MOLECULAR BIOLOGY may be a cut-throat sort of discipline, but its
practitioners are usually polite. Usually. Last March, even etiquette went out
of the window when a few determined researchers staged a multi-fronted attack on
one of the most established ideas of how genetic information is discharged from
the nucleus to fulfil its mission in the cell. Before then, most people had
thought that the cellular mechanism that activates genes, and the device at the
heart of it鈥擯ol II (for RNA polymerase II)鈥攚as so well understood it
could just about be relegated to textbook dogma.
The first assault came from yeast geneticists, Kevin Struhl of Harvard
University in Cambridge, Massachusetts, and Michael Green of the University of
Massachusetts Medical Center in Worcester. They had independently come up with
evidence that a group of proteins called TAFs, which had long been thought
essential for Pol II to function, were nothing of the sort. In fact, they said,
Pol II can easily release most genetic information from the limbo of the nucleus
in the total absence of TAFs. To some, that was like claiming it was possible to
play the piano without keys.
Freak out
Advertisement
鈥淭here was some serious freaking out going on.鈥� says Struhl, when the results
were presented at an elite meeting in Taos, New Mexico, in March last year.
A second assault came from Richard (Rick) Young of the Whitehead Institute
for Biomedical Research in Cambridge, Massachusetts. That March, he was pushing
his theory that Pol II was a mere cog in a still larger piece of apparatus (see
Nature, vol 380, p 82). According to Young, a huge complex of tens or
even hundreds of proteins (and not one of them a TAF) controlled Pol II by
combining with it to form a vast nuclear apparatus. He called the whole caboodle
the Pol II holoenzyme. And while most experts on gene activation believed that
the relatively diminutive Pol II complex was freshly assembled each time a gene
needed to be turned up, Young had the audacity to insist that his data showed
that the massive holoenzyme comes fully assembled. To Young, the Pol II complex
was like a molecular factory ship, floating inside the nucleus until called into
action.
Unless a gene is activated, the information contained in its DNA sequence
remains locked for ever in the secure environment of the nucleus. Understanding
how that information is set free is key to understanding fundamental biology
like how animals grow and develop from a single fertilised egg, or how some
genetic defects trigger disease. Pol II is the device that reads the original
DNA sequence and makes the RNA copy that is free to roam. Once the RNA molecules
have crossed through the nuclear membrane into the cell鈥檚 cytoplasm, they might
become part of a vital cell structure or, more often, move on to the protein
production lines where their genetic sequence provides the blueprint for every
type of protein from hormones to muscle fibres.
Pol II was first identified in 1969, so one might have been forgiven for
assuming that the exact way it worked was understood long ago. But, in fact, Pol
II is a very complex beast made up of around 12 protein parts. Some components
unwind the DNA double helix, some read the DNA sequence, and some select and
connect the correct RNA building blocks, or provide energy to keep the whole
thing running. And that鈥檚 just the tail end of the process.
Before Pol II can get to work, a molecular 鈥渟tarting gate鈥� lines it up with
the gene鈥檚 core promoter, a region of DNA close to where production of RNA
begins. This starting gate is constructed from proteins with names like TFIIA,
TFIIB, and the TAFs. Pol II comes under starter鈥檚 orders when it binds to the
starting gate. It is given the 鈥済o鈥� signal by activators, chemical messages sent
from various parts of the cell, including the cell membrane and so indirectly
from the outside of the cell. The activators bind first to a section at the
front of the target gene called the proximal promoter, which releases the
starting gate, setting Pol II off and securing the production of more of the
gene鈥檚 messenger RNA.
Meal time
After a good meal, for example, Pol II in the cells of the pancreas is
directed to make more of the RNA that codes for insulin, to ensure that there鈥檚
enough of the hormone ready for the next meal. Until Struhl and Green made their
announcement, most self-respecting molecular biologists would have gambled their
research grants on the idea that it was the TAFs that lifted the starting gates
at the activators鈥� command.
Historically, it was the biochemists who took the lead in studying gene
activation. They ground up cells, separated the different proteins and put them
in various combinations into test tubes along with strands of DNA. They were
able to get their mixtures to produce RNA, and, by a process of elimination, to
work out which proteins are essential to make RNA from DNA.
One of the pioneers of this approach is Robert Tjian of the University of
California in Berkeley, known as 鈥淭eej鈥� on the circuit. Since the 1980s, Tjian鈥檚
group on the West Coast, and Robert Roeder鈥檚 group at Rockefeller University in
New York, have between them identified most of the components that make up the
so-called 鈥減re-initiation complex鈥濃�攖hat is Pol II and its starting
gate鈥攁s well as numerous activator proteins which ensure that Pol II turns
on particular genes under particular circumstances.
