Fluorescent Proteins and the Story Behind GFP

Top

00:00:11.02 Hello, I'm Roger Tsien.
00:00:13.01 And I'm here today to tell you about fluorescent proteins.
00:00:16.14 which have made a major impact on microscopy
00:00:19.26 because they are genetically codable and
00:00:23.00 provide a relatively direct link from molecular and cell biology
00:00:26.04 into colors that we can directly see
00:00:29.17 particularly in the light microscope
00:00:31.12 but also, as you will see later on,
00:00:33.28 more macroscopic levels as well as eventually now
00:00:37.17 beginning to help us with super-resolution.
00:00:41.26 So, here this picture is, of course,
00:00:44.24 of the jellyfish, and
00:00:45.28 this jellyfish is where it all began.
00:00:48.13 And in a way, this is the creature that we most have to thank.
00:00:52.13 This is the jellyfish Aequorea victoria.
00:00:55.17 which was studied in Puget Sound
00:00:59.03 by Professor Shimomura, and in a moment,
00:01:01.24 we'll learn a little bit more about him soon.
00:01:02.27 This is the source of two of the most actually valuable proteins
00:01:07.06 in cell biology
00:01:08.21 First, there is the protein that actually enables this jellyfish to glow
00:01:12.23 And that's called aequorin, and
00:01:15.12 when the jellyfish is alarmed in the water,
00:01:18.11 and the water is disturbed, it emits a glow.
00:01:20.19 And then the partner of aequorin is the
00:01:24.00 green fluorescent protein, which changes the color from blue to green
00:01:28.01 to this day, we really don't have a well-accepted explanation
00:01:32.14 as to why the jellyfish wants to glow
00:01:35.21 why should it show this remarkable phenomenon
00:01:40.24 when the water's disturbed
00:01:42.29 nor do we know why when the jellyfish glows
00:01:45.28 why was it so important that it glows green instead of blue
00:01:49.21 why not just leave aequorin, or
00:01:52.10 if it really was important to glow green
00:01:55.07 why not just change aequorin in the first place to make it glow green directly
00:01:59.00 rather than invent an additional protein
00:02:01.23 but we are very grateful to the jellyfish
00:02:04.05 for having invented GFP.
00:02:06.27 So, the example - this is perhaps one of the only videos that we know of where we can show you the
00:02:15.04 glow of the jellyfish.
00:02:17.08 And this is a jellyfish upside-down
00:02:20.25 in a beaker in an aquarium at Alaska
00:02:24.18 where the jellyfish can still be found.
00:02:27.01 And we are illuminating with ultraviolet light right now.
00:02:31.27 And you may be able to see that there's a circle here.
00:02:36.18 There are little dots around the edge; that is the jellyfish glowing.
00:02:41.02 And this is the green fluorescent protein naturally made by the jellyfish
00:02:45.03 that we are exciting with the UV lamp.
00:02:47.06 I'm sorry the beaker is a bit fluorescent.
00:02:49.16 We should have got a better beaker that didn't have any,
00:02:52.01 but it's some yellow background.
00:02:53.19 But nevertheless, you can see the jellyfish there.
00:02:57.02 And it's upside-down and confined to this little beaker.
00:02:59.29 And what we're gonna do in a moment
00:03:01.11 is poke it to stimulate the flash.
00:03:04.18 So we're gonna start here, and you may be able to see
00:03:07.22 that the jellyfish is slightly moving around.
00:03:10.08 In this video, we are playing with the illumination
00:03:13.05 trying to get it nicely aimed at the jellyfish,
00:03:16.26 and in a moment, we're going to bring in this rod, with which we are going to poke it.
00:03:21.17 Then we turn out the lights, and we stir.
00:03:23.26 And during that stirring, that was the light emitted by the jellyfish
00:03:28.08 - not the light that previously had been applied from a UV lamp.
00:03:33.18 So that's it. That's the flash.
00:03:35.29 And maybe, you might have been able to see
00:03:38.19 it was sort of greenish - maybe.
00:03:41.04 Anway, the man who discovered that phenomenon is shown here.
00:03:46.07 This is Osamu Shimomura, and in 1962,
00:03:50.11 he published his first paper on aequorin.
00:03:54.06 And that was really the protein he was most interested in.
00:03:56.07 It had the really exciting job of turning chemical energy into light.
00:04:00.13 And he also mentioned that he found this contaminant
00:04:03.22 that got in the way a little bit of the purification
00:04:05.28 and that it was a greenish protein that fluoresced green
00:04:09.27 and that later he mentioned that it changed color of the jellyfish from the blue of aequorin to the green of GFP.
00:04:21.11 Later in 1979, he actually proposed the structure of the chromophore, GFP
00:04:25.27 based on surprisingly little evidence
00:04:29.00 but he got it almost absolutely, completely correct.
00:04:32.00 There were only very very small modifications from later work that came in.
00:04:37.06 This is a picture of him much later in 2008
00:04:40.19 in the rehearsal for the Nobel Ceremony.
