Through a Glass, Darkly December 4, 2000

CalTech used a grading system that I really liked. The entire freshman year was pass/fail. This allowed the students plenty of time to find their place in the scheme of things without the intense pressure of actually having grades that count. I mean, I could study 160 hours a week (allowing myself 8 hours to sleep) and I would never get a higher grade in physics than J. T. (not his real name) who never studied and slept most of the time. Better to study a more reasonable amount of time, go to a few parties and be happy with a B. Or also go to a few movies and get a C. You know what I mean!

Real grades came the sophomore year. But, you were allowed to take one class pass/fail each quarter as long as it was not specifically required for your degree. And you did not have to declare which class this was until 2 weeks before the end of the quarter. So, you could choose whichever class you wanted credit for but were not going to get a good grade in. My GPA was lowest my sophomore year (mainly because I had to take a second year of physics and math) and went up every year after that.

See, we were also required to take a humanity/social science class EVERY quarter. In order to fight the stereotype that CalTech students were too one-dimensional, we actually had to take more liberal arts classes than scientists at other colleges. So, I took Comparative Religions, British Commonwealth and Empire, and Art 101. In fact, I would say that taking Art classes was the single most important thing for raising my GPA. I mean, I did fine in the Biology classes and I took some Freshman classes my senior year (Introduction to Solid State Devices. Since it was only given pass/fail, it was not included in my one pass/fail class a quarter. Helped my GPA but it also served its purpose by exposing me to a field that I actually kind of liked.)

The best art classes were the movie ones. We would watch a movie once to get an impression, then a second time more carefully. One quarter it was about horror/science fiction movies. I wrote a paper examining Invasion of the Body Snatchers and The Thing (the '50s versions) with respect to their reflection of the Communist Menace. Great BS papers for which I got great grades.

But the best movie class was on directors. We saw Red River by Howard Hawks and the Mortal Storm by Frank Borzage. Chuck Jones gave a guest lecture and showed his personal prints of several of his cartoons (with the classic Duck Amuck and Duck!Rabbit!Duck! being my favorites). It was marvelous getting to look at some of these movies and see how the director was able to create a mood by moving the camera, or not.

So, I used these tricks to raise my GPA. But, and this is the reason I like this system, these tricks meant I was exposed to themes and ideas that I would not have been if I had strictly focussed on biology. I discovered that I really like how computers work. I enjoyed learning who the Zulus are and why they were able to be such a thorn in the British side. But I really loved the mechanics of movies: how a writer created characters and how a cinematographer put a picture on a frame of film and how a director put it all together.

Now, anyone who knows me knows I just love movies. I have seen some really awful ones (Good Burger starring Kenan and Kel being my burnt into my quivering brain forever) but most are just pedestrian. Sturgeon's Law states, in part, that "90% of everything is crud." Seems about right for movies. Some cruddy movies are even enjoyable. But these are seldom designed to make you think, to examine the movie and what it means, how it relates to life. To be able to write a great BS term paper on. I mean, no one is going to write a paper comparing The Grinch and X-Men as a reflection of a capitalist society in a post-WTO world!!

But every once in a while something comes along that jiggles those brain cells a little. Those are the movies I watch again and again. Like "The Seven Samurai". Or "His Girl Friday". Or "Blazing Saddles". Or "The Searchers". Or "The Matrix". I saw "Unbreakable" last week and just adore it. I would not expect it to be widely embraced but it is the best of its kind (Of course, the argument is what kind is it!). My wife and I spent 2 hours at dinner going over so many aspects of this movie. I could write such a great paper on this movie. Because everything just fits when it is looked at a certain way. Like his previous work, "The Sixth Sense", we spend most of the movie never quite sure where it is going or how it will be resolved. It is only at the end that clarity is restored. The foggy glass through which we have been looking clears up.

Now, what does this have to do with the column's topic? Heck, I don't know. I just write and somehow it all comes together. What I want to discuss are some recent insights into chromosomal structure and rearrangements as it relates to gene expression.

I sometimes feel many people just ignore chromosomes. Genes are just seen as strings on a bead. Polymerase just comes along and does its job. But what makes chromosomes chromatin are the proteins. The DNA is just packed with protein, such as histones, transcription factors, etc .These help protect the DNA or help the chromosomes maintain higher level structure, but their presence has always presented somewhat of a dilemma.

There are other questions. The only time it is easy to see most chromosomes is during specific times of cell division, when they condense into distinct bodies. The interphase chromosome is diffuse and very difficult to visualize. Well, new tools and recombinant DNA are helping to clear up the picture. Chromosomes undergo a lot of physical change in response to external perturbations. Histones are modified to allow access of the transcription machinery to the DNA. But even larger changes are now being seen, thanks to some novel techniques. The foggy image of an interphase chromosome is becoming clearer. (See, there it is. I knew my subconscious would find someway to connect the rambling of the first few paragraphs into this week's subject.)

