So much of the microscopy that we are doing nowadays is fluorescence microscopy and I'll be explaining what that is in this lecture. So why do we use fluorescence in the microscope? As you probably seen so far, there are two important aspects in making a great picture in your microscope. One is resolution, you want high resolution, and the second thing is contrast, you want high contrast. Otherwise you wouldn't see anything. And what fluorescence offers is this high contrast. So, it's demonstrated in this image here. We have part of the cell labeled in green, in green, we are looking at microtubules, so we label microtubules with the fluorescent dye and we have part of the cell labeled in red, and that part is actually the uh, nucleus. It's DNA, it's a red label attached to the histone protein and what you see here is that these labels stand out on a black background. So we have a very high contrast, we see what we want to look at whereas our background is totally absent.

so it gives an actually extremely high contrast. And what we already see in this specific image is that you get specificity. We can attach these fluorescent labels to specific proteins, to specific molecules, to specific dyes, to specific places in the cell or wherever we want them, so that we can target them, that they show what we want to see. Then fluorescence is quantitative, so if I calibrate my system correctly, areas where we have higher signal, like here, those are areas where I have more of my the protein that I am interested in. So in this case, there is a lot more tubulin here in this bright area than here in these dimmer areas. And then lastly, and very importantly, fluorescence is compatible with imaging of living cells. So this image here is not only a still image, but it's actually part of a sequence of images that I took. And when we now start playing those faster, we see that it's actually a dividing cell. The microtubules assemble in this spindle.

and the spindle then pulls apart the chromosomes into daughter nuclei. So fluorescence is a great thing, a great thing for microscopy. But what is it really? And what's kind of surprising to me still, is that humankind has been aware of fluorescence for only about a hundred and fifty years. So one of the first descriptions of fluorescence was by Sir William Herschel. He was studying the aspect of an extract of the bark of the cinchona tree. And he held up this extract in the light of the sun that came through his window, and he noted that the light of the sun induced a celestial blue light coming out of that extract of the bark of the cinchona tree. I will now try to repeat that one hundred and fifty year old experiment here in the lab, and I'll be showing you that now. So the bark of the cinchona tree is hard to come by, but luckily the active ingredient can be found in every supermarket in form of contained in tonic water. I'll take the tonic water, pour it in a glass.

and the other part is that we need UV light to illuminate it, and that we can buy in a form of what's called a urine tracker in a pet store. So, when we now shine this UV light in the quinine, in tonic water, you see that you get a blue light emanating from within the tonic water. It's another wavelength, it's a longer wavelength, than this light with which I illuminate, and that is why we see it a lot better. Our eyes are not very sensitive to this UV, but they are highly sensitive to the blue light in tonic water. Quinine is not the only compound that fluoresces, of course, otherwise it would be quite uninteresting. Already in the nineteenth century, scientists synthesized this dye called fluorescein that is highly fluorescent. And when I shine blue light in that, we get yellow fluorescence coming out. Another very well-known dye is rhodamine, and when you shine green light in that, you get this orange-yellowish fluorescence coming out. And so this is fluorescence.

okay so what we have seen is that we use light of a lower wavelength. So in this case of the quinine from the extract of the bark of the cinchona tree, we use UV light, and then the dye responds to that light by emitting light of higher wavelength. So in this case, blue light. So in general, except for very special cases, the emission light is of a longer wavelength, which means lower energy. So there is some energy loss going on in this process than the excitation light. This aspect of fluorescence is very important, and we often show that by measuring the spectra. So we measure the excitation spectrum, which is often very similar to an absorption spectrum, and we measure an emission spectrum. So here that is this curve here, which is the fluorescent light emitted by the dye. So we characterize these dyes by their excitation maximum and by their emission maximum. Those numbers tell you immediately a lot about how to use the dye in your actual experiments.

the difference between the excitation and emission maximum is called the Stokes Shift, after Sir Stokes who first realized that there was this difference. You can have dyes that have a small Stokes Shift, where the emission is very close to the excitation, or you have dyes with a very large Stokes Shift, where the emission is very far away from the excitation light. Now, so there is energy loss going on, and this causes the difference in wavelength. But what, how, does that happen? And one of the most useful ways of thinking about that is with the so-called Jablonski diagrams. Jablonski was a Polish scientist who did this work in the beginning of the twentieth century. And in these Jablonski diagrams, we are looking at the orbitals that electrons can be in when a dye observes light. The electron is going to be kicked to a higher orbital, in a higher energy state. This absorption goes very, very fast, so it happens in a time scale of femtoseconds.