00:04:42.08 And he is actually holding a UV lamp.
00:04:44.29 That's the violet that's on your left, and then next to it is the tube of GFP.
00:04:50.27 And this is the last tube in existence that is actually made from real jellyfish that he had kept in his freezer all this time.
00:04:58.05 Since then, everything has been made by molecular biology
00:05:02.02 and that's due to this man, here, Douglas Prasher at the center
00:05:04.22 of the photograph who in 1992 published the cloning of the gene for GFP.
00:05:10.23 And that was a long struggle. He didn't know what to look for.
00:05:15.11 And he could not predict that it would become green fluorescent.
00:05:18.19 In fact, that was a major worry
00:05:20.08 that there was no precedent for a protein that made by a biological organism would absorb blue light
00:05:27.27 and fluoresce green.
00:05:30.14 We didn't know whether it needed any cofactors.
00:05:35.01 Normally, in many cases, when biology does this sort of remarkable thing,
00:05:39.07 with light, it uses small molecule cofactors.
00:05:42.22 you're seeing me, for example, not because the proteins in your eye directly
00:05:47.06 absorb the light through the protein, but rather,
00:05:49.08 because the proteins bind retinal, the pigment that we have
00:05:53.07 to get from other sources, and that binds into the protein to make the functional protein.
00:05:59.15 And for all we knew, the fear was that maybe this protein used another cofactor
00:06:05.23 Or the chromophore structure that Shimamura had proposed might take many enzymes to generate out of the protein.
00:06:15.21 But, Marty Chalfie, who's shown here on the right side of the picture
00:06:20.01 then got the clone, the gene, from Prasher, as I did as well,
00:06:26.10 but Marty was first that proved that just that gene alone was
00:06:31.13 self-sufficient; in other words, if you took that gene
00:06:33.13 and expressed it in other organisms,
00:06:35.15 and he did it in E. coli and
00:06:37.23 in worms, then you would generate fluorescence.
00:06:40.21 And that proved that GFP was self-sufficient.
00:06:44.01 It knew how to make its own chromophore out of its own guts.
00:06:47.25 Fred Tsuji, not shown here, also did very very soon after Chalfie independently.
00:06:55.25 So, we now know that the chromophore structure
00:06:58.28 is something like this structure here
00:07:02.03 at the right, but it had to come from the
00:07:06.08 protein structure shown here
00:07:08.24 where just at this stage still amino acids.
00:07:11.03 There's a serine, a tyrosine, and a glycine
00:07:15.03 that somehow have to be modified,
00:07:16.20 and we have to introduce through a various number of steps,
00:07:20.05 which people are still now somewhat debating exactly how it happens.
00:07:24.16 We generated extra double bond; we generate a ring, and so on.
00:07:28.23 And all of this does require oxygen.
00:07:31.19 If you do not have oxygen in the atmosphere
00:07:34.02 when this protein is growing, it cannot become fluorescent or doesn't even absorb light - let alone become fluorescent.
00:07:40.27 So the oxygen is essential
00:07:44.27 and the only creatures which are biologically capable of using GFP
00:07:50.05 assuming we can get the DNA, the gene for GFP, in would be organisms that
00:07:56.16 cannot tolerate oxygen at any stage of their life cycle.
00:08:00.14 In other words, obligate anaerobes.
00:08:02.08 But they are a limitation. You do need oxygen.
00:08:05.15 So then, this may seem to be a completely unprecedented reaction,
00:08:10.06 but it turns out that the crucial first step is the attack of a backbone NH from Gly67
00:08:17.19 onto the amide bond of Ser65.
00:08:22.16 And this is a very strange reaction, you might think.
00:08:25.25 Since when does one amide attack another.
00:08:29.09 But in fact, it turns out to be fairly similar to a known
00:08:33.16 reaction in biochemistry
00:08:35.01 where the side chain amide of glycine attacks the side chain amide of Asparagine.
00:08:44.10 And that is promoted by bringing them closely
00:08:48.14 and it requires the special properties of glycine that it can curl up
00:08:53.17 and also that it's relatively unencumbered in its NH group.
00:08:56.27 And it's interesting that of all the fluorescent proteins that have ever been discovered,
00:09:02.01 the most conserved residue is not the one that contributes the most atoms to the chromophore
00:09:07.19 nor this other one that gets attacked,
00:09:10.06 but rather Gly67 - every known fluorescent protein has Gly67 in it.
00:09:16.27 And maybe it's because that's necessary for that first step.
00:09:19.29 Later on, the Asn-Gly goes into a different path.
00:09:26.15 It also forms a ring, but it spits out ammonia and so on.
00:09:29.28 And they diverge. Now there's one important other feature about this chemistry that I have to mention.
00:09:36.07 Oxygen comes in and at some stage pulls off the hydrogens
00:09:40.28 that were on the tyrosine
