Fluorescent Indicators. Martin Thomas, Cairn Research Ltd

Transcription

1 Fluorescent Indicators Martin Thomas, Cairn Research Ltd

2 ptical Measurements Are Sensitive! Electric current 1A = 6.25 x electrons/sec Squid axon voltage clamp 1mA charges/sec Microelectrode voltage clamp 10uA charges/sec, down to approx 1nA charges/sec Patch clamp 1pA 10 7 charges/sec ptical Recording Photomultipliers and electron-multiplying CCD cameras can detect single photons, but we need a lot to get good images - tradeoff between spatial and time resolution

3 ptical Units Radiant Power Watt (W) Lumen (lm) Radiant Intensity Watts / Steradian (W/sr) Lumens / Steradian (candela) Illumination Intensity Lux (lm m -2 ) Footcandle (lm ft -2 ) At 555nm: 1W 680 lm 2.8x10 18 photons sec -1

4 Fluorescence Fluorescence is is the absorption of light by a molecule, to form a short-lived excited state, followed by re-emission of light at a longer wavelength. A crib for becoming an instant expert: FLIM Fluorescence Lifetime IMaging FRET Fluorescence Resonant Energy Transfer FISH Fluorescence In Situ Hybridisation TIRF Total Internal Reflection Fluorescence FRAP Fluorescence Recovery After Photobleaching

5 Fluorescence Fluorescence is is the absorption of light by a molecule, to form a short-lived excited state, followed by re-emission of light at a longer wavelength. A crib for becoming an instant expert: FLIM Fluorescence Lifetime IMaging FRET Fluorescence Resonant Energy Transfer FISH Fluorescence In Situ Hybridisation TIRF Total Internal Reflection Fluorescence FRAP Fluorescence Recovery After Photobleaching FLAF Four Letter Acronym Fluorescence

6 - C CŌ CŌ CŌ CŌ - C - C - C - C CŌ CŌ - C CŌ CŌ - C EGTA indo-1 fura-2 fluo-3 BAPTA - BCECF ph indicator - C - C H CŌ a + indicator: CŌ CŌ - C CŌ SBFI Indicator structures Ca 2+ buffers and indicators CŌ - C - C S Donor Cl H CŌ CŌ S CŌ - C - C Cl - CCF2: gene expression reporter Flox-6: voltage sensor S Accept or Lectin Fl CŌ (stolen from Roger Tsien)

7 Crystal structure of S65T GFP

8 Ca Indicator Spectra Fura2 EX spectra measured at 510nm EM, EM spectra measured at 340nm EX Ca bound Ca free Indo 1 Absorption (=EX?) spectra, EM spectra measured at 338nm EX Ca bound Ca free EX spectra dotted, EM spectra continuous, data from Invitro

9 SBFI a and ph Indicator Spectra EX spectra measured at 505nm EM, EM spectra measured at 340nm EX a bound a free BCECF EX spectra measured at 535nm EM, EM spectra measured at 480nm EX ph 5.2 ph 9.0 EX spectra dotted, EM spectra continuous, data from Invitrogen website

10 Fura2 excitation spectrum

11 Indo1 emission spectrum

12 Fura8 from AAT-Bioquest Excitation spectrum shifted to longer wavelengths So no need for UV-transmitting optics But seems more susceptible to photobleaching than Fura2

13 But the affinity of most indicators is too high to record transients properly! Slow Timescale Fast Timescale Endothelial cells plus ATP Hepatocytes plus flash-released IP3 gden et al (1995), Eur J Physiol 429,

14 Camelions

15 Yellow Cameleon Emission Spectra

16 Fluorescence Quantification Ca + I FREE = CaI (at equilibrium) K D = [Ca][I FREE ]/[CaI], so [Ca] = K D [CaI]/[I FREE ], or [Ca] = K D [CaI]/([I TTAL ] - [CaI]) If only CaI is fluorescent, then we can write: [Ca] = K D F/(F MAX - F) where F MAX is the fluorescence at saturating [Ca], i.e. [CaI] = [I TTAL ] If I FREE is also fluorescent, the relation becomes: [Ca] = K D (F - F MI )/(F MAX - F) where F MI is the fluorescence at zero [Ca], i.e. [I FREE ] = [I TTAL ] A similar derivation for ratiometric quantification gives: [Ca] = K D x C(R - R MI )/(R MAX - R) where C is is the relative change in fluorescence between zero and saturating [Ca] at the denominator wavelength of the ratio (see Grynkiewicz, Poenie and Tsien, 1985, J. Biol. Chem 260 p3440 for full derivation) For both quantification methods, BEWARE of buffering and saturation effects, which can significantly alter the amplitude and time course of the recorded signals.

