Top
SCIENCE. STORIES. IMPACT.

iBiology

Bringing the World's Best Biology to You

Optimizing Detection of GFP

00:00:11;00 So, let me now go over some of the
00:00:14;00 factors that affect the detectability of
00:00:16;00 GFP, and this is one of the problems that
00:00:18;00 people working in GFP used to have a lot
00:00:22;00 of trouble with. And even now if you were
00:00:24;00 a student trying it for the first time
00:00:26;00 on a protein that hasn't already been
00:00:28;00 worked out, you may find difficulty in
00:00:31;00 getting your signal. And people throw up
00:00:34;00 their hands. "What's wrong with my
00:00:36;00 experiment?" Well if you analyze it,
00:00:38;00 there's a whole bunch of factors. And
00:00:40;00 here the four major categories, which
00:00:42;00 we're then going to break down. The first
00:00:45;00 is you just got to make enough GFP.
00:00:47;00 GFP is not infinitely bright. The next is
00:00:50;00 that having made the protein by protein
00:00:53;00 synthesis off the ribosome - presumably
00:00:55;00 that's how it's genetically encoded - you
00:00:58;00 then have to make sure that it does a
00:00:59;28 good job of post-translationally making
00:01:04;00 that fluorophore. The third category is the
00:01:08;00 molecular properties of the mature GFP.
00:01:11;00 And this is the most intrinsic factor
00:01:13;00 that takes a protein engineer to fix and
00:01:16;00 the one that say things like S65T-
00:01:19;00 improved. And nowadays this is less of a
00:01:22;00 problem because people typically know
00:01:24;00 how they can acquire either from their
00:01:27;00 friends or go out and buy a pretty
00:01:29;00 up-to-date GFP that's already been
00:01:30;00 optimized in this. And finally, you have
00:01:32;00 to remember that you have to detect
00:01:34;00 these signals in a competition with
00:01:36;00 noise and background and there is noise
00:01:38;00 and background and fluorescence. A lot of
00:01:41;00 the rest of this microscopy course will
00:01:42;00 be teaching you about how to optimize
00:01:45;00 that balance. But a few things can be
00:01:47;00 said even now. So, let's go back to total
00:01:49;00 amount of GFP. Obviously, you've got to
00:01:52;00 have enough copies of the gene and
00:01:53;00 duration of expression, and this is back
00:01:55;00 to your traditional molecular biology.
00:01:57;00 You have you also have to have a strong
00:01:59;00 enough and promoter and enhancer. The
00:02:01;00 translational efficiency has to be
00:02:03;00 correct and that includes the so-called
00:02:05;00 Kozak sequence, which is the sequence
00:02:09;00 right up at the front of the coding
00:02:12;00 amino acids, which has certain optima to
00:02:14;00 help the ribosome get started. It's
00:02:17;00 helpful if you have the correct codon
00:02:20;00 usage. If you, the original GFP, of course,
00:02:23;00 had jellyfish codon usage.
00:02:25;00 Now you can get genes which
00:02:26;00 have been optimized for mammalian
00:02:28;00 or bacterial and so on. And there was a
00:02:30;00 time in the old days when people had
00:02:33;00 trouble with it because they were using
00:02:35;00 the wrong codons in their nucleic acid.
00:02:38;00 And you also want to avoid mRNA splicing.
00:02:41;00 It turned out that plants could find a
00:02:46;00 splice site within the jellyfish gene
00:02:49;00 and splice it out and of course then you
00:02:52;00 didn't make a functional protein. Most
00:02:55;00 these problems now are less important
00:02:57;00 because so much engineering and
00:02:59;00 experience has been done. How about the
00:03:01;00 efficiency of post-translational fluorophore
00:03:03;00 for formation. Originally, if you tried to
00:03:05;00 express this heavily in bacteria,
00:03:08;00 especially if you did it warm, the
00:03:09;00 protein would just crash out and
00:03:11;00 denature because it didn't like being
00:03:13;00 made in such quantities. And if you have
00:03:16;00 chaperones available to help fold
00:03:18;00 proteins better that helps. In fact, this
00:03:20;00 has even been used as a bioassay
00:03:22;00 internally for the presence of
00:03:24;00 chaperones. Pick up a crummy GFP, it has
00:03:27;00 trouble folding and see whether you can
00:03:29;00 make it fold. And that may be dependent
00:03:31;00 on the chaperone that you want to assay.
00:03:32;00 A lot of the times when you fuse this to
00:03:36;00 a host protein, that fusion has some
00:03:39;00 unfortunate properties very hard to
00:03:41;00 predict and the protein just doesn't
00:03:45;00 like to express. And sometimes people
00:03:47;00 already know. My favorite host protein
00:03:49;00 just won't tolerate you putting
00:03:51;00 anything on the N-terminus. Well then
00:03:53;00 don't be surprised if you try to put GFP
00:03:55;00 there, it won't like it. Sometimes we
00:03:57;00 don't know that in advance. And commonly
00:03:59;28 if the N-terminus doesn't work try the
00:04:02;00 C-terminus. GFP can work at either one in
00:04:05;00 general. But for any given protein you
00:04:07;00 don't know until you try it, and even if
00:04:10;00 they've poop out and don't work for you
00:04:12;00 you can even try to put GFP, insert it
00:04:16;00 within a loop of, or a comfortable spot
00:04:19;00 in the host protein. Sometimes that
00:04:22;00 solves the problem. Of course, at this
00:04:23;00 stage you're doing a lot of work of
00:04:24;00 trial and error but that's what it may
00:04:27;00 take. And of course, you do have to give
00:04:30;00 it time. Those post-translational
00:04:32;00 reactions in those complicated sequences
00:04:35;00 that I showed you in the ---
00:04:37;00 with asparagine glycine, all that takes
00:04:39;00 time. We now have some that can go pretty
