Post on 14-Dec-2015
Seeing is believing? Practical tips on designing your Microscopy experiments
CMI Light Microscopy Lecture Series 2015
Question: what is the best way to image my sample?Answer: Choose the right methodology…..Question: what is the right methodology / to use?
Brightfield Microscopy
Phase Contrast DIC
Reflected Light
Transmitted Light
Fluorescence Microscopy
Widefield Scanning Confocal
Spinning disk Confocal
Darkfield
Multiphoton
TIRF
DeconvolutionMicroscope orientation : Upright/InvertedAlso, Fluorescence Techniques: Photoactivation, Photobleaching, FRAP, FRET, FLIM, FCS,
STED
PALM , STORM
Thick opaque
What do you want to see ? Is your sample small, dim or live?What Spatial/Temporal resolution do you need?What is it made of ?Do you need labels - Stained/unstainedIf you have to section – how thick?
Lightsheet
Orthogonal plane
Multiphoton
Incubated system
brightfield system
Which system should I use?
Spinning disk Thin samples
Remedi systems
• Upright fluorescence Microscope - volocity• Inverted transmitted light microscope and
fluorescence • Upright brightfield microscope • contact the CMI for any other imaging• www.imaging.nuigalway.ie
Question:
What is the best overall microscope for fluorescent microscopy applications?
• Anyone say confocal?• ok…Sharp clear images , 3d rendering, good for
bright fixed samples. • But everything in imaging is a tradeoff….• Confocal vs widefield – which to use?• confocal microscopes are still constrained by the
same diffraction limit of light as any other conventional (i.e., non-superresolution) microscope. Obtainable Resolution is the same.
What key factor makes the fluorescence microscope work?
What key factor makes the Confocal microscope work?
Scanning Confocal Widefield
Pinhole rejects out of focus lightOptical Sectioning ability
Entire field of view is observed
Works ok with 30 micron to 100 micron
Difficult with sections thicker than 30 micron
Point scan , time consuming, not real time
quick
Most of emission light is blocked Most emission light gets to the detector
High energy light sources necessary – photo damage of sample is concern for live cells and bleaching fluorophores
Lower energy sources
Frame rates in scanning systems low
Higher frame rates
Comparisons
A microscope for live imaging
Relative Sensitivity
• Widefield 100• Spinning-Disk Confocal 25• Laser-scanning Confocal 1
• See Murray JM et al, J. Microscopy 2007 vol. 228 p390-405
Other approaches to cleaning up out of focus light
• TIRF membrane localised events – 100 nm• Multiphoton: localised excitation, inherent optical
sectioning , deep penetration into thick samples, reduced photobleaching
• light-sheet microscopes illuminate sample planes one at a time with relatively low-intensity light, imaging in an orthogonal direction. thick samples , whole organisms, live samples, low bleaching.
• But are expensive set ups and not available NUIG.
Which imaging technique should I use?
Sample Thickness
Wide-field (+deconvolution)
Point scanning Confocal
2-photon confocal
TIRF (for samples at the coverslip)
Spinning Disk Confocal
Line-scanning confocal
1-5 mm
1-20 mm
10-100 mm
>20 mm
>50 mm
Sens
itivi
ty
FastSlow
Setting up the right experimental conditions
• sample prep is keyGarbage in = Garbage out
• Check your staining is optimised before spending time on a high end microscope
• Will you image live cells or fixed cells? • How will you stain them? • what will you grow them on? • What objectives will you use to image your cells,
and what tradeoffs does that decision represent?
Setting up the right experimental conditions
• Plan experiment carefully• Köhler illumination is recommended to produce
equally distributed transmitted light and ensure the microscope reaches its full resolving potential.
• Match sample to objective• Use correct immersion oil• Use correct cover slip , plates etc – optical grade• Transmitted light – don’t forget, not all about
fluorescence! overlay fluorescence and transmitted light image to generate amazing structural context
The most important piece of the optics in a microscope : the Objective
Enables ‘parfocality’ – changing objectives with minimal refocussing
NA = nsinθFinest observable details given by λ/2NA
r = 0.61λ/NAobj
Correct for aberrations
Compound lenses can correct aberrations Achromat: limited color correctionFluorite: good all around objectives, NA up to 1.3Apochromat: Highly color correctedPlan Apochromat: corrected for field curvature
Glass has a different refractive index according to each wavelength of light.
difference in focal length according to each wavelength
different colours are focused at different points.