Tjian, Roeder, and others had shown that around half of the two dozen or so
proteins that make up the Pol II starting gate are TAFs. The TAFs associate with
another protein called TBP that binds to the gene at the core promoter. Tjian
built a cottage industry around the TAFs, and although he claims it was never
his idea, it became accepted wisdom that activators release the Pol II starter
gate, and so activate genes, via the TAFs. 鈥淧eople had started to settle down
and believe that this was the whole story,鈥� says Steve Jackson of Cambridge
University, a biochemist who used to work with Tjian. 鈥淧eople thought they knew
how [gene activation] worked.鈥�
The biochemists had studied the process by trying to build the machinery that
switches on genes from its component parts. The geneticists took a different
tack. They deliberately smashed bits of the machinery and then looked to see
what happened. To do this, they bred mutant strains of the yeast
Saccharomyces cerevisiae. Because no one ever found yeast mutants with
completely defective TAF genes, most of the geneticists had also assumed that
TAFs are essential for life. Presumably without TAFs, Pol II never got to leave
the starting gate and the genetic information remained frozen in the
nucleus.
Struhl and Green quashed that theory with the surprise discovery that in a
new type of yeast mutant, the loss of any one of six TAFs had no effect on Pol
II鈥檚 ability to make fresh RNA, even though the yeast cells did eventually die.
Rather like the fabled fingernails on a corpse, the Pol II continued to make RNA
even when the cells were technically dead. The two geneticists concluded that
TAFs are essential for life, but not for RNA production.
鈥淵ou can wait a long time after the cells are dead and the activation will
still work,鈥� says Struhl. Yeast cells that have been 鈥渄ead鈥�, that is unable to
multiply, continue to activate their genes and make RNA for as long as eight
hours.
So astounding was that finding that Struhl and Green couldn鈥檛 resist
presenting it in a preliminary form at the Keystone Symposium in Taos. It
created 鈥渕ore than a furore,鈥� remembers Tjian. 鈥淸They] presented preliminary
data, and it was interpreted in the most extreme way. Like, TAFs are not
important at all; they鈥檙e irrelevant, an artefact.鈥� Struhl now concurs that the
original data were incomplete and open to different interpretations.
Making waves
The finding was, however, supported by some earlier experiments in yeast that
had been largely ignored. Teams led by Young and Roger Kornberg of Stanford
University in Palo Alto, had already managed to activate genes, and make RNA
from DNA by combining yeast proteins that did not include TAFs. Kornberg, a
biochemist, was understandably chuffed by Struhl and Green鈥檚 announcement at
Taos. 鈥淭he TAFs failed the genetic test with flying colours,鈥� he says.
Another wave of Pol II-bashing broke when Struhl and Green eventually
published their findings in the journal Nature in September last year
(vol 383, p 185 and 188). Their paper was accompanied by a 鈥淣ews and Views鈥�
article co-written by Young which virtually declared that the textbook version
of Pol II鈥檚 role needed revision. Young argued that the activators whipped up
the speed of Pol II, not by binding to TAFs, but via the components of the Pol
II holoenzyme. Nor was Young showing any signs of backing down on his theory
that the holoenzyme came ready assembled, simply needing to be swung into place
to activate a gene鈥� indeed at least two other research teams had by then
published evidence that backed him up.
But in the past few months a calm has fallen upon the Pol II battlefield. A
small, closed meeting held at Cold Spring Harbor Laboratories on Long Island
just before Christmas, was set up specifically to thrash out gene activation
controversies. It seems to have worked, at least in part. 鈥淎 remarkable meeting
. . . a very savvy group,鈥� says Green.
Green鈥檚 team has now identified 鈥渁 number of [yeast] genes鈥� that need
TAFs鈥攁ll of which help to control cell division. 鈥淭AFs are very important,
I wouldn鈥檛 have half-a-dozen people working on them in my lab if I thought they
weren鈥檛,鈥� he now says.
Armistice
And so a new game plan for how Pol II turns on genes is emerging. Given that
there are roughly 100 000 genes in the cell of a mammal, and that keeping an
organism running requires that each be precisely regulated鈥攌ept silent,
kept ticking over, or revved up to screaming point鈥攖he new plan has the
activators operating a panoply of different switches on Pol II鈥檚 starting
gate.
Only one of those switches鈥攑ossibly the one that controls certain
cell division genes鈥攊s made up of TAFs. 鈥淭he polymerase is going to be
tailor-made for each gene,鈥� Tjian predicts. 鈥淭AFs clearly are required for some
transcription, but they have selective effects,鈥� agrees Struhl. 鈥淎ctivators can
work through TAFs, or through components of the holoenzyme.鈥�
And how about some agreement on whether Pol II and all its assorted
attachments come preassembled, or are made afresh each time a gene needs
activating? All the signs are that there will be more hostilities before there鈥檚
an armistice.
