Don't Throw Out Those Introns

A somewhat accidental discovery and random meetings between proteins in a cell: These are the subjects of two new online articles. Each, it its way, involves a technological advance that will, in turn, lead to further scientific discovery.

The first involves a partnership between a physics group and a cancer-research group. Among other things, such collaboration is essential for dealing with large data sets - multiple gene expression patterns, for example. When the team made their discovery, they were looking not just at gene expression, but at pieces of genes. More precisely, they were investigating exons - bits of genetic code that are snipped out of the sequence and spliced together to make the protein instruction list - the messenger RNA (mRNA). For this they had a newfangled Affymetrix exon array that can probe both exons and introns - the lengths of gene code that get left out (and for that reason, are generally left out of the research data, as well).

Production overshoot (marked in green, left): gene expression profiles without and with the introns

At some point, the students conducting the research had the brilliant idea of holding on to the intron expression data, thinking they might come in handy as markers. And that, in their own words, is when they noticed "something very weird." Many expression profiles of the introns did not fit with those of the exons from the same gene. That finding was strange enough for them to go back and check it properly with additional arrays as well as more traditional sequencing methods.

What was going on? Introns only appear in the pre-mRNA - the fresh copy of the gene's full code before it is chopped up and spliced into mRNA. So the intron profiles they were seeing showed pre-mRNA production; the exon-only profiles were of the mRNA. And, in some cases, pre-mRNA production shot straight up - to ten times or more than that of the mRNA that followed. Why would the cell make this much excess pre-mRNA? The scientists think that the "production overshoot," as they call it, occurs when the cell needs a rush job on the manufacture of certain proteins. Like stepping on the gas to merge into fast traffic, it throws a lot of fuel into the first few minutes before easing up. When everything is up to speed, it apparently falls to the mechanisms for degrading mRNA to regulate the amounts.

The second study began with a question: Are protein interactions slower in the cell than they are in a test tube? Intuition said they should be quite a bit slower: Two proteins in a test tube solution have no problem meeting up, but the interior of a cell is a densely crowded place, with thousands of other molecules all jostling one another.

Using FRET technology, Weizmann scientists managed, for the first time, to create movies of protein interactions inside a living cell and to measure them. Something like the laser show at a rock concert, they coordinated flashy lighting effects to the action on stage: When the proteins interacted, one of each pair transferred energy to the second, causing its genetically-engineered fluorescent green glow to turn red.

Here the surprise finding was that in the live cells, these protein-protein interactions come about at a similar rate to that in a test tube. The scientists' explanation is that the crowding does slow the proteins down at first, but actually jostles them into meeting once they get close to one another. These proteins don't automatically recognize one another - they can bump into each other many times before an interaction takes place. Bouncing back and forth in the bustling mass of molecules increases the number of chance encounters between the two proteins, and thus the odds of interaction.

The original use of existing technology in the one and the new technique - the tweak to the technology - in the other are already inspiring new research. The physics-cancer research team, for instance, have already used their method to check gene expression in different situations and cell types, and they are planning to adapt it to mapping the protein production peaks to see if their ideas hold up, or whether more surprises are in store. And while the second finding actually tends to support the use of the test-tube studies for investigating protein-protein interactions, it may well lead to a wealth of new and interesting experiments in which molecular actions are imaged inside living cells.