Working Distance
In general, high NA lenses have short working distances
However, extra-long working distance objectives do exist
Some examples:
10x/0.3 WD = 15.2mm
20x/0.75 WD = 1.0mm
100x/1.4 WD = 0.13mm
Resolution of the Microscope
-2 -1 0 1 2
-2
-1
0
1
2
limited by the point-spread function
Y
Z
X
•Microscope objective collects a limited cone of light from the sample
•This limits the resolution achievable by the microscope
•Resolution can be measured by the blurring of a point object → the point-spread function
Resolution of the Microscopelimited by the point-spread function
NA X-Y Z
0.3 1017 16830
0.75 407 2690
0.95 321 1680
1.4 218 770
Light-gathering power
NA Brightness
0.3 0.09
0.75 0.56
0.95 0.90
1.4 1.96
Light-gathering power proportional to the square of NA
All things being equal, a higher NA lens will give a brighter image
Increasing magnification generally decreases brightness as light is spread out over more pixels
Choosing an objective
• Questions:– What resolution do you need?– How bright is your sample?
• For high resolution, you’ll need high NA.• For dim samples, you’ll want high NA, regardless of
resolution, to maximize light-gathering.– Dim, low-resolution samples (e.g. protein abundance in
nucleus):– Possible solution: bin camera to trade off resolution for
brightness
Choosing an objective
• Questions:– What resolution do you need?– How bright is your sample?
• When to use low NA?– Bright samples at low resolution / low magnification– If you need long working distance– If spherical aberration is a concern– If you want large depth of field to get whole structures
in focus at once (avoid Z-stacks)
Cover slips• Most objectives are matched to #1.5 or 170 micron
thickness coverslips• Its all about refractive index matching and reduction
of diffraction• Sample RI – Glass RI (1.52 – Culture medium
RI(1.33)• Some objectives have correction collar, allow for
variations in the coverslip thickness.
Immersion media• Five types of objectives are commonly used in biological research:
dry, water-immersion, oil-immersion, glycerin-immersion, and water-dipping objectives.
• Dry – used in air• Immersion – drop of liquid between sample and lens front surface. • Water dipping allow immersion in media (upright systems) • Oil has an RI (~1.52) very similar to glass. • oil immersion lens effectively couples the glass coverslip and glass
objective in a block of uniform material. • Glycerin has an RI of ~1.47, making it compatible with some
common sample mounting media, • water’s RI is ~1.33. • Matching the immersion media to the sample is critical for
microscopy.
Immersion media cont’• Example : oil-immersion objective imaging into an
aqueous media (e.g., culture media) will create spherical aberrations due to a mismatch in RIs.
• Spherical aberration is an optical distortion that causes objects to appear stretched or compressed, and which also results in an apparent loss of photons.
Spherical aberration prop to NA3
Use correction collar
Immersion oil considerations
Improving image contrast in fluorescence microscopy: Fluorescence image of stained cells using low autofluorescence immersion oil (left) and general immersion oil.
Which oil immersion lens to use?
• 60x / 1.2 for samples with RI ≈ 1.33• 60x/1.3 silicone for samples with RI ≈ 1.4• 60x/1.42 oil for samples with RI ≈ 1.52
Methyl Salicylate ( thick specimens) 1.53-1.54Living cells varies 1.35-1.4Fixed Cells depends on mounting Tissue 1.515Immersion Oil 1.515Glycerol 1.47Vectashield 1.45DPX 1.525Gel/Mount (Biomedia) 1.36Coverslip Glass 1.52Water 1.33Air 1.0
Some different refractive indices
But η prop T and λ!!!