00:09:44.01 connecting the beta-hydrogens of the tyrosine that connect it to the rest of the chromophore.
00:09:53.14 So O2 pulls off H2, and the product is hydrogen peroxide.
00:10:00.26 H2 + O2 does not give you water in this case.
00:10:04.07 You have to balance the number of hydrogens and oxygens.
00:10:06.23 H2 + O2 gives you H2O2, which is hyrdogen peroxide.
00:10:10.08 So there is one molecule of a slightly toxic substance
00:10:14.15 that's generated for every molecule of GFP.
00:10:16.23 And that's sort of required by the conservation of mass.
00:10:20.20 It has been now directly verified, and it is a potential problem
00:10:25.14 that GFP is not totally totally safe for the organism,
00:10:30.11 and making a heck of a lot of GFP could generate one mole of hydrogen peroxide per every mole of GFP you make.
00:10:36.22 And you have to keep that in mind.
00:10:38.29 Now, fortunately, most organisms that have grown up in air has some defense
00:10:43.00 against a slow bit of hydrogen peroxide that's trickling out.
00:10:46.10 And we seem to be ok, but it's something that you have to keep in mind.
00:10:49.29 I should also say that in this original scheme
00:10:52.16 we propose that dehydration went before oxidation, and actually,
00:10:57.08 that's not quite so clear at exactly which stage the oxygen comes in and makes hydrogen peroxide
00:11:02.20 maybe a little bit earlier than I drew here.
00:11:05.27 So the original jellyfish protein actually was not very strongly fluorescent.
00:11:12.09 This dotted line here is the spectrum of the wild-type GFP.
00:11:17.28 And it has a big peak at around 400 nm - just below 400 nm.
00:11:24.10 In other words, the jellyfish actually GFP glows best in the UV.
00:11:30.11 And it only has a minor peak out here in the sort of green that gave it its name
00:11:35.16 and that's the one everyone wants to use.
00:11:37.21 Why the jellyfish deliberately crippled this protein
00:11:42.01 so that actually 5/6th of the time roughly, it is generating the UV,
00:11:48.01 and only 1/6th of the time, the visible, we really don't know.
00:11:51.05 Nevertheless, mutagenesis soon showed that particular mutations at Ser65
00:11:58.27 which you might think wasn't that close to the chromophore
00:12:01.29 though it contributed some of the atoms turns out to clean up the spectrum enormously
00:12:06.18 and get rid of this UV peak
00:12:09.06 and accentuate the visible peak
00:12:11.16 and also make the protein more stable,
00:12:13.19 and so these are the ones, the descendants of these improved proteins are the ones that
00:12:17.25 everyone uses nowadays including the very popular
00:12:21.13 so-called eGFP - e standing for enhanced,
00:12:24.14 which was this mutation S65T + a folding mutation to make it fold a little bit better.
00:12:32.14 Another feature I should say is that the jellyfish originally
00:12:36.05 grew and lives in Puget Sound or very cold water.
00:12:40.24 It never had any reason to worry about the ability to fold
00:12:44.27 at warm temperatures, but so many of us
00:12:46.29 want to work on mammalian tissue
00:12:48.29 at 37 degrees, this original protein just sorta collapsed
00:12:53.05 and basically wouldn't fold officially.
00:12:55.21 And a lot of mutagenesis of which this was just the beginning
00:12:58.28 then gradually made the protein able to tolerate warmer temperatures and fold more efficiently.
00:13:04.16 So later on, the crystal structure appeared almost simultaneously
00:13:10.19 from two groups in 1996
00:13:12.07 one of them was of this mutant S65T
00:13:15.25 and it showed a beautiful 1-stranded beta-barrel.
00:13:20.16 So the GFP is almost a perfect cylinder
00:13:23.01 made out of these strands that cross in this sort of helical lattice
00:13:27.04 around the outside, and then up the middle is
00:13:31.03 an alpha-helix that carries the chromophore.
00:13:34.06 And the chromophore in this model here is the bit in the middle that has the red and blue atoms that show up at atomic level
00:13:42.27 whereas everything else is just beta-strands or alpha-helix.
00:13:46.21 And this revealed why, in a way, no other enzymes were necessary.
00:13:53.07 They couldn't have been necessary because the chromophore's completely deeply buried inside this cylinder
00:13:58.18 so nothing else could have ever gotten that.
00:14:00.27 So it was essential in a way that the protein learned to do surgery on its own guts
00:14:05.21 and thereby generate the chromophore
00:14:08.27 in that chemistry I was just discussing where the hydrogen oxygen somehow has to get in,
00:14:12.28 hydrogen peroxide gets out.
00:14:15.11 Simultaneously, the other structure which was more of the original wild-type GFP
00:14:20.21 happened to be in a different crystal form
00:14:23.02 and showed that the GFP was also dimeric.
00:14:26.05 And that's another form of GFP, and actually,
00:14:29.12 there's an equillibrium between the two.
00:14:31.10 And it turns out that the GFP has a hydrophobic patch
