Commissioning Results of The Protein Microcrystallography Beamline TPS 05A

Size: px

Start display at page:

Download "Commissioning Results of The Protein Microcrystallography Beamline TPS 05A"

Constance Carson
5 years ago
Views:

1 Science Advisory Committee Meeting Commissioning Results of The Protein Microcrystallography Beamline TPS 05A Construction Team Leader: Yuch-Cheng Jean

2 Outline I. Beamline Introduction II. Commissioning Results III. Future Works IV. Summary

PX Users Community Currently there are 71 domestic user groups and 42 international user groups Japan (10) Hokkaido University Kyoto University Nagoya University Osaka University Photo Factory

3 PX Users Community Currently there are 71 domestic user groups and 42 international user groups Japan (10) Hokkaido University Kyoto University Nagoya University Osaka University Photo Factory Yokohama City University Korea (1) Yonsei University China (2) Chinese Academy of Science Beijing Normal University Taiwan (71) Hong Kong (4) University of Hong Kong The Chinese University of Hong Kong Thailand (8) Mahidol University National Center for Genetic Engin. and Biotech. Suranaree University of Technology Chulalongkorn University Singapore (17) Institute of Molecular and Cell Biology Nan-Yang Technological University National University of Singapore Singapore Immunology Network

4 Tough Projects viruses and large macromolecular assemblies multi-protein and tertiary complexes membrane proteins pathogenicity-related proteins extreme-environment metalloproteins structure-based drug designs structural genomics projects

5 Challenges & Beamline Requirements Challenges: large unit cell (viruses, ribosomes, large complexes) small & weakly diffracting crystals (membrane proteins) high mosaicity crystals high throughput (structure-based drug screening) Beamline requirements: small beam divergence small beam size high flux density wide and accurate energy tunability 1 Å phasing 2 Å phasing high positional stability beamline automation remote user access

6 Optical Design a 3m in-vacuum undulator (IU22) with a minimum gap of 7 mm a double-crystal monochromator (DCM) a vertical focusing mirror (VFM) a horizontal focusing mirror (HFM) (35 m) (29.85 m) (27 m) (24.5 m)

7 Beam Properties X-ray beam properties Unique features energy range: kev beam sizes: 65 x 36 mm 2 (H x V) beam divergences: mrad (H x V) 12.4 kev: > 1x10 13 photons/s small beam sizes low beam divergences high flux densities high position stability wide and accurate energy tunability beamline automation remote access Aperture Size (mm) Defined Size (mm, H x V) Defined Flux (Phots/sec) Flux Density (phots/s/mm 2 ) Time to Henderson Limit (sec) Full Beam

End Station Features integrated beam shaping 5 to 50 mm beam, user selectable changeable beam divergence 0.

$precision PHI axis < 1 mm SOC, suitable for microdiffraction maximum PHI rotation speed up to 130 deg/s$ suitable for inverse-beam experiments high-speed and large-size area detector suitable for shutterless

suitable for inverse-beam experiments high-speed and large-size area detector suitable for shutterless

8 End Station Features integrated beam shaping 5 to 50 mm beam, user selectable changeable beam divergence 0.1 to 0.5 mrad divergence, user selectable on-axis sample video parallax error-free and real time imaging high precision PHI axis < 1 mm SOC, suitable for microdiffraction maximum PHI rotation speed up to 130 deg/s suitable for inverse-beam experiments high-speed and large-size area detector suitable for shutterless fine-slicing exp. grid scan use the best diffracting area helical scan mitigate the radiation damage automatic sample changer enable high throughput crystal screening remote access capability Microdiffractometer MD2 Fast Detector MX300HS

9 Outline I. Beamline Introduction II. Commissioning Results III. Future Works IV. Summary

Beam Size and Flux 65 36 mm 2 @ full beam 39 25 mm 2 @ 50

photons/s at 1 Å 23 23 mm 2 @ 30 mmf pinhole 15 16 mm 2 @ 20

10 Beam Size and Flux mm full beam mm 50 mmf pinhole full beam = mm 2 (H x V) flux = 1 x photons/s at 1 Å mm 30 mmf pinhole mm 20 mmf pinhole mm 10 mmf pinhole 7 4 mm 5 mmf pinhole

11 Beam Size and Flux Aperture Size (mm) Full Defined Size (mm, H x V) Defined Flux (Phots/sec) Flux Density (phots/s/mm 2 ) Time to Henderson Limit (sec) Cal Intensity Fluctuation 200um Mea um 1.0 % 50 Cal Mea % 100um 30 Cal Mea % 20 Cal Mea % 10 Cal Mea % Cal um 5 Mea % simulation energy at kev, 3 m- IU22, 500 ma, gap 7 mm, with heat load on 1st crystal. measurement energy at kev, 3 m- IU22, 150 ma, gap 7 mm; measured values are scaled to 500 ma. flux density = defined flux / defined size Henderson limit ~ 2E7 Gray ( = total exposure dose for best 2Å data )