RI of commonly imaged tissues
Biswas and Gaupta 2002
Correct use of immersion oil• See guidelines (http://
imaging.nuigalway.ie/equipment/fluoview1000brb.html)
• Never use oil on air objective• Know how to distinguish air/immersion objectives• Always check what objective you are using – know
the NA• Know how to clean oil lenses and how to use oil on
lenses
Microscope Care• Keeping the microscope optics clean is
important for high-quality imaging. Dust, fingerprints, excess immersion oil, or mounting medium on or in a microscope causes reduction in contrast and resolution. DIC is especially sensitive to contamination and scratches on the lens surfaces.
• Always keep microscopes covered when not in use.
• Make sure that all ports, tubes, and unoccupied positions on the lens turrets are plugged.
• Plastic plugs are usually supplied with the microscope.
Areas to watch out for• The external surface of the front lens of the objective• The surface of the camera sensor and its protective
glass cover• Both surfaces of the cover slip• The surface of the microscope slide• The surface of the camera adapter optics• Surface of the upper lens of the condenser• Other glass surfaces in the light path e.g. bulbs
of halogen- or high pressure lights, fluorescencefilters and beam splitters, collector lenses,contrast and heat filters.
Oil contamination
Clean (left) andoil contaminated (right)objective front lens.Toad, kidney,stained with Trichrome.Planapo 20/0.80. Bright field
Spherical Aberration
An imperfect image despite clean opticscaused by spherical aberration:Correction collar of the planapo 40/0.95 objective correctly (left) and incorrectly adjusted (right). Frog, small intestine,stained with Azan.
Dust on the optics
Dust on the internal optics (upper right),extremely dirty camera (bottom right)and clean optics (left).Frog, small intestine, stained with Azan.Planapo 10/0.45. Bright field.
When cleaning lens surfaces, avoid touching the lens surface with anything (even lens paper if possible).
IMPORTANT NOTE: Never use Kimwipes or commercial facial tissue, because they may contain a filler that is part Silica material (glass). One pass of a standard tissue could ruin an objective. Some objectives cost in order of €5K!
Remove dust by using a hand blower. Don’t use compressed air blower.
Remove water-soluble contamination using distilled water with a small amount of lens cleaning solution.
Lens Cleaning
Cleaning OilClean oil, other contaminants using proper optical micro lens cleaner.Pure Isopropanol , pure petroleum ether or Ted Pella Optical Lens Cleaner
• Optical Cleaning Solution L (from Carl Zeiss®)85% petroleum ether15% isopropanol
Remove most immersion oil by passing a high-quality lens tissue over the objective or condenser front element.Do not leave oil on lens when finished imagingWipe excess off when changing from objective to objectiveBe very very careful – slow down. Clean objective lenses by holding a piece of doubled lens paper over the objective and placing a few drops of solvent on the paper. Draw the paper across the lens surface so that the solvent flows rapidly in a circular pattern over the recessed lens surface. Finish the stroke with a dry portion of the paper. Repeat as necessary.Avoid soaking a lens with solvent, to prevent damage to lens cements.
Cleaning lensesTo clean recessed front elements of dry objective lenses or to remove stubborn dirt, use a cotton-tipped applicator that has been soaked in cleaning solution and then shaken to remove excess fluid. Rotate the cotton tip over the lens surface to clean.Again, first use the optical lens cleaner solution to remove the oil from the surface.Use a detergent solution or ethanol to clean the surfaces of the eyepiece lenses.Do not use xylene as it may solubilize enamel surfaces
Cleaning lenses
IMPORTANT NOTE: Do not rub the surface of the lens. No area on the tissue should come in contact with the lens twice. This prevents dust and dirt removed from the lens from coming back and possibly scratching it.
Wash hands – minimise oils/crud from fingers getting onto the lens surface.
Cleaning - Spiral motion
Cleaning is achieved using a spiral motion fromthe center to the rim. Never wipe using zig-zagmovements as this will only spread the dirt.