00:14:35.28 where these amino acids mentioned here - three of them from each of the subunits get together
00:14:44.26 and this greasy patch enables dimerization.
00:14:48.24 And this is something to keep in mind again.
00:14:50.13 And one of the slight problems with wild-type GFP is that it likes a little bit to dimerize
00:14:55.10 And dimerization is generally bad when we want to tag a particular protein
00:15:01.20 because that forces the protein that we're really interested in, the partner, to also become dimeric
00:15:07.11 when it wouldn't otherwise have been.
00:15:09.18 We have forced it to be dimeric by fusing it to GFP.
00:15:13.12 And that often changes the cell biology and messes things up.
00:15:17.11 It doesn't matter so much if you're just trying to light up a cell
00:15:20.15 and you're using isolated protein because
00:15:22.20 then the fact that it wants to be a dimer, if it's not stuck to anything,
00:15:26.07 nobody else cares, so to speak.
00:15:28.23 Nevertheless, there's been a lot of interest in the dimerization, and we now know exactly how to get rid of it.
00:15:35.09 For example, if you take the A206, which is one of the ones mentioned
00:15:41.21 and change it to a lysine, A206K
00:15:44.13 completely destroys the dimerization because
00:15:47.09 then a lysine and a lysine replacing the alanine and alanine are now
00:15:51.11 in each other's face, and the electrostatic repulsion
00:15:54.09 between the two plus charges blows apart this dimer interface
00:15:58.01 and we supress the dimerization.
00:16:00.08 So in anytime that you are in any concern about it, there are mutations that can be used
00:16:05.08 to fix the dimerization problem, and they don't seem to have any bad effects on any of the other properties of GFP.
00:16:11.04 So can we get other colors?
00:16:15.00 And there was a lot of interest in getting other colors for various reasons.
00:16:19.17 The very simplest is that sometimes we want to follow several different proteins simultaneously
00:16:23.19 or follow different cells, which are marked with different and by making them different colors,
00:16:29.27 we can distinguish all of these.
00:16:31.22 So the original color is pretty close to this - the green output
00:16:38.15 if you forget the UV that the wild-type produces, you wouldn't have seen it with your eyes anyway
00:16:44.06 the residual green is very similar to this green from S65T
00:16:48.21 except S65T puts all of its energy into this green instead of only 1/6th of it.
00:16:54.26 It turns out to be possible to make a blue, a cyan, and evena somewhat yellowish version.
00:17:02.01 It's not awfully yellow; it's a yellowish-green still. But we call it YFP.
00:17:06.08 And these were found by various substitutions of the chromophore or around the chromophore.
00:17:12.20 So for example, if you change the original tyrosine in the chromophore, which is here,
00:17:20.27 and change it to a histidine, then we get the blue.
00:17:23.22 If we change it to a tryptophan, we get the cyan.
00:17:27.08 Ánd by the way, the UV peak was when this tyrosine was neutral.
00:17:32.27 And the green peak was when the tyrosine is negatively charged.
00:17:38.09 And this turns out to be crucial for my next lecture where we often, people now play with this equilibrium
00:17:44.06 and make it subject to environmental influences.
00:17:47.17 And by changing the ratio basically of the UV to the visible peak, we can make a GFP that switches on due to environmental influences.
00:17:56.02 And finally, the yellowish version is by taking an amino acid that's remote in the primary sequence that's
00:18:03.01 threonine 203 changing it to things like tyrosine or phenylalanine
00:18:08.28 and thereby it's close in 3-dimensional space.
00:18:11.23 It stacks above the real chromophore, and
00:18:13.28 the pi-pi stacking is what we believe is responsible for the yellow shift
00:18:18.18 that makes it a somewhat yellowish-green.
00:18:21.16 If we now look at the range of fluorescent proteins up to now,
00:18:26.02 everything was based on GFP from the jellyfish with mutational engineering
00:18:31.06 but it turns out that fluorescent proteins are fairly widely distributed in nature.
00:18:36.04 In fact, there are homologs in vertebrates
00:18:39.18 though they happen not to be fluorescent
00:18:41.29 if they were, we'd all be walking around like sort of the green giant.
00:18:46.10 But in fact, the ones that are all fluorescent all come from
00:18:51.03 the phyllum Cnidaria
00:18:52.27 which is the glorified Latin name for basically the family of coelenterates that includes
00:18:58.13 jellyfish, and corals, and some other related organisms.
00:19:02.03 And it's not just the jellyfish that can make fluorescence, but it turns out that corals
00:19:07.22 and that was the brilliant discovery of a Russian group, first author Matz et al.
00:19:13.02 And they have subsequently have done a lot of the phylogenetic tree by doing this classic molecular evolutional and genome analyses.