17 Ratiometric Measurement There must be a change in the shape (not just the amplitude) of the spectrum, so not possible for all indicators Corrects for changes in indicator concentration and excitation light intensity Easier quantification, as R MAX will be a known constant on a given experimental setup, whereas for nonratio indicators F MAX needs to be determined for each experiment Can utilise changes in either excitation spectrum or emission spectrum Detection of changes in emission spectrum is somewhat easier (use two detectors) Detection of of changes in excitation spectrum requires sequential measurements at different wavelengths (e.g. use a monochromator) A versatile measurement system should be able to handle both methods May be possible with either method to use more than one indicator at the same time (if their spectra are sufficiently different)

18 Ratio change on K+ perfusion: movement artefacts on the 340nm and 380nm traces don t show up on the ratio (vertical location of 340 and 380nm traces is arbitrary)

19 Loading Indicators Into Cells Indicator needs to be membrane-impermeant in order to remain in the cell, but how do we put it there? Microinjection can work well for large cells, but increasingly difficult for smaller cells. Shotgun approach using coated gold beads can also work well (e.g. for tissue slices). ther membrane disruptive techniques such as electroporation may also be possible. Can introduce polymeric (dextran linked) forms for cells that tend to expel small foreign molecules (e.g. plants). Ester loading can work well for smaller cells; esterified forms are membrane permeant, and so can enter cell, then are hydrolysed to the active free (membraneimpermeant) forms by internal esterases. But internal compartments other than cytosol may also be loaded, depending on location of the esterases. Genetic expression of a protein-based indicator is potentially very attractive, as expression can be targeted to specific cell compartments and the cells come readyloaded. But problems here include time and effort to make the constructs, and generally more limited signal range than for chemically synthesised indicators. In ALL CASES the presence of the indicator itself (e.g. buffering effects), as well as the means of its introduction, may disturb what you are trying to measure, so do remember this when planning experiments (e.g. use several different concentrations).

20 A fluorescence microscope (lympus BX series)

21 Filters and Dichroic Mirrors Transmission spectra for FM-143 filter set Specialist coatings permit filters with almost any optical profile to be created

22 High pressure Arc Lamps Arc Profile: Luminous intensity variation along lamp axis 75W lamp < 1mm Luminous Intensity distribution curve

23 Arc Lamp Emission Spectra Xenon Arc : Mercury Arc :

24 High Intensity LEDs Blue wavelengths now especially powerful, but more power in the green would still be useful. Available wavelengths now down to 340nm (Fura2 excitation!) Lower spectrum is of a white lamp actually blue with a phosphor.

25 Diffraction Grating Diffractive maxima occur at angles where the path length differences are an integral number of wavelengths. Thus longer wavelengths are diffracted at greater angles.

26 Monochromator peration Image of input slit formed at exit slit Triangular bandpass profile ptics must be matched to rest of optical system to maximise throughput

27 Detection Systems PHTMULTIPLIERS: Cheap, fast, sensitive, can use adjustable diaphragm to determine field of view, can easily record data in same acquisition system as e.g. electrical data. Why not buy one or two to use alongside your expensive CCD camera? PHTDIDES: More background noise than photomultipliers, so not suitable for weak signals. Avalanche photodiodes (APDs) are better in this respect, but still inferior. However, photodiodes do have a higher quantum (i.e. detection) efficiency, so may be better for strong signals, where photon noise dominates. CCD CAMERAS: More expensive, generally slower and more limited signal range, but give nice pictures for journal covers. 3D deconvolution in software can give reasonable depth resolution, but you need to record a vertical stack of images in order to to this. Cameras and software now much improved over earlier systems. LASER SCAIG CFCAL: Unlike the above methods, out-of-focus light is rejected, allowing true vertical sectioning. Potentially much more detail, but tradeoff is longer acquisition time for a full image, although e.g. linescan mode can be fast. Range of usable dyes restricted by availability/cost of appropriate lasers. Two-photon systems potentially even better, but at a price. IPKW DISC CFCAL: Scans multiple points on the specimen simultaneously, and images onto a CCD camera. ice idea in principle, also gives nice pictures, but depth resolution is poorer. ne problem seems to be that the optimum hole pattern in the spinning disc depends on the microscope objective, so a compromise here.

28 Dotted lines show how response can be extended into the UV if the device has a quartz instead of glass window. Photodetector QE

29 ipkow Disc Microscope

30 The Cairnfocal DMD-Based Confocal Microscope Attaches to the sideport of a fluorescence microscope The nonconjugate side images from the DMD's off pixels Light from DMD collected at 24 degrees focusses normally onto the cameras 62 is the DMD 66 and 66' are the conjugate and nonconjugate cameras 86 and 86' are the excitation light sources (generally just use 86)

31 The Cairnfocal DMD-Based Confocal Microscope

32 Cairnfocal Data! From Ashley Cadby Lab, Sheffield

33 oise on ptical Signals Detector will contribute noise, but this will be a constant amount, although it can vary widely between different types of detector. ptical signals are inherently noisy because of statistical variation in arrival times of individual photons this is termed shot noise. Shot noise increases with the square root of the signal level. Therefore stronger signals are also noisier, and the signal-to-noise ratio increases only with the square root of the signal level, rather than linearly as one might have expected. But there may be considerable scope for increasing the optical signal, as collection efficiency increases with the square of the numerical aperture (A) of the system. Can also increase illumination intensity of course, but there are practical limits (e.g. photobleaching). Since noise sources are uncorrelated, the overall noise level is the square root of the sum of the squares of the individual components, so the total noise level is determined primarily by the largest one (more so than if they added linearly). For weak signals, detector noise may dominate, so low noise here is important, e.g. photomultipliers, image intensifiers. For stronger signals, shot noise is more likely to dominate, favouring use of more efficient (higher QE) detectors, e.g. avalanche photodiodes, back-thinned CCD chips. Latest CCD cameras with on chip multiplication may offer high QE AD low detector noise, but the statistical nature of the multiplication is expected to increase the effective shot noise by 2, so even this is not a perfect solution!