Cleaning - To recap1. When starting to clean, don’t forget to use a dust blower except when fluids (such as immersion oil) are to be removed.2. Never wipe lenses with dry swabs or tissue – this causes scratches!3. Do not use abrasive materials e.g. leather wipes, dry linen cloths or polystyrene sticks.4. Do not apply any solvents before trying distilled water (a film of distilled water can be generated by breathing on the surface), except when grease/oil is to be removed.5. Do not use any disposable cotton swabs ( e.g. Q-TipR) instead of cotton or ITW Texwipe CleanTipsR swabs, as the former are not free from contamination.6. Do not use any of the optical spray cans containing pressurized liquid air. The pressurized air from these sprays leaves a slight, but difficult to remove, residue.7. Never use acids or ammonia to clean objective front lenses.8. Never try to clean the internal optical surfaces, cameras or adaptor optics.
Preparation of Slides• Be careful with any mounting media – make sure its
set, do not get onto objectives, or use too soon• Make sure sealer (nail varnish) for cover slips has
set before using the microscope• Image in good time – don’t leave the slides too long.
Store in fridge. Keep away from room light. (for immunohistochemical stains)
Pitfalls in microscopy
• Bleaching , NA, filters, aberration• Experimental setup
– Fluorophore cross talk– Fluorophore saturation– Detector saturation
• Image acquisition– Offset– Binning, under, over sampling
• Software resolution issues
Fluorophore fading, bleaching, quenching
• Reduction of fluorescence intensity• All fluorochromes are subject to the
process of photo-bleaching, which is the irreversible chemical destruction which takes place during excitation.
• Living cells, too, may be damaged by the intense light.
• Very important: restrict the excitation brightness and duration to the exact amount needed.
Note: • Modify light with neutral density
filters or a motorised attenuator.• When the light is not needed for
excitation, close shutter.
TIME
False Data Interpretation
Mouse brain tissue slice 18 micron stack, 50 z-sections, 0.33 micron step40x 1.2NABlue Dapi nucleiGreen GFP dendritesRed Cy3 microtubule associated protein MAP2
Image A: xy merge at top of stack
Image B: xy merge at middle of stack
Image C: xz slice through the stack
GFP and DAPI extend through whole section Cy3 is restricted to top and bottom z sectionsConcl: Incomplete penetration of antibody
NA determines resolution
Diatoms AB, CDNA , not magnification determines the resolving power of the lens
100x UPlanApoWith adjustable NA
NA determines intensity
Intensity proportional to (NA)4
triple-labeled MDCK epithelial cell monolayer, 10x/0.25NA (Plan)
10×/0.45 NA objective (C-Apochromat)
High resolution images are necessary for colocalisation studies
10x
MDCK cells were labeled for tubulin (green; Alexa Fluor 488) and cytokeratins (red; Rhodamine Red-X)
100x
Is there colocalisation ?
High res image shows they are Non overlapping cytoplasmic networksLow res images can give false interpretation of colocalisation.
Emission filters must be selected to match your fluorochrome combination
Q: How many different fluorophores are present in this sample?
543-nm excitation and an HFP 488/543 main beamsplitter
560-nm long-pass filter
Actin filaments, nuclei, and cell–cell junctions all appear red
560–615 band-pass emission filter for the red channel and a 650-nm long-pass filter for the blue channel sequential scan
11-nm spectral bands across a total range of 552–723 nm, lambda scan, spectral unmixingTetramethyl Rhodamine (TRITC; labeling actin), Rhodamine Red-X (labeling desmosomes-cell junctions), and To-Pro3 (labeling nuclei)
The effect of chromatic aberration on images.
0.1 micron Tetraspeck beadsA: 100x/1.35NA, deltavision image restoration, minimal lateral shiftB: Z axis – 600 nm shift, chromatic distortionC,D corrective z-shift
E: nuclear structures alexafluor green, dapi, blue, xy mostly superimposedF: z azis not superimposedG: correction applied as calculated for beads
Chromatic aberrations can lead to incorrect conclusions.
The effect of spherical aberration on images
Mitotic spindle of HeLa cellsGreen: microtubulesRed: centrosomesBlue: Chromosomes 100x/1.35NADeconvolution deltavision
A-C: xy axis D-F: xz axis
Different oils of RI cause mismatch
Correct match is ?
Spectral Bleed/ cross talk
Multi-colour image of a GFP/YFP double-labelled sample . In this sample, GFP was fused with the H2B histone protein and YFP with tubulin.