and femtoseconds to me don't mean much, so I like to express this in the distance that light travels during that time period. This is like those extremely fast. It's pretty amazing that light travels only like 0.3 microns during this time period. Once that electron finds itself into these higher orbitals through a process called internal conversion, it quickly falls back to what you call ground level of the excited state. And that internal conversion is a tiny little bit of heat that dissipates this way, so we have a little bit of energy loss here, and that happens on the level of time scales of picoseconds, and light will travel only 0.3 mm during that time period. Then the electron can hang out here for what is in comparison to this previous time scales almost eternity, something on the order of many nanoseconds, then it will fall back into the real ground state and emit a photon. That photon will be of a wavelength less than the wavelength, sorry, of an energy that is less

than the energy that was put in initially, and that means that the photon that is coming out is going to be of a higher wavelength. Now, not only this return to the ground state is not the only process that can take place when a dye is in an excited state. It's also possible that the energy is transferred to other molecules, or even the electron can fall back to the ground state but not emit a photon. So there is also non-radiative decay. So we have a total number of events, some of which result in light coming out, and some of which don't. And we express the ratio of those two as the quantum efficiency. And the quantum efficiency is a very important parameter of the dye. It not only depends on the dye itself, it also depends on its environment. For instance, what other molecules are around the dye. But nevertheless, quantum efficiency is something you want to know, and that you want to measure, or that you want to be told what it is. Another important aspect is the brightness.

and the brightness in the end is determined by how well the dye absorbs the excitation light and what fraction of the excited state returns to the ground state on the emission of fluorescence. So it's determined by both the absorption coefficient and the quantum efficiency. Now, another important aspect here is what we already talked about: how long does it take between absorbing a dye and for the dye to return to the ground state. And that is what we call the life time. It is mainly determined by how long it takes before we return to the ground state out of the excited state. Usually, this is on the order of nanoseconds. Nevertheless, it's also a parameter that is dye specific, that is also dependent on the environment of the dyes. And there's a whole form of fluorescence microscopy called LifeTime microscopy that actually reads out LifeTime instead of only reading out fluorescence itself. Nevertheless, what we are now going to talk about is microscopes that discriminate between the excitation light and the emission light.

so how do we do that in a microscope? Well, the main thing is that we use filters. We use filters that filter the exciting light, so that we get out of our light source only the light that is of the wavelength that will excite our dye. That light will bounce off of the dichroic mirror, and then excite our sample. After it went through the objective, the sample will start fluorescing, and that fluorescence travels back through the same objective and will now pass through the dichroic mirror. And then there is an emission filter that will let through fluorescence, but that will completely block any scattered excitation light so that we only see the fluorescence on our detector, which I have symbolized here with an eye, but that can also be a CCD camera, or what not. So this is the simple, basic principle of a fluorescent microscope. So the most important aspect here, and the thing that makes it different from a normal microscope, is filters, and these filters need to be very good.

becuase they really need to discriminate this excitation light from the emitted light so what options do we have there? Well, you can have filters that are based on absorption that absorb certain wavelengths of light. Those are not that great in practice because their spectra are very broad, so they will also absorb a bit of the emission light. So, they will not do exactly what we want them to do. The type of filters that we mainly use these days are interference filters. So what are those? Interference filters are basically thin alternating layers that have a different refractive index so that we get interfaces there that can reflect the light. So you can almost think about them as thin transparent layers that are about the wavelength of the light or half the wavelength of the light with semi-reflective coatings in between. So what now happens when light hits this filter is that part of it will be reflected, reflected, part of it will move through. The light that reflects will interfere with itself.

and when the wavelength is such that it matches the wavelength of the light, you will get constructed interference, and that constructive interference will then, in effect, let the light pass through this filter. Whereas, when the wavelengths are different from the two, um, the distance between the layers, we get destructive interference, and it will be reflected out. The filter manufacturers can completely compute how these layers will behave, so they can design in the computer what wavelengths this filter should pass through and what wavelengths should be reflected, and you can basically order any type of filter that you may possibly need. Now, just to make this a little bit more clear, there is one example of interference that you are highly aware of, and that everyone of us has seen, and that is what you see in cell bubbles. So in cell bubbles, when light hits that, you will see reflection of all kinds of color of light. The cell bubble itself has a thickness that is on the order of half the wavelength of a lot of the visible light.