00:19:22.01 So a lot of you know, of course, tropical corals have beautiful colors.
00:19:27.20 Every scuba diver knows that or even snorkeler
00:19:31.12 but what we didn't realize is that so many of those colors are actually due to
00:19:34.08 homologs of GFP,
00:19:36.21 and particularly, one that attracted a lot of interest and was the subject of that first paper by Matz et al.
00:19:43.29 was Discosoma, which has a beautiful red fluorescence.
00:19:48.14 And that paper was about the cloning of the gene
00:19:52.17 for a red fluorescent protein that they called Ds for Discosoma - DsRed.
00:19:59.01 And it proved to be this beautiful red color.
00:20:02.18 At that time, it was not clear right away how did the corals manage to coerce the chromophore into absorbing green light
00:20:12.09 now instead of blue
00:20:13.18 and fluorescing red instead of green.
00:20:16.05 But chemical analysis of the chromophore eventually showed
00:20:22.20 that was the corals had figured out was add an extra double bond, which is between the carbon in this place and the nitrogen here
00:20:35.21 and this forms a very unusual structure called an acylamine
00:20:39.06 that is basically completely unstable
00:20:41.11 in ordinary solution and is only held together
00:20:43.06 by virtue of being inside the protein
00:20:45.19 but that extra double bond here in addition to all the double bonds that were up here
00:20:50.17 now extends the chromophore and it recruits this ketone or carbonyl group that used to be part of amide
00:20:58.23 and this extended chromophore then turns out to beautifully explain the red color as even shown by quantum mechanical calculations.
00:21:06.00 And when this structure was determined, some people in my lab were really
00:21:11.27 pleased with themselves, and they used the actual bacteria expressing the protein and used this ink, and with little toothpicks, they drew the structure on a petri dish.
00:21:22.16 So you can both see the beautiful red color and the structure that gives rise to to it all in one picture.
00:21:28.14 Now DsRed made by the corals as typical was made by the coral for its own reasons
00:21:35.14 and to this day, as usual, we really don't have a definitive explanation that everyone accepts
00:21:41.13 for why the coral should want to have these colors.
00:21:45.08 But whatever it is, the protein was not ideal for cell biological use.
00:21:50.07 The biggest problem was that it turned out to be a very tight obligate tetramer.
00:21:55.07 GFP was a weak dimer, and even the weak dimerization gave problems.
00:22:01.19 But DsRed being a tight tetramer often prevents proper trafficking, or fusions.
00:22:06.18 It's even worse because it's really strong and tight,
00:22:10.05 and four copies of your protein that you're trying to label red
00:22:14.10 now get fused together
00:22:16.23 And an example here was connexin-43, which is a constituent of gap junctions
00:22:22.29 and its detailed structure is shown here.
00:22:25.05 If you fuse it to GFP, you can get a fluorescent gap junction
00:22:28.14 and it trafficks reasonably well and makes this fluorescent plaque in a micrograph, of course,
00:22:35.01 showing that there's a boundary between one cell and another
00:22:38.09 which is slightly fluorescent, but when you do the same with DsRed,
00:22:42.03 the plaque is of course trying to make this connexin-43 wants to make a hexamer on its own.
00:22:48.26 Two of them eventually get together to 12-mer or decamer; meanwhile, the connexin-43 is trying to make a tetramer.
00:22:56.14 The two clash with each other, and you get a mess.
00:22:58.21 And the protein can never make a gap junction.
00:23:01.03 There are additional problems like the DsRed took a really long time to turn red,
00:23:06.29 and it didn't finish the job.
00:23:08.00 It left some green behind.
00:23:09.19 But the fact that it went through green stage is an interesting clue.
00:23:13.02 Eventually, these were all fixed by mutagenesis.
00:23:16.06 It was much more difficult than with GFP because that tetramer was really hard to break,
00:23:21.25 But eventually now, we now have a whole gamut of different colors.
00:23:26.07 of fluorescent proteins derived from DsRed through monomeric forms,
00:23:32.26 and it turned out to be not too hard to change their color.
00:23:35.08 So over here are the four that came from the jellyfish -
00:23:38.24 blue, cyan, green, and yellow
00:23:41.00 and these are all, by the way, these are all little tubes of protein made in E. coli and purified and simply a photographed by their own fluorescent light
00:23:51.20 and then we eventually got these additional colors in my lab
00:23:56.26 and in order to separate them and keep them easy to remember, we gave them the names of fruits
00:24:02.22 which each color references. Like this one is honeydew.
00:24:06.28 because honeydews are sort of yellowish green.
00:24:09.03 And then there's tangerine and strawberry and cherry and so on
00:24:13.14 - all these different beautiful red colors.