12 Beam Position Stability energy scan with knife-edge scanning of a tungsten wire hor. position variation = 23.3 mm p-t-v ver. position variation = 32.9 mm p-t-v energy set to 12.4 kev, monitoring beam image for 12 hours hor. beam drift = 9.2 mm per hour ver. beam drift = 3.4 mm per hour

13 Energy Resolution and Stability rocking curve scan of DCM energy resolution ~ 2.2x10-4 peak shift of Se-edge within 5 hrs energy drift = 0.05 ev per hour

14 Characterization of Beamline Flicker Error At low resolution, systematic error dominates and the ISa value derived from the XDS software is a good indicator for it. The higher flicker error in this beamline really degrade the low resolution data quality a lots comparing with TLS data. It is better to use longer exposure time with beam attenuated instead if a low mosaicity good crystal is used for anomalous phasing. Insulin (size:~ 200 mm) TPS-05A TLS-13B TLS-15A Flux (phs/s) 1.4 x x x Fluctuation (%) Beam size (um) XDS Mosaicity ( ) Distance (mm) 120 mm 200 mm 200 mm Frame width ( ) Exposure time (s) 1 ~ 98 1 ~ 50 1 ~ 54.5 Oscillation range ( ) Completeness (%) Number of data sets 5 8 8

15 Fast Detector Leads to Shutterless Fine-slicing Mode With a high-speed detector and diffractometer, the shutterless fine-slicing data collection is feasible. jitter error can be eliminated background scattering can be reduced but readout noise will increase for CCD drtector Data quality comparison of shutterless and shuttered mode Data quality comparison of shutterless fine-slicing and thick-slicing mode detector : MX300HS sample : insulin resolution : Å dose : 1.5 MGy sample : insulin frame width = 0.75X and 5x mosaicity dose : 0.9 MGy

$Fast Detector Leads to Shutterless Fine-slicing Mode With a high-speed detector and diffractometer, the shutterless fine-slicing data collection is feasible.$

16 Fast Detector Leads to Shutterless Fine-slicing Mode With a high-speed detector and diffractometer, the shutterless fine-slicing data collection is feasible. jitter error can be eliminated background scattering can be reduced but readout noise will increase for CCD detector What is the optimal fine-slicing frame width? photon counting PAD detector no readout noise highest resolution data charge integrating CCD detector readout noise < 1 X-ray photon highest resolution data optimal frame width = 0.5 x mosaicity optimal frame width = 1.0 x mosaicity Frame width/mosaicity Frame width/mosaicity Acta Cryst. (2012). D68, 42-56

17 Throughput Comparison of Shutterless and Shuttered Mode Shutterless mode, 180 images 0.1 sec exposure exposed time: 0.1*180 sec dead time: 0.2 sec total time: ~18.2 sec 1 sec exposure exposed time: 1*180 sec dead time: 0.2 sec total time: ~180.2 sec 10 sec exposure exposed time: 10*180 sec dead time: 0.2 sec total time: ~ sec Shuttered mode, 180 images 0.1 sec exposure exposed time: 0.1*180 sec dead time: 3.1*179 sec total time: ~572.9 sec 1 sec exposure exposed time: 1*180 sec dead time: 3.1*179 sec total time: ~734.9 sec 10 sec exposure exposed time: 10*180 sec dead time: 3.1*179 sec total time: ~ sec ~ 30 time faster for 0.1 sec exposure

18 Micro-beam Provides New Data Collection Strategies Radiation damage has placed a limit on the diffraction resolution on all protein crystals. Since the crystal lifetime is proportion to its scattering volume, it is very difficult to collect a complete high resolution dataset from one microcrystal. How to achieve high resolution from microcrystals or micro-beam? multi-crystal strategy (crystal size < beam size) helical-scan strategy (crystal size > beam size) serial-crystallography approach Multi-crystal small-wedge strategy Helical scan strategy

19 Multi-crystal Small-wedge Strategy microcrystals huge multi-crystal effect typical crystals No multi-crystal effect (Lactase, P222, B = 9.6 Å 2, BEST simulation, 2014 DLS-CCP4 Workshop)

data set collected by normal rotation 5 partial data sets

20 Data Quality Comparison of Normal Rotation and Helical Scan a rod-shaped crystal: 300 x 60 x 60 mm 3 a complete data set collected by normal rotation 5 partial data sets are collected by stepped helical scan 30mm N P2 P4 P1 P3 P5

