GE s DeltaVision OMX Microscope

Transcription

1 GE s DeltaVision OMX Microscope SIM Theory, Sample Prep & Analysis Patrina Pellett, PhD Applications Scientist GE Healthcare, Cell Technologies May 27th, 2015

2 OMX Capabilities Summary 3D structured illumination microscopy Theory Applications Sample prep Reconstructing SIM images & artifacts Spherical aberration caused by refractive index mismatch Photobleaching Honeycomb OMX Seminar May

3 DeltaVision OMX Types of imaging experiments possible DV Localization Microscopy Sub diffraction limit resolution 2D & 3D-SIM Ring TIRF & PK DV OMX Compatible with common probes OMX Seminar May 2015 Ultrafast widefield imaging 3

4 SIM Theory

5 3D-Structured Illumination Microscopy How does this technique work? Use 3D striped light pattern to illuminate sample Extract high resolution information in Fourier space 2 fold increase in resolution both laterally & axially Large contrast improvement in XY & Z Compatible with standard dyes & fluorescent proteins OMX Seminar May

6 Moiré Fringes Encode high resolution information Stephen Cody, Monash Micro Imaging OMX Seminar May

7 Moiré Fringes Encode high resolution information Stephen Cody, Monash Micro Imaging OMX Seminar May

8 How is the striped pattern generated? Use interference to get 3D pattern at sample Back aperture Split the laser with a diffraction grating. Focus 3 beams onto back aperture to create a sinusoidal pattern at the sample plane. OMX Seminar May

9 3 Beam Interference Creates Pattern in Z 3D pattern gives 2x increase in axial resolution Focal Plane Z X OMX Seminar May

10 3D-SIM Imaging Image with striped pattern at 5 phases & 3 angles Phase 1 Phase 2 Phase 3 Phase 4 Phase 5 Angle 3 Angle 2 Angle 1 OMX Seminar May

11 Structured Illumination Imaging Image sample with a striped pattern 1 OMX Seminar May

12 Structured Illumination Imaging How many images are required? 3D-SIM: 135 images per micron per color 5 phases per angle (5 images) 3 angles per slice (3 x 5 images) 9 slices per micron (3 x 5 x 9 images) 125 nm step size 2D-SIM: 15 images per slice 5 phases per angle (5 images) 3 angles per slice (3 x 5 images) OMX Seminar May

13 SIM Theory Summary Illuminate with striped pattern at 5 phases & 3 angles Create Moiré fringes Extract high frequency info in Fourier space Obtain high resolution info not accessible by conventional microscopy OMX Seminar May

14 Structured Illumination Microscopy Benefits of using this technique Advantages of SIM ~ nm resolution XY Large volumes possible ~ nm resolution Z Compatible with many sample types Compatible with standard probes Up to 4 colors in SIM Live cell compatible OMX Seminar May

15 DLM vs 3D-SIM Things to consider when choosing a method Localization Microscopy Down to 20 nm resolution Special probes required Multi-color tricky Limited live cell Limited z range TIRF region or very clean samples required Structured Illumination ~ nm resolution Compatible with common probes Up to 4 colors in superresolution Live cell compatible Large volumes possible Compatible with many sample types OMX Seminar May

16 OMX SIM Applications

17 DeltaVision OMX Capabilities SIM Summary 2D or 3D-SIM possible 4 EMCCD cameras 100x 1.42 NA objective 405, 488, 561, 642 nm lasers λ Resolution* XY (nm) Resolution* Z (nm) Blue Green Red Far Red *measured on bead samples OMX Seminar May

18 OMX 3D-SIM Large Z stacks possible OMX 3D-SIM Volume View Lanni Lab, CMU 20µm thick Z stack! OMX Seminar May

19 OMX 3D-SIM on Thick Tissue Improved contrast & resolution TNFR2 Iba-1 Overlay Widefield OMX 3D-SIM 10 µm Max projections Bethea Lab, Drexel University OMX Seminar May

20 OMX 3D-SIM on Thick Tissue 14µm thick SIM stack of brain tissue Widefield Volume View OMX 3D-SIM Volume View Bethea Lab, Drexel University TNFR2 Iba-1 OMX Seminar May

21 OMX 3D-SIM Resolve centrosome structure Widefield OMX 3D-SIM 5 µm Max projections Mennella Lab, Sick Kids OMX Seminar May