00:24:14.15 The one exception that really isn't monomeric is so-called tandem dimer Tomato.
00:24:20.16 And that is actually two copies of the original 11-stranded beta-barrel that are permanently genetically fused to each other.
00:24:28.08 And in that way, they satisfy each other.
00:24:30.25 This protein is not a tetramer; it's a dimer. But it's internally satisfied.
00:24:35.21 It's got two copies, and when you fuse it, you fuse both together as one unit.
00:24:42.02 And that's why we call it a tandem dimer.
00:24:44.17 So I wanted to mention then, derived from these new coral proteins, there have been
00:24:52.17 actually there's some varieties that were made from original jellyfish as well.
00:24:56.28 It now turns out that there are proteins that can be deliberately changed with light.
00:25:03.05 Of course, some proteins bleach, and that can be very useful in techniques
00:25:08.14 like fluorescence recovery after photobleaching.
00:25:11.01 But an even more spectacular case is this protein called dendra
00:25:16.09 which starts out green like DsRed.
00:25:21.15 But whereas DsRed spontaneously changes from green to red,
00:25:25.09 this one sits green indefinitely until you shine blue light on it
00:25:29.21 and that is what switches it from green to red.
00:25:32.29 So we can use this as an optical highlighter
00:25:36.04 to, for example, in this case track the migration of
00:25:40.13 a protein called fibrillarin
00:25:42.05 these authors fused Dendra to fibrillarin and locally
00:25:47.00 turned, shined blue light on, turned some of it from green to red and then watched
00:25:51.25 the subsequent fate of just those proteins that had been labeled.
00:25:55.24 So those proteins that were in the spot first lost their green color - not quite completely.
00:26:02.13 It might have been better if it had been complete, but they lost it considerably.
00:26:06.04 And then there's some recovery as proteins move around into this illuminated zone.
00:26:10.16 And meanwhile, the red protein made in the spot then over time spreads out.
00:26:16.15 And we can watch. And these are 1, 2, 3, 4 are different regions of interest in the original that were being tracked.
00:26:24.11 And even more spectacular form of photochemistry is shown by Dronpa
00:26:29.13 which is reversible, and it starts out green fluorescent,
00:26:33.08 and as you continue to excite its green fluorescence using blue-green light,
00:26:37.17 it fades away and reaches essentially down to nothing.
00:26:41.23 And you might think that this protein is dead, and most fluorescent proteins had you done it, you'd be right.
00:26:47.08 But this one, when you shine violet light on it, springs back to life
00:26:51.29 and comes back fully completely, just about, rejuvenated.
00:26:56.27 And we can do this cycle after cycle after cycle, and we show here,
00:27:00.27 these are compressed time course of probably 50 cycles,
00:27:07.07 and there's a slight bit of rundown as you go through many many many cycles.
00:27:12.01 But you can do this over and over again.
00:27:13.12 And the applications for repetitively following movements inside cells or as you will later hear from other people for doing super-resolution microscopy.
00:27:24.04 There are major implications that have been exploited, but I don't have the time to go into them here.
00:27:28.08 But the authors were so pleased with themselves that they
00:27:31.06 simply laid down a film of protein, a wet film of protein
00:27:35.26 and then they shined its own name successively using this trick of
00:27:41.08 starting fluorescent, bleaching it completely, shining the letter D,
00:27:47.21 a silhouette of the letter D,
00:27:48.22 onto this film of protein, and then they
00:27:51.13 lit up the letter D, and then they erased it again.
00:27:54.01 and then they wrote it again, this time the letter R, et cetera and wrote out its name.
00:27:58.17 Ok, so I don't have time talk more about these very interesting photochemically active proteins.
00:28:09.04 I'm now gonna turn briefly to the question of how do we look inside a more intact animal.
00:28:13.01 especially animals that are not quite as transparent as the ideals.
00:28:19.00 Now, the ideals may be zebrafish and a lot of microscopists like zebrafish or C. elegans because
00:28:25.05 they are transparent, but many of the other important ones like mice or flies are somewhat more opaque.
00:28:33.21 And in general, most organisms, especially mammals are nearly completely opaque
00:28:41.02 down in short wavelengths
00:28:42.03 below 600 nm, and that's because of the peaks of hemaglobin.
00:28:46.20 After all, that's what makes us pink.
00:28:48.27 why flesh is red, and as long as it's there,
00:28:53.00 you can't easily excite through much thickness of this tissue.
00:28:58.03 A standard demonstration is that if I try to shine a green laser through even the thinnest of my fingers,
00:29:03.16 nothing will get through, but if I shine a red laser,