21 Phasing Power Comparison of Normal Rotation and Helical Scan a rod-shaped crystal: 300 x 50 x 50 mm 3 a sulfur SAD data set collected at 7 kev by normal rot. 5 partial data sets are collected by stepped helical scan 30mm N P1 P2 P4 P3 P5

22 High Resolution Study of Large Homogeneous Crystal sample: Lysozyme (size: 900 x 150 x 150 mm 3 )

9 (6.9) Multiplicity 8.8 (7.4) Completeness (%) 98.5 (91.6) Unique reflections 52694 (2390) Refinement statistics R work / R free (%) 18.7 / 21.0 R.m.s.d.

23 Structure Determination of Large Homogeneous Crystal by MR X-ray statistics Lysozyme Wavelength (Å ) 1.0 Beam size (mm) 50 Dose (MGy) 4.53 Cell dimension (Å ) 78.9 / 78.9 / 37.0 Space group P Resolution (Å ) ( ) Mosaicity R sym (%) 5.9 (27.1) Mean I/s 26.9 (6.9) Multiplicity 8.8 (7.4) Completeness (%) 98.5 (91.6) Unique reflections (2390) Refinement statistics R work / R free (%) 18.7 / 21.0 R.m.s.d. bonds (Å ) R.m.s.d. angles ( o ) Overall B factor (Å 2 ) (Counter level: 1.8 s) PDB structure (2LYZ) TPS structure a backbone RMSD value of Å over 121 Ca atoms

Structure Determination of Large Homogeneous Crystal by S-SAD X-ray statistics Insulin Wavelength (Å ) 1.0 Beam size (mm) 50 Dose (MGy) 3.04 Cell dimension (Å ) 78.

8) Completeness (%) 100.0 (100.0) Unique reflections 17407 (1711) Refinement statistics R work / R free (%) 18.4 / 19.5 R.m.s.d. bonds (Å ) 0.022 R.m.s.d. angles ( o ) 1.

24 Structure Determination of Large Homogeneous Crystal by S-SAD X-ray statistics Insulin Wavelength (Å ) 1.0 Beam size (mm) 50 Dose (MGy) 3.04 Cell dimension (Å ) 78.0 / 78.0 / 78.0 Space group I2 1 3 Resolution (Å ) ( ) Mosaicity R sym (%) 6.2 (152.4) Mean I/s 81.9 (2.1) Multiplicity 37.7 (32.8) Completeness (%) (100.0) Unique reflections (1711) Refinement statistics R work / R free (%) 18.4 / 19.5 R.m.s.d. bonds (Å ) R.m.s.d. angles ( o ) Overall B factor (Å 2 ) (Counter level: 1.8 s) RMSD = Å (PDB / S-SAD, 47 Ca atoms) RMSD = Å (PDB / MR, 49 Ca atoms) RMSD = Å (S-SAD / MR, 46 Ca atoms)

The 1 st User Group and Experiment 2015.12.

25 The 1 st User Group and Experiment TPS-05A beamline operation team and Prof. Andrew Wang s group members

$255 Å over 224 Ca atoms diffraction resolution is 2.$

26 The 1 st User Group and Experiment The 1 st protein structure solved at TPS 05A TPS structure TLS structure a backbone RMSD value of Å over 224 Ca atoms diffraction resolution is 2.15 Å a substrate was found in the active center

27 The 1 st User Group and Experiment The 1st paper using TPS-05A data was submitted Structural Elucidation of a Z,Z-Farnesyl Diphosphate Synthase from the Wild Tomato Solanum habrochaites (Plant Physiology, submitted) User s Comments: data collection time is less than 1 min, very fast the computing system for data processing is very efficient support staff is very patient and professional very nice working environment

28 Commissioning Phase User Experiments May 26 th to June 13 th : large unit cell:c.-j. Chen small crystal:a.-c. Ma ultra-high resolution:t.-l. Li high-throughput screening:a.-h.-j. Wang important protein:n.-l. Chan, Y.-J. Sun, Prof. M.-C. Ho, User s Experiences of TPS-05A

29 Outline I. Beamline Introduction II. Commissioning Results III. Future Works IV. Summary

30 Future Works make the beam more stable, smaller and stronger DI/I = 2% for 5 mm beam < 1% full beam size ~ 65 x 36 mm 2 55 x 21 mm 2 flux density ~ low DCM heat load effect rigidity of mirror adjustment mechanism and support structure rigidity of diffractometer support structure reduction of thermal expansion by temperature control active and passive vibration attenuation slope error of mirror surface minimum gap of undulator increase the level of beamline automation establish the automatic data collection and processing workflows for multi-crystal data collection strategy for microcrystals user training July 2016, Micro-beam data collection and processing workshop