22 OMX 3D-SIM Centrosome structure resolved in multi-colors Green Red Overlay OMX 3D-SIM Widefield 1 µm Mennella Lab, Sick Kids Max projections OMX Seminar May

23 OMX 3D-SIM Resolve inner & outer mitochondria Widefield Z stack movie OMX 3D-SIM Z stack movie Vaughan Lab, UW OMX Seminar May

24 OMX 3D-SIM Resolve ring structure at base of cilia Green Red Overlay Widefield OMX 3D-SIM 1 µm 1 µm Shah Lab, Harvard Max projections OMX Seminar May

25 OMX Live Cell 3D-SIM Imaging Resolve primary cilia structure Widefield Image OMX 3D-SIM Image Max projection Sec61β Shah Lab, Harvard OMX Seminar May

26 OMX Live Cell 3D-SIM Imaging Image ER dynamics at 123 frames per second Widefield Image OMX 3D-SIM Image Max projection Sec61β Vaughan Lab, UW OMX Seminar May

27 OMX Live Cell 3D-SIM Imaging 2-color imaging for 200 time points at 96 fps Widefield Image OMX 3D-SIM Image Max projection Vaughan Lab, UW OMX Seminar May 2015 Mito-GFP ER-mCherry 27

28 OMX Live Cell 3D-SIM Imaging Image mitosis with minimal phototoxicity Widefield Image OMX 3D-SIM Image 915 images per time point! Max Projections Scott Lab, UW & HHMI EB3-GFP OMX Seminar May

29 Sample Prep Recommendations for SIM

30 SIM Sample Prep Recommendations Mount sample on #1.5 coverslips #1.5 Coverslips are 170 µm thick High precision #1.5 coverslips require less optimization! Objective Microscopes are optimized for #1.5 cover glass! OMX Seminar May

31 SIM Sample Prep Recommendations Do NOT mount sample on glass slide Glass Slide x #1.5 Coverslip Mounting on the slide introduces an uncontrollable variable! Objective The working range of a high NA objective is very small <200µm! OMX Seminar May

32 SIM Sample Prep Recommendations Mount sample on #1.5 coverslips Glass Slide #1.5 Coverslip Mounting on the coverslip helps keep the sample in the working range of the objective! Objective OMX Seminar May

33 SIM Sample Prep Recommendations Do NOT mount tissue sections on glass slide Glass Slide x Tissue Section #1.5 Coverslip Mounting on slide introduces additional uncontrollable distances Objective Tissue sections are not flat Uneven surface can add additional distance from objective Makes matching RI difficult OMX Seminar May

34 SIM Sample Prep Recommendations Mount tissue section on #1.5 coverslip Glass Slide Tissue Section #1.5 Coverslip Objective Keeps sample within working range of objective & can help with imaging deeper structures OMX Seminar May

35 SIM Sample Prep Recommendations Do not use DAPI in the mounting media Glass Slide Coverslip DAPI can cause high levels of background! To image with DAPI, perform a separate labeling step & wash well! Objective Use glycerol based mounting media with an anti-fade agent OMX Seminar May

36 OMX Sample Prep Recommendations Mount with non-hardening mounting media Non-hardening varieties (Vectashield) help preserve 3D structure OMX Seminar May

37 OMX Sample Prep Recommendations Mount in the middle of 75mm glass slide 75 mm Slide Frosted area 25 mm Mount coverslip in the middle 25 mm of slide 25 mm 25 mm OMX Seminar May

38 OMX Sample Prep Recommendations Other compatible sample holders 35 mm imaging dishes (#1.5 coverslip) 4 well or 8 well chamber dishes, middle wells (#1.5 coverslip) OMX Seminar May

39 OMX Sample Prep Recommendations Use dyes that match lasers & filter settings Available laser lines for 3D-SIM Imaging: 405, 488, 561, 642nm Possible filter setting combinations: Channel Emission Filters (nm) DAPI FITC mcherry Cy OMX Seminar May

40 SIM Sample Prep Recommendations Summary Least strict requirements of all the super resolution methods Get sample as close as possible to #1.5 coverslip Use photostable dyes Consider using antifade agents in mounting media Minimize refractive index mismatch OMX Seminar May

41 DLM vs 3D-SIM Things to consider when choosing a method Localization Microscopy Down to 20 nm resolution Special probes required Multi-color tricky Limited live cell Limited z range TIRF region or very clean samples required Structured Illumination ~ nm resolution Compatible with common probes Up to 4 colors in superresolution Live cell compatible Large volumes possible Compatible with many sample types OMX Seminar May