00:29:05.28 laser pointer, a lot will get through,
00:29:08.17 and that's because a red laser is about 630 nm
00:29:12.05 beyond the shoulder, so it would be really nice
00:29:14.19 if we had fluorescent proteins that could be both excited and would emit longer than 600 nm.
00:29:22.26 And the record for current proteins derived from the jellyfish, coral, cnidarian family is just barely over 600 nm
00:29:32.08 and sort of marginal.
00:29:33.01 But we'd like to get out say to the 700 or so.
00:29:37.14 And it turns out that this is possible, but you have to be willing to go to a whole new protein family.
00:29:43.13 And this protein family is now back to cofactors.
00:29:49.16 These are small organic molecules.
00:29:51.20 This is what we were afraid the jellyfish and corals were using.
00:29:55.19 Now, we actually use them, and these pigments are derived
00:29:58.26 from the heme biosynthesis pathway.
00:30:01.12 Particularly, this is heme with a porphyrin, which is four pyrroles in a big circle
00:30:11.02 and with an iron in the middle
00:30:12.24 and when we break down heme or any organism that uses heme, that's essentially the entire kingdom of life - almost
00:30:22.22 when we break it down with an enzyme called heme oxygenase, the first breakdown product is called biliverdin.
00:30:28.26 Biliverdin is something that every adult human being makes about a quarter gram of every day
00:30:34.21 Normally, it's sort of kept sequestered, but if any of you have had a big bruise,
00:30:40.21 and you turned a sort of black and blue, a lot of that color is broken down heme that makes biliverdin.
00:30:47.09 And later bilirubin here, which is part of what gets excreted.
00:30:52.15 So it turns out that weak biliverdin is phylogenetically ubiquitous
00:30:58.28 Some organisms elaborate it into known fluorescent pigments that have these more elaborate cofactors.
00:31:06.28 But it turns out that bacteria use bilverdin in proteins that sense light.
00:31:11.16 And by mutagenesis, it was possible to prevent them from sensing light and instead use the energy to fluoresce.
00:31:18.22 These now make infrared fluorescent proteins that absorb at 680-some nanometers and emit just over 700 nm, hence, their name as infrared fluorescent proteins.
00:31:30.17 And they are genetically encodable; you have to make sure that the organism either provides biliverdin,
00:31:38.18 or the cell provides biliverdin. And as I said, most of our cells make it.
00:31:41.16 If necessary, you could can supply the extra by injection if you don't think they make enough of their own.
00:31:50.08 And so up to now, this has been the only way to decisively get beyond 600 nm.
00:31:55.15 And an example of its use here is when this is transfected into the liver;
00:32:01.28 now we can see the liver through the skin of an intact mouse
00:32:05.04 Far far better than even a relatively red-shifted protein that was derived from corals
00:32:11.29 And in turn, that is better than GFP,
00:32:14.22 which is essentially hopeless.
00:32:16.24 You cannot see the liver
00:32:18.24 of a corresponding GFP-transfected mouse that's got that.
00:32:21.14 Instead, all you see is the autofluorescence or the fur around the outside.
00:32:26.24 Whereas this stuff is shining through the abdomenal cavity.
00:32:31.11 And there's many things more that need to be done to improve this relatively recent work.
00:32:35.25 There's actually thousands of phytochromes; this is extremely widely diversified gene family.
00:32:41.16 There may be possible to get longer than what I just described.
00:32:45.07 And it should be helpful for lots of interesting in vivo macroscopic fluorescence imaging
00:32:52.08 somewhat similar to microscopy but at a more a less-space resolution but greater depth inside intact animals.
00:32:59.06 It provides colors that are orthogonal even to the wide range we've already got.
00:33:05.18 It could be an acceptor for fluorescence resonance energy transfer if it's quantum yield was improved a bit more.
00:33:11.23 It may be also good for the photoacoustic type of imaging
00:33:19.06 and many other tricks, and finally if heme oxygenase activity is itself biologically important in a lot of metabolic activities,
00:33:27.21 it's responsible for making carbon monoxide and helps makes cyclic-GMP,
00:33:32.18 and this is a possible way we could detect that activity in a cell that doesn't have much of its own and doesn't have any other source of biliverdin.
00:33:41.01 So in conclusion, these are some of the people in my lab who contributed or that we collaborated with
00:33:48.07 who contributed most to it,
00:33:50.16 and once again, I'm showing you the fun that you could have with these multi-colored living inks made out of fluorescent proteins in which some people in my lab with more artistic talent than I did
00:34:02.15 actually tried to draw a sunset with a green flash as you might see it from the beach not far from our lab.
00:34:10.20 Thank you very much.