31 Outline I. Beamline Introduction II. Commissioning Results III. Future Works IV. Summary

32 Summary We have built a low-divergence, high-intensity, energy-tunable and small sample available beamline in TPS. Using apertures makes TPS 05A a suitable facility on which to perform experiments on all but the smallest of microcrystals. This beamline will open many scientific opportunities and will provide user in Taiwan and surrounding areas an essential tool to tackle difficult but important biostructures. In-house X- ray machine TLS PX beamlines TPS 05A beamline Suitable unit cell < 150 Å < 350 Å < 1000 Å Suitable crystal size > 200 mm 200~50 mm 50~5 mm Data collection time ~24 hr ~0.5 hr ~0.5 min Scientific opportunities: viruses and large macromolecular assemblies multi-protein and tertiary complexes membrane proteins pathogenicity-related proteins extreme-environment metalloproteins structure-based drug designs structural genomics projects Challenging crystallography: large unit-cell crystals microcrystals high mosaicity crystals weakly diffracting crystals structural genomics projects structure-based drug designs Å MAD/SAD exp.

33 Summary How NSRRC Help National Program in Taiwan : National Research Program for Genomic Medicine (NRPGM) One of the focuses: Structural genomics research NSRRC TLS 13B1 & 13C1, high-throughput crystallography beamline 2005 and 2006 open to user : National Research Program for Biopharmaceuticals (NRPB) One of the focuses: Targeted drug discovery NSRRC TLS 15A1, high-throughput crystallography beamline 2013 open to user : Taiwan Protein Project (TPP) Flagship project: Structural analysis of cancer related protein-phosphoprotein complexes involving kinases and phosphatasesrelated to cancer NSRRC TPS 05A, micro-crystallography beamline 2016 open to user

34 Acknowledgements Many thanks to the technical help from: SSRL-MXG, for the automatic data acquisition and control system BluIce/DCSS. RIKEN SPring-8 Center, for the automatic multi-dataset processing workflow KAMO. Protein Diffraction Group F. Chao C.H. Huang C.C. Tseng C.K. Chou C.H. Chiang Y.H. Chen Y.C. Liu W.T. Huang Construction Team: Beamline Group D.G. Liu C.H. Chang C.Y. Liu C.F. Chang H.W. Chen

35 Commissioning Phase User Experiments May 26 th to June 13 th : large unit cell:c.-j. Chen small crystal:a.-c. Ma ultra-high resolution:t.-l. Li high-throughput screening:a.-h.-j. Wang important protein:n.-l. Chan, Y.-J. Sun, Prof. M.-C. Ho, User s Experiences of TPS-05A

Large Unit Cells Crystal Size: 350 x 420 x 50 µm 3 Space Group: P2 1

86 Å Resolution: 3.45 Å (> 2.5σ) 3.20 Å (> 1.2σ) Completeness: 97.

36 Large Unit Cells Crystal Size: 350 x 420 x 50 µm 3 Space Group: P Cell Parameters: a = Å b = Å c = Å Resolution: 3.45 Å (> 2.5σ) 3.20 Å (> 1.2σ) Completeness: 97.5% Data merged from 4 data sets (0.3 /frame), 1 crystal <By courtesy of C.-J. Chen>

Large Unit Cells Crystal Size: 350 x 420 x 50 µm 3 Space Group: P2 1 2 1 2 Cell Parameters: a = 209.56 Å b = 731.00 Å c = 192.86 Å Resolution: 3.

37 Large Unit Cells Crystal Size: 350 x 420 x 50 µm 3 Space Group: P Cell Parameters: a = Å b = Å c = Å Resolution: 3.45 Å (> 2.5σ) 3.20 Å (> 1.2σ) Completeness: 97.5% Data merged from 4 data sets (0.3 /frame) at 4 regions of 1 crystal <By courtesy of C.-J. Chen>

38 Large Unit Cells Comparison between TLS and TPS TLS BL15A TPS 05A <By courtesy of C.-J. Chen>

86 Å Number of reflections : 391,968 Number of

39 Large Unit Cells a = Å, b = Å, c = Å Number of reflections : 391,968 Number of atoms : 44,022 Current structure refinement: R/R free = 0.30/0.38 <By courtesy of C.-J. Chen>

40 Ultra-high Resolution Two crystals: ~ 400 x 400 x 100 mm 3 Two ultra-high data sets collected at 19 kev by continuous helical scan Two data sets are merged to obtain 0.85 Å resolution data 30mm 0.86Å <By courtesy of T.-L. Li>

Ultra-high Resolution Data collection parameters and data processing statistics - Space group: C222 1 - Cell parameters: 101.2 185.3 99.2 90.0 90.0 90.0 crystal1 crystal2 merge No.