and you will get the same interference effect, where some of the colors will pass through and other colors will be reflected, and that gives this beautiful color patterns. Now we will break and show you some of these filters, and will also show you now here how these filters are assembled in a filter cube. Okay, so I have an example here of such an interference filter. This one is called HQ47040, and 470 is the center wavelength of this filter, and the 40 means bandwidth, so this one will pass blue light. So if I have my blue LED and I hold that up, I see the blue light coming through, and you will see that this LED indeed passes it through. Now with a green light source, and hold that up, and go through this filter, you will see that it gets a lot dimmer, and not only that, you may also notice that it's actually a little bit of blue coming through, and that is because this green LED is not purely green.

but also has a slight bit of blue light in it. Then, the next part is this dichroic mirror. So, this one is designed, again, such that it reflects blue light. So, you will not see coming anything through, but if I reflect it to you, you may be able to see that. And then the last part is this emission filter that will transmit the green light and reject the blue light. So, in effect, when we would now place these three in this configuration in the microscope, illuminate from this side, go up, we have the objective in the sample here, come back through, and go through this emission filter. That's the principle of a fluorescence microscope. These filters, this is exactly the way they are arranged in these cubes that were designed by Ploem, where we have an excitation filter, a dichroic mirror, and then an emission filter here. And so, these will, as expected, reflect, and I'll now turning it to you.

so that you can see it, so that the exciting blue light will be reflected to the specimen, the specimen will fluoresce green, and that will make it to the eyepiece or to the camera sensor. So we've seen now these filter cubes. There's an excitation filter, there's a dichroic mirror, and there's an emission filter. What's very important to know, and the information you should always have about these filters, is their spectra. So, what are the wavelengths that are being transmitted, and what are the wavelengths that are being reflected, and those spectra then will look something like as follows. So we have an excitation filter that passes through this band-pass, these wavelengths, these colors of light. The dichroic mirror is made such that it will reflect everything that is lower than a certain cutoff band and pass through everything that is higher, and then the emission filter will pass through the higher wavelengths.

but reject absolutely reject. The excitation light. So, when you now assemble a fluorescent microscope, your task will be to select filters that match well the dyes that you'll be working with. And what you will be doing in that case is to try to overlay the spectrum of the fluorophores with the spectrum of your filters. So, I have here a spectrum of a dye that is probably Cy3, the Cyanine dye. It excites in the green areas, around five-fifty, and it has an emission maximum around six hundred. So, if you use a band-pass filter that is centered around six hundred and fifty nanometers, that will easily reject the excitation light because it's far away from that. But you also note that the integral of it, that it only passes through a fraction of the emitted fluorescence. And that means that you are losing something like seventy-five percent or so of the emitted fluorescence. And in general, you really want to get all the photons that you can, so it's not a perfect situation.

so if you now switch to a band-pass filter that is slightly lower, that is around, centered around six-hundred and ten nanometers, then your catching a much larger fraction, like seventy-five or eighty percent of the emitted photons, and you will be much happier because your image will be much brighter. Now, we'll now switch to the lab and show how we are using these filter cubes to collect actual images. Here we have a real fluorescence microscope, so we have a light source back here. This is a mercury arch lab light sources about that. That light then passes through the main body, and then hits here, the dichroic mirror. So, and I'll show you now how this works. So, this is basically a carousel that you can take out, and this carousel contains, or turret, contains these same filter cubes that we have been looking at in the lecture. So, this one is sitting here, the light hits the excitation filter. I can take these cubes out like this.

so the light hits the excitation filter. It's sitting in the microscope like this. It goes through the excitation filter, hits the mirror, goes up. The fluorescence light comes back, makes its way through the mirror, and then here at the bottom, we have the emission filter that goes down and in the end, so the camera or the eyepiece. And now, since this thing is motorized, once it's sitting in the microscope, it can, with the push of the button, put in another one of these filter cubes. And these cubes have been selected to visualize the dyes that we are interested in. And so usually there's one for things like DAPI, for things like GFP, or fluorescein, and a cube for something like mCherry or rhodamine, okay. So when I now put this back into the microscope and then open the shutter here in the back that blocks the light at the moment, we get blue light coming out of the objective, and with a push of the button.