Green fluorescent protein (GFP) has generated so many spin offs. There are mutants that fluoresce at different wavelengths, giving us cyan fluorescent protein. In fact, we have so many colors that we should be able to develop a living wall in color.

A recent paper in Nature Cell Biology describes the use of one of these proteins to examine gene activity in single, living cells. A stable cell line was created containing an 18.5 kilobase plasmid, or actually multiple copies ranging from 10 to 1000 per haploid genome. Each plasmid had 256 copies of the lac operator, 96 copies of a tetracycline-responsive element controlling a promoter expressing cyan fluorescent protein (CFP). The location of the lac operator region could be visualized by transfection with a EYFP/lac repressor fusion to look at in vivo expression or by adding exogenous EGFP/lac repressor to fixed cells. Transfection of another plasmid, pTet-On, results in protein production, which is this case was CFP, easily imaged in a microscope.

What they saw was a specific spot in the nucleus that contained the transfected gene. When CFP expression is induced, there is a noticeable decondensation of the chromatin. This would begin to occur 30 minutes after the inducer was added and extend out to 4 hours. In a great set of photos, they examined the position of the transfected locus over time along with the presence of protein expressed from the locus, – in a single cell! You can see the decondensation of the chromosomes at 30 minutes and the appearance of protein at 3 hours.

Pretty neat to see the effect on gene expression in a living cell. This approach presents a lot of possibilities for examining other processes. Another recent paper in Molecular Cell uses GFP as a marker of chromosome structure. Chromatin insulators are regions that almost act as anti-enhancers. They can interfere with the ability of an enhancer to activate transcription. But they also can act to buffer some chromosomal position effects. That is, the ability of a gene to be transcribed at all can depend on its position in the chromosome.

There are a group of proteins that interact with chromatin insulators. Using a GFP fusion with one of these proteins, it was possible to visualize in a single nucleus just where these insulator sequences were located. Surprisingly, these locations were all at the periphery of the nucleus. Even though the insulator sequences are found throughout the chromosome, they come together in discrete locations. Here is a proposed model of this structure. This suggests that these insulator sequences may be very important in overall chromosome structure, particularly of non-transcribed genes.

What is interesting is that following heat shock, in which there is a tremendous change in which genes are turned on or turned off, these insulator bodies break down and the sequences are moved towards the center of the nucleus. So changes that reform transcription events in the cell also transform the structural effects of these insulators.

A lot of data is starting to come together regarding the effect of chromosome structure on expression. These new tools are allowing us to see the remodeling that takes place when transcription occurs. Transcription is controlled by machines containing polymerases and other proteins. But it is probable that these machines do not float free throughout the nucleus. They are localized to specific regions. Genes that are to be transcribed are brought to them. So, the purpose of enhancers may not be to directly increase transcription. Maybe they only increase the chance that a sequence will be moved towards one of these transcription machines (see Lies and Statistics). The more enhancers bound by factors, the greater the probability.

This could also explain some of the theories about heterochromatin. (see Book, Worm). That column discussed the possibility that heterochromatin is the site for new genes. They are expressed at low levels and, if they are useful, eventually move to euchromatin and are expressed in high levels. This last part was hard to explain. How does something "move" from heterochromatin to euchromatin. An actual translocation seems unlikely.

But, what if something like enhancers can move around in the nucleus on retroposons (see The Intro(n) and the Outro(n) and Genomic Reformer)? More of these elements will increase the likelihood of transcription by moving the mountain to Mohammed. And it appears that that this act might be the distinguishing difference between euchromatin which is diffuse and heterochromatin which is condensed. Move the genes to the transcriptional apparatus and the chromosome has to remodel itself, becoming more diffuse. Keep them at the periphery and they remain condensed. Perhaps the main difference between euchromatin and heterochromatin is the presence of enhancer sequences.

Makes for a nice model for gene expression and the evolution of novel genes. It would suggest that expression of recombinant genes would be greatly enhanced (pun intended) by the proper placement of the gene, preferably near some important enhancers. Otherwise, it would be more likely to be heterochromatic and not expressed highly.

Boy, did I link together a lot of my earlier columns!! Hope it all makes some kind of sense. I like it since it explains why we have hetero- and euchromatin, and how one might become the other. This view of dynamic chromatin, moving about to new locations or reforming and modifying itself to respond to conditions, has mainly come about because of the green fluorescent glow of a modified protein from a jellyfish through the dark glass of cell biology. I guess that comparative religion class did come in handy (See 1Corintians 13:12).