41 Ultra-high Resolution Data collection parameters and data processing statistics - Space group: C Cell parameters: crystal1 crystal2 merge No. of frames/ oscillation angle ( ) 600/ / / 0.2 Exposure time per image (s) Dose (MGy) Mosaicity Resolution (Å ) Completeness (%) 99.0 (97.9) 99.0 (97.8) 99.9 (99.8) I/s 9.37 (2.04) (2.02) (2.08) R merge (%) 7.2 (58.4) 5.8 (63.9) 8.8 (81.0)

$Small and Weakly Diffracting Crystals:$ $and weakly diffracting crystals grown from LCP$ mesophase becomes opaque upon snap-cooling

42 Small and Weakly Diffracting Crystals: Multi-crystal Data Collection Challenges: Small and weakly diffracting crystals grown from LCP The crystals are no longer visible because the mesophase becomes opaque upon snap-cooling Raster Scan Phi = 150º Phi = 240º 3.5Å <By courtesy of A.-C. Ma>

43 Small and Weekly Diffracting Crystals: Multi-crystal Data Collection 17 crystals are used to collect data at 12.7 kev and a complete data set is obtained to 3.36 Å diffraction resolution - Space group: P2; Cell parameters: No. of frames/ oscillation angle ( ) 1390/ 1 Exposure time per image (s) 1 Dose (MGy) data sets merge Mosaicity Resolution (Å ) Completeness (%) 99.6 (99.7) I/s 7.14 (2.04) R merge (%) 54.8 (82.5)

$diffracting Phi = 229.$

44 Large Inhomogeneous Crystals: Multi-position Data Collection Challenges: The crystal is large but inhomogeneous Weakly diffracting Phi = 229.6º Raster Scan (Beam size: 20 mm) Phi = 319.6º <By courtesy of S.-J. Sun>

$25 Å diffraction resolution Space group: C2; Cell parameters: 203.364 111.307 121.920 90.0 92.197 90.0 position1 Position2 position3 merge No. of frames/ oscillation angle ( ) 310/ 0.5 310/ 0.$

45 Large Inhomogeneous Crystals: Multi-position Data Collection 3 positions are used to collect data at 12.7 kev and a complete data set is obtained to 3.25 Å diffraction resolution Space group: C2; Cell parameters: position1 Position2 position3 merge No. of frames/ oscillation angle ( ) 310/ / / / 0.5 Exposure time per image (s) Dose (MGy) Mosaicity (º) Resolution (Å ) Completeness (%) 97.3 (93.7) 98.7 (95.1) 66.2 (62.9) (100.0) I/s 5.97 (1.06) 5.22 (0.82) 4.36 (0.64) 8.73 (2.03) R merge (%) 22.4(122.8) 28.8 (134.4) 12.8 (105.7) 33.4 (147.7) Redundancy 3.0 (2.8) 3.0 (2.3) 1.7 (1.5) 7.2 (6.0)

46 Se-SAD Phasing Energy: 12.7 kev Data resolution: 1.6 Å 1 Se atom in one protein (119 residues) ARP/wARP traced the model of ~100%, except for disorder N- and C-terminals <By courtesy of C.-J. Chen>

strategy: helical scan, merging data from multiple crystals) TLS BL15A1, 2.

47 Summary of Users Assessment World quality beam: extremely intense, tunable beam size (5-50 μm) and energy Fast data collection Excellent beamline support staff Radiation damage (careful planning of the data collection strategy: helical scan, merging data from multiple crystals) TLS BL15A1, 2.75 Å X-ray absorption fine structure spectroscopy (XAFS) TPS 05A1, 2.7 Å Computing power Debugging. (Blu-Ice, k-arc, ) Training courses In-ring dinning areas

48 Thank You for Your Attention

Nature Methods: doi: /nmeth Supplementary Figure 1. Resolution of lysozyme microcrystals collected by continuous rotation.

Nature Methods: doi: /nmeth Supplementary Figure 1. Resolution of lysozyme microcrystals collected by continuous rotation. Supplementary Figure 1 Resolution of lysozyme microcrystals collected by continuous rotation. Lysozyme microcrystals were visualized by cryo-em prior to data collection and a representative crystal is