVICKPEVNGYGGSPIGCQCDRDANQLFVADMRLGLLVVQTDGTFEEIAKKDSEGRRMQGCAYCAFDYEGNLWITAPAGEVAPADFTISLREKFGSIYCFTTDGQMIQVDTAFQCPAGIAVRHMNDGRPYQLIVAEQPTKKLWSYDIKGPAKIENKKVWGHIPGTHKGGAAGMDFDEDNNLLVANWGSSHIEVFGPDGGQPKMRIRCPFEKPANLHFKPQTKTIFVTEHENNAVWKFEWQRNGKKQYCETLKFGIFGSLEHHHHHH >CE0 Amino Acid Sequence MEIPVIEPLFTKVTEDIPGATGPVFDKNGDFYIVASPQLELARGGARTDSNNGAGEILRIDLKTGKKTVICKPEVNGYGGSPIGCQCDRDANQLFVADMRLGLLVVQTDGTFEEIAKKDSEGRRMQGCAYCAFDYEGNLWITAPAGEVAPADFTISLREKFGSIYCFTTDGQMIQVDTAFQCPAGIAVRHMNDGRPYQLIVAEQPTKKLWSYDIKGPAKIENKKVWGHIPGTHKGGAAGMDFDEDNNLLVANWGSSHIEVFGPDGGQPKMRIRCPFEKPANLHFKPQTKTIFVTEHENNAVWKFEWQRNGKKQYCETLKFGIFGSLEHHHHHH


IV. REFERENCES

1. Kunkel TA Rapid and efficient site-‐specific mutagenesis without phenotypic selection. P Natl Acad Sci USA 82(2):488-‐492 (1985).

2. Siegel JB, et al. Computational design of an enzyme catalyst for a stereoselective bimolecular Diels-‐Alder reaction. Science 329(5989):309-‐313 (2010).

3. Otwinowski, Z., Minor W. Processing of X-‐ray Diffraction Data Collected in Oscillation Mode. Methods In Enzymology: Macromolecular Crystallography, Part A 276, 307-‐326 (1997).

4. Mccoy, A.J., Grosse-‐Kunstleve, R.W., Adams, P.D., Winn, M.D., Storoni, L.C., Read, R.J. Phaser crystallographic software. Journal Of Applied Crystallogrphy 40, 658-‐674 (2007).

5. Collaborative Computational Project, Number 4. The CCP4 Suite: Programs for Protein Crystallography. Acta Crystallographica Section D 50, 760-‐763 (1994).

6. Scharff, E.I., Koepke, J., Fritzsch, G., Luecke, C., Rueterjans, H. Crystal structure of diisopropylfluorophosphatase from Loligo vulgaris. Structure 9(6), 493-‐502 (2001).

7. Emsley, P., Cowtan, K. Coot: model-‐building tools for molecular graphics. Acta Crystallographica Section D-‐Biological Crystallography 60, 2126-‐2132 (2004).

8. Krissinel E,. Henrick K. Secondary-‐structure matching (SSM), a new tool for fast protein structure alignment in three dimensions. Acta Crystallographica Section D 60, 2256-‐2268 Part 12 Sp. Iss. 1 (2004).

9. Murshudov, G.N., Vagin, A.A., Dodson, E.J. Refinement of Macromolecular Structures by the Maximum-‐Likelihood Method. Acta Crystallographica Section D 53, 240-‐255 (1997).

10. Winn, M.D., Isupov, M.N., Murshudov, G.N. Use of TLS parameters to model anisotropic displacements in macromolecular refinement. Acta Crystallographica Section D 57, 122-‐133 (2001).

11. Vagin, A.A., Steiner, R.A., Lebedev, A.A., Potterton, L., Mcnicholas, S.L.F., Murshudov, G.N. Organization of Prior Chemical Knowledge and Guidelines for its Use. Acta Crystallographica Section D 60, 2184-‐2195 Part 12 Sp. Iss. 1 (2004).

12. Laskowski, R.A., Macarthur, M.W., Moss, D.S., Thornton, J.M. PROCHECK: a program to check the stereochemical quality of protein structures. Journal Of Applied Crystallography 26, 283-‐291 (1993).


increased diels-alderase activity through backbone remodeling guided … · increased...

Documents