Feature: Functional genomics writ large

A forensic analysis of the cellular train wreck of a human tumour will usually detect between 200 and 600 mutation-mangled genes. It’s only two decades since cancer researchers thought half a dozen mutation 'hits' were sufficient to tip a call into cancerous growth.

Dr Dubravka (Duka) Skalamera says cancer is now known to be the result of a much more complex process of genomic degradation. “An average tumour contains around 2000 mutations,” she says. “Up to 600 of these will lie within genes, while the rest are in non-protein coding regions.”

Skalamera is the manager of Australia’s most sophisticated high-throughput facility for exploring gene function, the Arrayed Retroviral Expression Cloning (ARVEC) facility at the University of Queensland Diamantina Institute within the Princess Alexandra Hospital in Brisbane.

The roll-call of mutant or dysregulated genes typically includes the usual, high-profile suspects: oncogenes, tumour-suppressors, transcription factors, protein kinases and signal receptors. Others have poorly defined or multiple cellular roles and some are completely unknown.

Many will be uninvolved in cancer, casualties of collateral damage in the random processes of tumorigenesis. But others will be unrecognised contributors to cancer, that have been silenced or constitutively activated by random mutation. “There are around 23,000 genes in the human genome. We know the precise function of fewer than 3000 of them,” Skalamera said.

“While the function of some of the rest can be inferred by similarity to genes in other species, thousands of human genes still remain only electronically annotated and identified by an alphanumeric code. Our limited understanding of gene function is compounded by the fact that one gene can have multiple roles in different types of cells.”

The challenge of distinguishing between the malefactors and the blameless in the more than 60 per cent of genes of unknown function is formidable. “Now we have all the sequence information from the Human Genome Project, we’re beginning to realise how complex the human genome is,” Skalamera says. “If you hope to assign functions to all those genes, you need a very high-throughput system.”

---PB---

Cell wells

“With the ARVEC facility, the idea was to do everything in micro-well plates, then add robotics to that, so we could pipette into tens of thousands of wells, and do thousands of experiments in parallel every week.

“That requires very expensive equipment, access to expensive gene libraries, vector systems to deliver the genes into cells, and people with the specialised knowledge and skills to help researchers do their experiments.”

Skalamera says ARVEC staff have the expertise to provide guidance in all aspects of screening, from project design and assay development, to imaging and bioinformatics. The facility operates on a collaborative, partial cost-recovery basis.

“It’s one thing to do a single experiment, sitting alone at a microscope; it’s another to do thousands of experiments in a week, involving millions of cells, and record and interpret the results,” she says.

According to Skalamera, the basic approach involves growing the cells in arrays of microplates containing 96 or 384 wells and “throwing agents on them that will either knock out a single gene, or increase its expression”, replicating the phenotypic effects of loss-of-function or gain-of-function mutations.

Gene-knockout experiments can be performed either with short interfering RNAs (siRNAs) or short hairpin RNAs (shRNAs). The siRNAs come from a commercial Dharmacon Smartpool siRNA library, marketed by Thermo Scientific, while the shRNA library is an Open Biosystems product.

“For over-expression experiments, we created our own lentiviral vector library, which is unique in Australia,” Skalamera says. “The international research community is also interested in what we’ve done.”

From Open Biosystems and Imagenes, the ARVEC centre purchased 17,100 open reading frames (ORFs) derived from the Mammalian Gene Collection, representing almost every currently available human gene. Researchers used Invitrogen’s proprietary Gateway insertion technology to transfer the cDNAs into a lentivirus vector.

Commonly used retroviral vectors are only able to transduce actively dividing cells. ARVEC’s lentivirus vector has the ability to transduce non-dividing cell lines and integrate its packaged transgene into the host cell’s genome.

Skalamera says this allows researchers to work with differentiated cells from biologically relevant tissues – both normal, and cancerous. A robotic microscope collects images of cells in multiple single colour channels in each well. Each colour tracks a different cell phenotype.