Revolution in genetics tipped as brave new world of RNA revealed

A new branch of genetics is being born. It might legitimately be called junk science, but Prof John Mattick has dubbed it 'Rnomics'... and it's the stuff of revolution.

All the so-called 'junk' DNA littering the human genome, and the genomes of all living organisms that aren't bacteria or viruses, suddenly assumes enormous significance, Mattick told the 28th annual Conference on Protein Structure and Function in Lorne last week.

Mattick, director of the Australian Genome Research Facility in Brisbane, now has abundant evidence that junk DNA actually encodes myriad RNA sequences that are never translated into proteins. Rather, they interact with each other, with the DNA of genes, and with the messenger RNAs that instruct protein synthesis.

Collectively, they comprise an intricate, multi-layered operating system for the collection of proteins in the cell -- the proteome. The implications of this dawning realisation, says Mattick, are enormous.

His landmark talk at the Lorne Conference, 'Programming of complex organisms: the hidden layer of non-coding RNA', proposed that junk DNA is actually the genome's 'ghost in the machine' -- a self-organising operating system that ultimately determines how, when and where genes shall work, as well as coordinating their interactions.

He says the first inkling of the existence of the system came to him a decade ago, when he was considering the implications of introns -- the large stretches of non-coding DNA that separate a gene's protein-coding sequences into modules.

If introns are junk DNA, why are intron sequences often more highly conserved than the protein-coding sequences of the genes in which they are embedded? Evolution only conserves DNA sequences that are functionally important.

The sheer mass of non-coding DNA in eukaryote genomes -- it comprises about 98 per cent of the human genome -- demanded explanation. "It took me 10 years, but it got me thinking," says Mattick.

"Non-coding RNA sequences are being produced in parallel to the protein coding sequences of messenger RNAs. The only possible function they could have is in networking, and communications, in real time, with genetic activity.

"It has interesting resonances with the sort of information transactions taking place in other complex systems, like computers and the human brain."

Biology's biggest mistake

Mattick says he thinks the failure of biologists to "consider the massive amounts of transcription involving non-coding DNA in the genome will be regarded as perhaps the biggest mistake in the history of biology.

"Every cell in the nematode worm C. elegans has a defined fate, and it's probably true for human cells, with the exception of cells that respond to immunological challenge or exercise," he says.

The RNA-nome, or simply 'Rnome', is probably also responsible for the amazing fidelity of the processes of differentiation and development, he believes.

Mattick says that it has always been assumed that the coordinated activity of transcription factors, acting through gene promoters and enhancers to produce proteins, is all that is required to coordinate the enormously complex processes of cell differentiation and the development of complex organs and structures.

But he now believes the proteome is merely the visible machinery of a far more complex system -- that the structure and function of proteome are actually specified by non-coding RNAs.

"If you had to assemble a Boeing 747, with tens of thousands of different components, you need massive amounts of information to put them together," he said.

"What we've been doing is studying the Boeing 747, and working out what the individual components do, but we have had very little information about what is required to put them together to make them work as an integrated whole."

Genome conundrum

One of the conundrums to emerge from the Human Genome Project is that the human genome, once thought to comprise at least 100,000 genes, actually contains a mere 40,000-odd genes.

Geneticists have explained the economy of the human, and other eukaryotic genome, in terms of the fact that most genes have multiple protein 'recipes' created by shuffling and resplicing of their exons, or protein-coding modules. The shuffling occurs during transcription -- it is messenger RNAs that are shuffled and spliced, not the genes themselves.

But the nature of the process that coordinates all this RNA splicing activity has been a mystery. Mattick is convinced that an enormous, multi-layered system of non-coding RNAs, embedded within introns and the non-coding tracts of DNA between genes, coordinates RNA splicing and many other fundamental aspects of gene activity.

Non-coding RNAs also appear to be involved in regulating gene function. At one level, non-coding RNA molecules appear to be an essential accessory in the formation of protein-DNA complexes in promoter sequences -- the DNA 'switches' that regulate gene activity.

At the other end of the process, 'antisense' non-coding DNA may form double-stranded complexes with messenger RNAs, to prevent them being 'read' by ribosomes -- the genome invented antisense gene silencing long before genetic engineers conceived the idea.

Most non-coding DNA, far from being ancient evolutionary junk, is probably a evolutionary innovation in eukaryotic cells -- a way of diversifying and coordinating the activity of a limited repertoire of genes, and facilitating the evolution of increasingly complex organisms.

Information theory predicts that, as the complexity of the genetic machinery increases, the complexity of the system that assembles and coordinates it must increase geometrically.

And that, says Mattick, probably explains why more than 90 per cent of the DNA of a complex organism like a human being does not encode genes. But nor is junk -- it encodes the RNA operating system.

Mattick believes small, non-coding RNAs may ultimately be accountable for most basic genetic processes, including the manipulating chromatin in readiness for gene activity, or gene replication during meiosis and mitosis. It probably controls so-called 'imprinting' processes like methylation, that determine which genes are switched on or off in particular tissues.

"There's a whole new RNA world, appearing before our eyes," says Mattick. "The implications are enormous."

He says genetic engineers must understand, and learn how to manipulate these non-coding RNA systems, as well as the genes themselves, if they wish to perform truly sophisticated gene surgery on plants and animals.

Non-coding regions of genes have been used as convenient locations to insert transgenes -- could the disruption of vital non-coding RNA sequences during gene insertion explain why some gene transfers fail, or don't work the way they should?

Mattick says that in addition to well-defined, single-gene disorders affecting protein function, there are likely to be huge number of subtle disorders that affect the non-coding RNA machinery -- the 98 per cent of the genome that does not code for genes presents a barn-door target for mutation.

For similar reasons, the chief source of individuality in living organisms may be subtle variations in the non-coding system that coordinates gene function, rather than small variations, called single nucleotide polymorphisms, in the genes themselves.

Mattick suspects that the way in which evolution has created such enormous complexity from such a compact set of genetic code holds lessons for engineers and computer programmers seeking to design compact systems or software that will give rise to complex, emergent behaviour.
-- Graeme O'Neill will report on the latest developments at the Lorne Protein Conference all week