42 Have questions? Need help? Need support? Call or me! OMX Seminar May

43 Reconstructing SIM Data Minimizing artifacts caused by refractive index mismatch

44 Reconstructing SIM Data How to minimize haloing or doubling artifacts Mismatched oil Matched oil RI Immersion Oil RI Immersion Oil Haloing/doubling artifacts can be minimized by optimizing the refractive index of the immersion oil OMX Seminar May

45 Reconstructing SIM Data What causes the haloing or doubling artifact? PSF with Mismatched oil PSF with Matched oil Z X RI Immersion Oil RI Immersion Oil SIM algorithm assumes a perfectly matched PSF! When it detects out of focus light from mismatched PSF, it assumes this is real signal & reconstructs it OMX Seminar May

46 Reconstructing SIM Data Minimize haloing artifact by matching the RI of immersion oil to your sample The OMX comes with an oil kit with 18 oils with different refractive indices ranging from to in increments of OMX Seminar May

47 Reconstructing SIM Data How to find the best oil to minimize artifacts RI Immersion Oil Z Coverslip Focal Plane Into the Sample X If the flaring/ asymmetry of the PSF is towards the coverslip or up, the RI of the oil is too high! Go down in oil! OMX Seminar May

48 Reconstructing SIM Data How to find the best oil to minimize artifacts RI Immersion Oil Movie Starts at Coverslip If the flaring/ asymmetry of the PSF is towards the coverslip or up, the RI of the oil is too high! Go down in oil! OMX Seminar May

49 Reconstructing SIM Data How to find the best oil to minimize artifacts RI Immersion Oil Z Coverslip Focal Plane Into the Sample X If the flaring/ asymmetry of the PSF is away from the coverslip or down, the RI of the oil is too low! Go up in oil! OMX Seminar May

50 Reconstructing SIM Data How to find the best oil to minimize artifacts RI Immersion Oil Movie Starts at Coverslip Coverslip Focal Plane Into the Sample If the flaring/ asymmetry of the PSF is away from the coverslip or down, the RI of the oil is too low! Go up in oil! OMX Seminar May

51 Reconstructing SIM Data How to find the best oil to minimize artifacts RI Immersion Oil Z Coverslip Focal Plane Into the Sample X If the flaring of the PSF is symmetrical on both sides of the focal plane, the oil is matched! OMX Seminar May

52 Reconstructing SIM Data How to find the best oil to minimize artifacts RI Immersion Oil Movie Starts at Coverslip Coverslip Focal Plane Into the Sample If the flaring of the PSF is symmetrical on both sides of the focal plane, the oil is matched! OMX Seminar May

53 Reconstructing SIM Data Affects of RI mismatch on SIM reconstructions Immersion Oil Immersion Oil Immersion Oil Immersion Oil 5 µm Max Projections Haloing or doubling artifacts are minimized in SIM reconstructions with the oil used results in a symmetric PSF OMX Seminar May

54 Reconstructing SIM Data Matching oil for multi-color samples Can only match the oil for 1 wavelength Meaning the wavelengths with suboptimal oil matching will have some haloing or doubling artifacts What can be done about this? 1. Match for the most red shifted wavelength because the blue shifted wavelengths are more forgiving 2. Match for the color you care the most about OMX Seminar May

55 Reconstructing SIM Data Which oil should you start with? Excitation Wavelength (nm) Imaging at the coverslip Imaging + 5µm from coverslip Imaging + 10µm from coverslip For imaging at 37 C increase oil by or 4 steps in oil kit For every + 5µm imaging into the sample go up or 1 step in the oil kit OMX Seminar May

56 Reconstructing SIM Data Minimizing artifacts caused by photobleaching

57 Reconstructing SIM Data Photobleaching gives artifact similar to haloing Anything that gives an asymmetric PSF will give the haloing/doubling artifact Do not bleach more than ~30% across the stack If have square edges in reconstruction, change imaging parameters to minimize photobleaching: Lower %T, increase exposure Increase %T, decrease exposure Smaller z stack (fewer # of optical sections) Use all channel then z imaging setting Use antifade reagents OMX Seminar May

58 Reconstructing SIM Data Photobleaching gives square edges Z stack movie Max Projection Sample bleached by ~90% while acquiring a 3 µm z stack OMX Seminar May