Collect imagesfor different fluorophores sequentiallywhenever possible to avoid crosstalkbetween image channels when multiplefluorophores are excited simultaneously.
How to get rid/minimise bleed through?
Multi-colour image of a GFP/YFP double-labelled sample, after spectral unmixing . The visible result is a pronounced chromatic resolution of both structures now displayed in green and red. The tubulin on top of the bright green nuclei is even detectable.
Do spectral separation and linear unmixing Need to be able to separate contributions from each dye using linear simultaneous equations applied at each pixel of the image.
• Be careful and choose the best filters for the experiment
• Use sequential scanning
Also essential for colocalisation experiments
Fluorophore saturation5% 10% 30% 50%
5% 10% 30% 50%
More out of focus fluorophores excited Poorer z axis resIncreased photo bleaching / toxicity
START WITH LOW POWER LASER , HIGH PMT VOLTAGE , GRADUALLY INCREASE LASER POWER
All images look the same regardless of increase in input light
Offset – setting black levels 0.1 0.05 0 -0.05 -0.1 Corrected
0.1Offset
Background intensity decreases but info is lost at offset of 0 and lower values.Better to collect image with high background and correct post acquisition.
Background fluorescence decreases precision of fluorescence intensity measurements`Should be reduced
Detector saturation
Increasing Laser Power
If PMT gain, laser power, lamp intensities of camera exposure times are too high , detector can saturate causing a loss of structural information within bright structuresBest to have NO saturated pixels (red) – use HI-LO look up table (LUT).
Sampling frequency: undersamplingand oversampling
Binning
To increase the signal-to-noise ratio the camera pixels can be averaged or ‘binned’.i.e. 4x4 binning means:each block of 4x 4 pixels is averaged into one pixel four times largerBinned images are smaller, good for large data sets or live cell imaging, fasterBut for detailed work, binned images can be blurry because of undersampling
The number of pixels in the image must be sufficient to distinguish features of interest:
Under sampling1x1
0.108 mm/pixel2x2
0.216 mm/pixel3x3
0.324 mm/pixel4x4
0.432 mm/pixelCameraBinning
SameDisplaySettings
ApplyDifferentContrast
Brightness
Increased S/N due to camera binning
Images show similar detail at low magnification
Undersampling1x1
0.108 mm/pixel2x2
0.216 mm/pixel3x3
0.324 mm/pixel4x4
0.432 mm/pixel
Zoomed in part of the images above – at high mag, higher sampling is neededLooking at cell morphology is fine (first image set) but to resolve smaller cellular structures, 0.1-0.2 micron is required
Oversampling
Distance (m)
0.0 0.5 1.0 1.5 2.0
Inte
nsi
ty
0
200
400
600
800
1000
1200
1400
1600
1800Zoom 1Zoom 3Zoom 6Zoom 8Zoom 10
Zoom 10.140 mm/pixel
Zoom 30.047 mm/pixel
Zoom 60.023 mm/pixel
Zoom 80.017 mm/pixel
Zoom 100.014 mm/pixel
http://www.svi.nl/NyquistCalculator
Resolution improves from 0.14 to 0.047 micron pixelsBut not from 0.047
Confocal microscopes can sample with a very small pixel size but for most apps , a 0.1-0.2 micron pixel size is ok. Smaller pixel sizes mean bigger files but the resolution is limited by the props of light and the system optics- pixel sizes of <0.05 micron are similar
Software settings and Image display100 ms 400 ms 700 ms 1000 ms
Variable exposure times , same image display settings
Brightness and contrast adjusted to give similar display intensities
Changing image display settings can highlight dim features.
Lower exposure times or laser powers may still reveal what you want to see…
Image display - colour
Same image with variable colour displays :
3D intensity plot
Blue, red channels can be difficult to visualise; better to use greyscale images or pseudo colour LUTs to bring out dim features within images.
Image display - colour
Zhang, H. et al. J. Neurosci. 2005;25:3379-3388
If possible only usecolour in overlays.