I put now another dichroic in place. Here we have green light for rhodamine or mCherry, and if I go on, we get red light for something like a dye like Cyfive and so. Just by switching these, we can select the wavelength of the light for excitation and the filter used for emission and select the fluorophore that we want to image. So we've now seen how we can use the fluorescent microscope to collect individual channels, images of individual fluorophores. Then what we can do in the computer is put all those images together, give them a nice color and end up with things like this. So here we have a staining of mitochondria in red, of DNA in blue, and what looks like actin stress fibers in green. This works very, very well. There are also alternative strategies to doing this, and one is what you may have noticed in the lab part is that it actually takes quite a bit of time to move that turret.

so going from one position to the other, it takes, you know, quickly like half a second, or so. The turret is pretty big, so it has a lot of vibrations. If you want to go faster, the whole thing starts moving more. Those are all bad things, and things you don't really want. So instead of using these filter turrets, what you can do is put your excitation, and possibly also your emission filters, on a filter wheel. Now, I brought such a filter wheel, and I can show that to you here. So it really is a wheel that has individual filters in them. There is a motor that can, through this rubber band here, move that around. And we can put, by activating the motor, we can switch between these different filters. And we can do that at a speed that generally is much higher than you can do on a turret. So we then first take an image where we put the first excitation filter in place. So for instance, in this case.

we have a blue excitation filter. We get blue light out. We get green fluorescence. And when we move the filter wheel to the next position, which has a yellow filter, we get red fluorescence coming out. So we collect now the second image, and that goes much, much faster. Now you may have noticed that there is something funny going on because this dichroic filter here, apparently, reflects both the blue light and the yellow light, but it passes through the green light. And so that means that the dichroic filter needs to have pretty special properties, and it's very different from what I've shown you so far. And indeed that is the case, and luckily the filter manufacturers are able to make such interference filters so they can make filters that reflect at certain wavelengths and then pass through the light in between those wavelengths. And some of the spectra of filters you can buy look a lot better than what I'm showing you here. It's literally amazing what the filter

manufacturers have accomplished and that has been a also a great driving force behind the whole progress in fluorescence microscopy. So, we had our image of our red dye, um, and then we put them all back together in the computer in these multi-colored images. Now, also that emission filter, we could, instead of having an emission filter, we could, which, would have to be a multi-band pass emission filter, we could replace it by a second filter wheel, and then we have a full control and very fast control of both the excitation and emission filters. One practical aspect that you need to be aware of is photobleaching. So, fluorescence dyes don't last forever, they only have so many cycles that they can go through and over time, their fluorescence tends to bleach out. So for instance, in this movie that I'm showing you here, um, where I've coupled GFP to the microtubule tip tracking protein EBI we saw that at the beginning.

we had bright fluorescence, and then over time, it completely disappears. Now, this is regretfully a common aspect of most dyes, and that is something that we usually really don't like and that we want to avoid. So, what can you do about it? First of all, you can select dyes that are resistant against fade, or do not bleach as fast as others. So, for instance, this dye, fluorescein, that I showed to you before, tends to bleach extremely, extremely fast. There are now other dyes on the market that have the same excitation and emission properties but that bleach much, much slower, so you would almost never want to use fluorescein anymore, but use one of these replacements. You can try to label more densely, so each dye has a certain chance that it will photobleach, so if we now put a lot more labels there, it will take much longer before my signal is gone. Then, very importantly, you want to

change the environment of the dye so that bleaching decreases. So, for instance, oxygen is important in this photobleaching process. So, a common strategy is to remove molecular oxygen from your sample as much as possible, for instance by using glycerol. You're already doing a lot better, or by using enzymatic systems that take out the oxygen from the sample. There are also a lot of compounds that you can add to your mounting medium that will reduce bleaching. And lastly, and very importantly, you really want to use all of the photons that you can get out of your dye. So, you don't want to leave on that excitation lamp when you're not looking, when you're not observing your sample. So, always close the shutter immediately when you're not looking, or when you're doing camera-based imaging, you want to take the shortest exposure times possible, using the lowest amount of excitation light that you can, and you also want to synchronize your camera with the shutter as tightly as possible.

so that there is no dead time that you're not actually collecting that fluorescent light. Now, I'd like to finish this presentation, going back to that movie that we saw in the beginning. You know, in the end, you're using fluorescence simply because it's a great tool to let us look at cells and let us look at individual aspects of cells. To me, it's still just amazing that we can see these living cells and components in these living cells while they are living. And you know, the beauty of these kind of movies and images is just utterly stunning, and it makes it a joy to actually work with fluorescence microscopes.

Microscopy: Introduction to Fluorescence Microscopy (Nico Stuurman)

Free biology talks by the world's leading scientists that show the process of science and the wonders of discovery. Please visit our other YouTube channels for more free videos:

iBiology: Seminars by the world's leading scientists 
iBiology Science Stories: Stories of discovery and the series "How I became a scientist"
iBiology Career and Professional Development: Advice and training videos on how to navigate a career in science and beyond
Wonder Collaborative: A new initiative aimed to bring the stories of science to the general public


View our website at https://www.ibiology.org, and join our newsletter here: https://www.ibiology.org/join.html