59 Reconstructing SIM Data Pay attention to max counts during acquisition Take note of max intensity count before starting experiment Watch how values change during acquisition Adjust imaging parameters to keep bleaching under 30% OMX Seminar May

60 Reconstructing SIM Data Minimizing the honeycomb artifact

61 Reconstructing SIM Data Diffuse labeling & bad S/N gives honeycomb Regular/hexagonal pattern is referred to as honeycomb Common with diffuse labeling or samples with bad S/N Caused by: Little to no mid range frequencies to detect (imaging the SIM pattern) Masking of the SIM pattern by high background Going from Fourier space to real space Typical honeycomb OMX Seminar May

62 Reconstructing SIM Data How to minimize the honeycomb artifact Sometimes you can t: Not all samples are appropriate for SIM Diffuse labeling and high background will give honeycomb SIM works best with samples that have discrete structures Increase S/N without photobleaching Increase Wiener filter constant This removes high frequency data Decreases resolution! Have good controls OMX Seminar May

63 Reconstructing SIM Data Get best S/N to help minimize honeycomb Low Signal to Noise High Signal to Noise S/N = 11 S/N = 57 Increase %T or exposure until the background signal starts increasing OMX Seminar May

64 Reconstructing SIM Data Increase Wiener C to reduce honeycomb Wiener C = Wiener C = Wiener C = SIM Image - Tubulin SIM Image - DAPI 5 µm Single Slice OMX Seminar May

65 Reconstructing SIM Data Affect of using different Wiener Filter Values As this constant is increased: The honeycomb pattern starts to decrease The DAPI signal starts to look smoother and less noisy By increasing this constant you remove high frequency info! You lose resolution by doing this! Typical values are For diffuse labeling or bad S/N use or higher OMX Seminar May

66 Adjusting the Wiener Filter Constant Increasing this constant decreases resolution Wiener C = Wiener C = SIM Image - Tubulin FWHM = 140 nm FWHM = 167 nm Reconstructing with the higher Wiener filter constant resulted in a loss of 27nm in resolution! OMX Seminar May

67 Reconstructing SIM Data How to adjust the Wiener filter constant 1. In SI reconstruction window adjust values in the Wiener Filter Constant box 2. For data with good signal to noise ratio typical values are Adjusting values in this box applies the same Wiener Filter Constant to all channels 4. For structures with diffuse labeling or bad signal to noise, try increasing this value to or higher OMX Seminar May

68 Reconstructing SIM Data How to use channel specific Wiener Filters In the SI Reconstruction tool, check the Use Channel-Specific Wiener Filters 2. Click on Set Up Wiener Filters to open the OMX Channel Wiener Filter Options 3. Select the correct drawer 4. Input Wiener Filter constants for each channel, i.e. higher values for diffuse signals like DAPI & lower values for other channels OMX Seminar May

69 Reconstructing SIM Data How to minimize the honeycomb artifact Not possible with all samples! Increase signal to noise without photobleaching For diffuse labeling increase Wiener filter constant Use for normal samples Use for diffuse labeling (results in a loss of resolution!) Report this value in the methods section of your paper! OMX Seminar May

70 SIM Summary

71 Structured Illumination Microscopy Key factors to successful SIM imaging Get structure of interest as close as possible to coverslip Determine the best immersion oil for your sample Get highest signal to noise ratio possible without photobleaching OMX Seminar May

72 DeltaVision OMX Types of imaging experiments possible DV Localization Microscopy Sub diffraction limit resolution Live cell 3D- SIM with Blaze Ring TIRF & PK DV OMX Compatible with common probes OMX Seminar May 2015 Ultrafast widefield imaging 72

73 GE, imagination at work and GE monogram are trademarks of General Electric Company. DeltaVision is a trademark of GE Healthcare companies. DeltaVision OMX products are for research use only - not for use in diagnostic or therapeutic procedures. All goods and services are sold subject to the terms and conditions of sale of the company within GE Healthcare which supplies them. A copy of these terms and conditions is available on request. Contact your local GE Healthcare representative for the most current information. Alexa Fluor is a registered trademark of Life Technologies Inc. ATTO is a registered trademark of ATTO-TEH GmbH 2013 General Electric Company All rights reserved. First published in February 2012 GE Healthcare Bio-Sciences Corp. 800 Centennial Avenue P.O. Box 1327 Piscataway, NJ USA

74