Image display – gamma factor
brightness contrast All Threelinear gamma 0.5gamma 2
Image Intensity
Dis
play
Int
ensi
ty
Same image with varying display settings: our eyes process image non-linearly
Using the gamma correction brings out cell edges and dim features
Software tools• Volocity (here on campus via CMI)• Photoshop• ImageJ• Incarnations of Image J:• FIJI (is just image J….
Fiji is an image processing package, a distribution of ImageJ (and soon ImageJ2) together with Java, Java 3D and a lot of plugins organized into a coherent menu structure. Fiji compares to ImageJ as Ubuntu compares to Linux.
• Also …Mac Biophotonics Image J, another collection of useful plugins wrapped up into the main image J container.
Deconvolution
• No microscope is perfect, and some distortion is inevitable. • But it is possible to create a mathematical model of the
microscope’s optical characteristics using the instrument’s so-called “point spread function” (PSF), which describes the behavior of an infinitely small fluorescent point.
• Deconvolution uses this model to back-calculate the original appearance of the imaged field of view.
• Deconvolution thus cleans up a pre-existing image by digitally reconstructing it.
• With good deconvolution and z scanner, possible to generate 3d images with widefield
Convolution and deconvolution• Stray light from out of focus areas above or below the
focal plane causes glare, distortion and blurriness within the acquisition
• These image artefacts are collectively known as convolution
• Number of ways to counter this: :confocal , TIRFM• Another way: deconvolution• mathematical method for eliminating these image
artefacts after image acquisitionWe need to know the image PSF…But what is this?
What is the PSF ?
• The PSF is the image of a point source of light from the specimen projected by the microscope objective onto the intermediate image plane, i.e. the point spread function is represented by the Airy disk pattern
• PSF of an individual objective or a lens system depends on NA, objective, λillumination, contrast mode (e.g.brightfield, phase, DIC).
• Models spread/distortion/diffraction of light in 3 dimensions
PSFGlare , distortion , blurinessconvolution
deconvolution
After imaging and If we know the psf…
Stack of images:
What does the PSF look like?
500 nm excitation60 x oil lens1.4 NAn = 1.515
500 nm excitation10 x air lens0.4 NAn = 1
Volocity and deconvolution
• Can deconvolute images from widefield, confocal, spinning disk systems
• Create calculated psf using input data or/• Use measured PSF (from fluorescent beads)• Deconvolute from image• Easy to do and worth it!• Volocity available to all on campus – see
imaging.nuigalway.ie
Checklist for optimizing images for quantitation
Increase signal:
✓ Choose a bright (high quantum yield, high extinction coefficient) and photo-stable fluorophore
✓ Image through a clean No. 1.5 coverslip
✓ Mount specimen as close to the coverslip as possible
✓ Use high NA clean objective lens with lowest acceptable magnification
✓ Choose fluorescence filter sets that match fluorophore spectra
✓ Align lamp for Koehler illumination
✓ For fixed specimens, use a glycerol-based mounting medium containing anti-photobleaching inhibitors
✓ Remove DIC Wollaston prism and analyzer from light path
✓ Use a cooled CCD camera with at least 60% quantum efficiency
✓ Use camera binning
Checklist for optimizing images for quantitation
Decrease noise: ✓ Use a cooled CCD camera with less than 8 electrons readout noise and
negligible dark noise ✓ Use amplification (e.g., EM-CCDs) only when signal is limiting ✓ Increase signal (see above) to reduce relative contribution of Poisson noise
Decrease background: ✓ Clean coverslips and optics ✓ Optimise fluorophore labeling protocol to minimize nonspecific labeling ✓ Mount specimens in minimally fluorescent medium (e.g., without phenol
red) ✓ Use band-pass filter sets that block autofluorescence ✓ Turn off the room lights ✓ Close down the field diaphragm to illuminate only the object of interest ✓ When out-of-focus fluorescence is high, consider using deconvolution,
confocal, or TIRF
Checklist for optimizing images for quantitation
To recap , we talked about…
• What systems to use• Widefield vs confocal • Care of your microscope, how to clean• Sample prep• Practical tips for generating good images• Pitfalls when generating images• Deconvolution