​Drosophila and I: A Journal of My Time with Flies 

What is Fluorescence? 

 

When walking through life, we literally see how light interacts with matter by and large via scattering and reflection. Fluorescence is also a phenomenon of light-matter interaction; however it can seem a little bit more intricate.

Imagine a single photon (a “packet” of light with no mass) flying through space and then literally “hitting” something. This something is our fluorophore, a structure that fluoresces when getting hit by light of a certain quality. In biology, fluorophores are organic molecules with delocalized p-electron systems (yes, think aromatic systems like benzene, just a bit extended) that are able to absorb our photon. Upon the absorption event, two things are happening: 1.) Our photon is “unloading” its energy into the electron system of the fluorophore. It puts the electron system into an energy rich, excited state. 2.) Due to the crash, some of original energy that is brought to the table by our photon gets transformed into either heat, mechanical energy (the molecule either rotates or accelerates as a result), or both.

Now, the excited state of a fluorophore is rather short-lived, let’s say a couple dozens of nanoseconds (10^-9 s) only. After that time, the fluorophore emits a photon back into space and returns thereby to its ground state (it is not excited anymore). This emitted photon has a couple of properties that are worth keeping in mind: 1.) the direction in which it is getting emitted is pretty much random; it “forgot” where the original photon that excited our fluorophore came from. 2.) The emitted photon possesses lower energy than the photon that excited the fluorophore, because the heat and the rotational energy transferred to the fluorophore cannot be recaptured by the emitted photon.

A photon is a discrete energy packet. We perceive this energy as color, at least within a certain range of the energy spectrum. In a way, a rainbow can be described as a result of an energy-sorting machine of the photons of the sun light, as it displays white sun light as discrete colored components that can be assigned relative energies to each other. UV, followed by violet light have the highest energies, while red & far red have the lowest. We can use a second descriptor of colored light: Its wavelength. Blue light has a wavelength of approx. 450-490 nm, red light hovers around 660-700 nm. In summary, the shorter the wavelength, the higher the energy of a photon, the more the color is shifted towards the blue end of the visible spectrum. The opposite is true as well: The more red-shifted a color, the longer its wavelength, the lower its energy.

The fluorescence process depicted in a partial Jablonski Diagram (left), culled from the Huang Lab Website at UCSF. hu is a descriptor of a photon. The horizontal lines depict (idealized) orbital energies, the arrows follow the path of an electron through an absorption/emission cycle. To the right is an idealized absorption/emission spectrum of a hypothetical dye.

How does our absorption-excitation-emission event look like from an observer’s perspective? Let’s say our original photon is blue, with a wavelength of 480 nm. It get’s absorbed by a fluorophore, and after an excitation-emission cycle a photon of the wavelength 540 nm is emitted. We will perceive this photon as green colored light. We can also state that the emitted light is red-shifted by 60 nm (the wavelength of the emitted photon is 60 nm closer to red light than the excitatory photon).

 

The event cycle of absorption, excitation and emission with a red-shift is called fluorescence. It has its roots deep in quantum mechanics.

 

In real life, if a fluorophore is excited with a discrete wavelength (let’s say a 488 nm blue light laser), the emission follows a spectrum (a mixture of wavelengths within a given range and distribution determined by the physical properties of the fluorophore).

 

Further looking: http://huanglab.ucsf.edu/Lectures/2012%20UCSF%20Fluorescence%20dyes.pdf

Further reading: The Molecular Probes Handbook, online edition

 

What Makes a Good Fluorophore

 

 

For once, it should absorb as much light as possible (as every absorbed photon might give rise to an emission photon). The absorption property is quantified by the molar extinction coefficient e. The higher e the better.

The fluorophore needs to be photostable (able to absorb and emit as much photons as possible without getting bombed to pieces). A molecule with great absorption/emission properties but that gets burnt to a pile of steaming ashes during the process is not useful for the experimentator, and even harmful for a cell (phototoxicity).

The next parameter to consider is the quantum yield, F. It is the ratio of emitted to absorbed photons. By definition, this value cannot exceed 1, because of a thermodynamic law not permitting perpetuum mobiles. But a good fluorophor should have a F very close to 1.

The spectral properties of a good fluorophore should produce a decent red-shift. In other words, the wavelength of maximal absorption should be at least 40-50 nm away from the wavelength of maximal emission. The reason that this is a desirable property lays in the practical difficulty to distinguish excitatory light (of which there is a lot) and emitted light (of which there is very little). The bigger the wavelength differences of these two types of light are, the easier is it to distinguish between them and filter away the excitation light from the detector.

Lastly, a good fluorophore should have a narrow emission spectrum. In other words, the majority of emission wavelengths should be confined to a couple of dozend nm. Current experiments cram as many different fluorescent dyes into a single experiment as possible, making a narrow emission spectrum a prerequisite to avoid the “bleeding” of emissions emanating from one dye into the emission spectrum of another.

Now you know why it is difficult to make a good fluorescent dye!