This Talk

Speaker: Roger Tsien

Audience:

Student
Educators of Adv. Undergrad / Grad
Researcher
Educators

Recorded: May 2012

Talk Overview

Live cell imaging has been revolutionized by the discovery of the green fluorescent protein (GFP). Roger Tsien talks about fluorescent proteins and covers the history of GFP, how it folds and becomes fluorescent, how it has been mutated to produce additional colors (blue, cyan, yellow). It also covers novel photoswitchable and photoactivatible fluorescent proteins, whose color can be changed by light, and new infrared fluorescent proteins.

Questions

Red fluorescent proteins were derived from which organism?
Which of the following is true for the generation of the GFP fluorophore?
1. Hydrogen peroxide is generated in the reaction
2. The chemistry to produce the GFP fluorophore occurs a few seconds after protein folding
3. The fluorophore will form in anaerobic environments
4. The glycine carbonyl is crucial for the cyclization reaction
The main effect of the S65T mutation is to
1. Make the protein fold better
2. Shift the emission spectrum
3. Shift the excitation spectrum
4. Produce faster maturation of the fluorophore
Which of the following is true:
1. Dendra photoactivates with blue light and shifts its emission from blue to green
2. Dronpa’s unique property is that its fluorescence disappears but can be photoactivated by red light
3. Blue fluorescent protein would be a good choice for deep tissue imaging
4. Infrared fluorescent proteins require the cofactor biliverdin

Answers

View Answers

Speaker Bio

Roger Tsien

Dr. Tsien was a Professor at the University of California, San Diego, a Howard Hughes Medical Institute Investigator and a member of the National Academy of Sciences. In 2008, Tsien shared the Nobel Prize in Chemistry for the discovery and development of green fluorescent protein, GFP. His lab continues to develop new fluorescent proteins as… Continue Reading