00:04:42;00 fast and do it in only a few minutes, but
00:04:44;00 originally it took four hours from the
00:04:48;00 original jellyfish. So that's been
00:04:49;00 improved a lot by further
00:04:51;00 engineering. And as I said, we now have
00:04:53;00 fairly warm temperature tolerant
00:04:57;00 versions. As I said, you must have oxygen
00:04:59;00 present at least during the maturation.
00:05:01;00 Once the protein is mature you don't
00:05:04;00 have to have oxygen still there. The GFP
00:05:06;00 works just fine but it did need it just
00:05:08;00 that once, once per each molecule. So, then
00:05:14;00 the molecular properties. And we want to
00:05:16;00 improve the wavelengths of excitation
00:05:17;00 and emission, make them ones which are
00:05:19;00 easily detectable. And there'll be a lot
00:05:21;00 more on that in this lecture. We want to
00:05:23;00 improve the extinction coefficient, which
00:05:25;00 is the ability to absorb light, and
00:05:27;00 fluorescence quantum yield, which is once
00:05:30;00 you've absorbed a light how probable are
00:05:32;00 you to reemit it. GFP now is actually
00:05:35;00 pretty good. The good GFP, the best ones
00:05:37;00 now, have quantum yields of seventy
00:05:40;00 to eighty percent which is, and
00:05:41;00 theoretically you can never go up above
00:05:43;00 a hundred percent. So, there's not much
00:05:45;00 room left to improve that one. But we are
00:05:48;00 all still very interested usually in
00:05:50;00 preventing photoisomerization or
00:05:53;00 bleaching unless we specifically want a
00:05:55;00 trick that it requires bleaching GFP
00:05:58;00 sometimes, and I'll briefly mention some
00:05:59;28 of those. But most people, for most
00:06:02;00 applications, just want thing to shine as
00:06:05;00 brightly as long as possible. And this is
00:06:07;00 still an area of active investigation,
00:06:09;00 how to improve that. And I mentioned
00:06:11;00 importance of knowing whether it's
00:06:13;00 dimerizing, and in some cases getting rid of
00:06:15;00 it. Particularly when you are fusing it to a
00:06:17;00 host protein. Finally, competition with
00:06:21;00 noise and background signals. You must
00:06:23;00 remember that a lot of the culture media
00:06:25;00 are also fluorescent at similar
00:06:27;00 wavelengths. Particularly if they have
00:06:29;00 riboflavin, if they have a lot of serum
00:06:32;00 proteins and so on. And that will
00:06:34;00 generate a haze of fluorescence that has
00:06:37;00 nothing to do with GFP but will obscure
00:06:39;00 small amounts of GFP. So, you want to get
00:06:42;00 rid of those if you're ever having a
00:06:44;00 detection problem. The location of GFP
00:06:47;00 actually matters somewhat, particularly
00:06:50;00 if you wanted
00:06:51;00 the fewest copies of GFP you want to
00:06:54;00 localize it into a tight little spot,
00:06:57;00 ideally a micron or so in size, preferably
00:07:01;00 where you know an advanced way to look.
00:07:02;00 And then you only have to ask do I see
00:07:05;00 it, a dot there, or do I not? And at that
00:07:08;00 level people can often see just a few
00:07:10;00 copies. And in the very best cases it has
00:07:13;00 been proven you, in not very general
00:07:16;00 cases, but you can occasionally see just
00:07:18;00 one copy of GFP. Though typically, you're
00:07:20;00 using so much laser light that you
00:07:22;00 bleach it very very quickly. So, compared
00:07:26;00 to that if it's just diffuse, it's much
00:07:28;00 harder to see the difference between
00:07:29;00 diffuse GFP and the diffuse background
00:07:33;00 that cells already have. They already
00:07:34;00 have autofluorescence even when the
00:07:37;00 culture media is very clean. You need to
00:07:40;00 have very good filters, which don't
00:07:42;00 overlap. And a lot of people who are
00:07:45;00 first starting fluorescence microscopy
00:07:47;00 just pick any filters they got around
00:07:49;00 them, that happened be rattling
00:07:51;00 around in the drawer underneath the
00:07:52;00 microscope. And those are often not
00:07:54;00 optimized and we need ones which
00:07:56;00 accurately excite the GFP but do not
00:07:59;00 enable crosstalk. And crosstalk is when, say,
00:08:03;00 a little bit of the excitation light can
00:08:06;00 leak through the emission filter, so
00:08:07;00 scattered light then instead of
00:08:09;00 fluorescence can contribute. And that's
00:08:12;00 something you just got to buy better
00:08:13;00 filters and dichroic mirrors. Finally, of
00:08:16;00 course, we need good detectors. Your eye
00:08:18;00 is actually quite an inefficient
00:08:20;00 detector and nowadays it's just a matter
00:08:22;00 of spending money on a sufficiently good
00:08:25;00 camera. And I'm sure other people in the
00:08:27;00 course will tell you a lot more about.

This Talk
Speaker: Roger Tsien
Audience:
  • Educators of Adv. Undergrad / Grad
  • Researcher
  • Educators
Recorded: May 2012
More Talks in Microscopy Series
All Talks in Microscopy Series

Talk Overview

This lecture discusses factors controlling the detectability of GFP within the cell and how to optimize these factors so as to maximize its detectability.

Speaker Bio

Roger Tsien

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

Playlist: Microscopy Series

Reader Interactions

Leave a Reply

Your email address will not be published. Required fields are marked *

iBiology and iBiology Courses are part of the Science Communication Lab (SCL). Our mission remains the same, to connect people to science. However, our focus has shifted to producing and evaluating cinematic films for education and the public, which you can find on the Science Communication Lab website. For more information, please see this blog post!