 

Green Fluorescent Protein

 

 

We use GFP as a genetically encoded fluorescent probe. What does that mean? We want to visualize the expression pattern of the individual enhancer-traps. These are the cells that express the LexA::GH transcriptional activator. That implicitly states that the cells that do not express LexA::GH should give as little signal as possible. The way to do it is to cross a LexA::GH target encoding GFP (LexAop2-mCD8::EGFP to be precise, more about this later). Now all cells that express LexA::GH express, by extension, GFP. These cells will fluoresce green when excited with blue light, while the rest of the cells show only background signal caused by autofluorecence.

            What might get lost here is the convenience of GFP. By a mere cross we can label a subset of cells, while the others remain dark. We can do that because GFP is encoded by a gene that gets transcribed and translated by the cellular machinery present in the cell. No dance involving permeabilization, fixation, antibody labeling or introduction of other fluorescent chemical probes required.

      GFP itself was derived from a jellyfish and has been actively researched for the past two decade or so. The work has been awarded a Nobel Price for chemistry to Chalfie, Shimomura and Tsien in 2008. The primary amino acid chain of GFP forms a neat tertiary structure called a beta-barrel encasing the chromophor (the part of GFP that does the biophysical labor of fluorescence) that is derived out of amino acids of the primary sequence, making it truly unique.

The GFP construct used here has a couple of features that make it different from the original jellyfish GFP. For once, the GFP sequence has been modified so that its spectral properties (higher e and n, as well as increased red-shift of emitted light and a narrower emission spectrum) are improved for laboratory use. This puts the “E” (enhanced) in front of the GFP. It makes EGFP brighter and easier to image. Furthermore, EGFP is fused to a “tag” (mCD8, a short stretch of amino acids from the mouse CD8 molecule) that localizes EGFP to cell membranes. Why is this an improvement? Take for example an axon of a neuronal cell. The diameter of the axon is tiny, with very little cytoplasm in its volume. Even if GFP gets somehow in there, there would not be enough GFP around to be noticed and these cell extensions would be poorly detectable or even remain invisible. The membrane tag assures that these sub-cellular structures become visible.

 

Further reading on chromophore: http://www.cryst.bbk.ac.uk/PPS2/projects/jonda/chromoph.htm

 

Epifluorescence Microscopy

 

Normal” microscopy uses a white light source that interacts with a biological specimen via reflection, absorption and scattering. In some cases it uses more complex interaction forms like phase contrast. What all these forms of microscopy have in common is that (at an upright microscope) white light shines from the bottom up, through the object. The image gets captured by a high magnification lens system and gets funneled into a recording system, a camera or your eye. In essence, a microscope is acts as a big (and somewhat expensive) magnifying glass.

Epifluorescent microscopy integrates the principles of fluorescence outlined above into microscopy. The “Epi” stands for the fact that the excitation light & the observation of the emission are taking place on the same side of the object that is being imaged. It is just one way to build a microscope system. Doing fluorescent microscopy requires extra equipment a standard light microscope has not. They all handle the extra light requirement.

 

1. A fluorescent light source. In most systems, this involves some Mercury [Hg] arc lamp bulb. This is necessary to produce the required light intensities for fluorescent microscopy.

2. Excitation filter. The light from the Hg arc bulb is filtered by a so called bandpass filter. In the ideal case, it blocks all light, except of certain wavelengths. For example, for EGFP, the filter specification is 470/20. That means that all light of 470nm +/- 20nm (this is blue light) can pass through, the rest cannot. This restriction is necessary, as unfiltered Hg lamp light would cause a massive onslaught of autofluorescence in which the signal would be lost or greatly diminished. Excitation filters are indispensable especially if multiple dyes are present in the sample that is being imaged. Only one dye at a time should get excited, not all the others. The other excited dyes might “bleed through” into the emission signal of the dye that is targeted for observation.

3. Dichroic mirror, also called a beam splitter. This is a funky semi-transparent mirror that provides a cutoff system. It reflects all light up to a certain wavelength, and is transparent for light of longer wavelengths than the cutoff. It allows for the observation point (eye or camera) to be in a different path than the Hg light bulb.

4. Emission filter. They come in two different flavors: Bandpass, similar to the excitation filter, but its permissive window is red-shifted relative to the excitation filter. The second type is a longpass filter, that lets all light from a certain wavelength on through. Why do we need an emission filter? Because the brightness of the reflected excitation  light is so much brighter than the brightness of the emission light. The emission signal would be drowned out by the reflected light, like, totally

5. A CCD camera as recording device. These are frequently black/white cameras, because they are more sensitive than color ones. This comes at a certain cost. For the camera, all photons register as grey values. The color information of the emission light of each specific dye is lost, and is re-introduced later by writing the gray values into the separate color channels of a RGB file. However, the person doing the imaging needs to keep track what filter setting are connected to what color channels of the imaging file. Otherwise the information for EGFP will end up in the red channel and cause confusion.

 

The excitation and emission filter, as well as the dichroic mirror are grouped as a filter cube. Each fluorophor has its own filter cube with the correct set of ex/em filters and dichroic mirror. These filter cubes are arranged in a wheel or slider to allow fast exchange.