Back to Index

Chris Goodnow, Facility Director

Australian Phenomics Facility (APF) The John Curtin School is the lead participant in establishing a new Major National Research Facility, The Australian Phenomics Facility, aimed at elucidating how our body's health, function and form – our phenome – is encoded in our DNA genome. Many of the health problems we face today – cancer, autoimmune diseases, allergy, cardiovascular disease, osteoporosis – stem from a discordance between genetics and environment. Our genetic code was selected to survive for a shorter period of time in a very different world, and it is inevitably out of step with the lifestyles we lead today. In some cases we know what to do to fix this imbalance, like putting on a hat and sunscreen to offset genetic susceptibility to the sun. For many disease, however, the solutions are not so straightforward. Rapid advances in gene technology are bringing us to the point of visualizing the imbalance between our genetic code and lifestyle-environment. A worldwide effort has already determined the spelling of the thirty-five thousand genes in our genetic 'dictionary' collectively referred to as our genome. The epic challenge for the next decade is to decipher what these gene words mean. How are they combined into the molecular language that guides the cells in our body and determines how well we cope with different environmental stresses? Which gene products are good targets for new drugs to prevent or cure common diseases? Differences in the spelling of those gene words underlie differences in susceptibility to modern diseases and divergent responses to particular therapies, and gene-fingerprinting technology now allows a list of spelling differences to be compiled from each person's unique genetic dictionary. This great opportunity to advance health care hinges on establishing the function of genes and how they interact.

The appearance and performance of our body – our phenome – is shaped not by individual genes acting in isolation, but by interacting combinations comprising hundreds of different gene products. Moreover, each gene typically serves multiple functions by engaging in different combinations. Subtle changes in individual genes or their control elements can produce radically different effects on the phenome than complete ablation of the gene, as illustrated by two publications from users of the APF this year (P Papathanasiou et al Immunity 2003; J Jun et al Immunity 2003).

The John Curtin School is the lead participant in forming the Australian Phenomics Facility, catalysed by an AUD$11.5M capital grant from the Commonwealth's Major National Research Facilities. The new national facility grows out of the School's Medical Genome Centre, which has been pioneering technologies to produce libraries of mice with informative gene alterations for the last five years, and ways to search those libraries for new genome-phenome connections. School staff serve many key roles to help establish the new facility. The APF combines the ANU staff and expertise with complementary experts at Monash University in the Monash Institute of Reproduction and Development, at the Garvan Institute, and the Institute for Molecular Bioscience at the University of Queensland.

Design and Construction of new facility

The foremost challenge in scaling up the activities of the Medical Genome Centre into a National Facility is the provision of new facilities to house expanded libraries of mice, phenotyping and mapping technologies, infrastructure to manage data and maintain frozen sperm, and space for researchers to access all of these resources.

Construction of the new building has been led by Andrew Kemp and Simon Butt of Manteena Pty Ltd, in close cooperation with Chris Culverwell and the Division of Facilities and Services of the ANU. All trade and services packages have now been let, and the building is on track and on budget for completion and occupancy in April 2004. This is a remarkable achievement so far, particularly when faced with very overheated building demand in Canberra fuelled by the January 2003 fires and a bullish property market.

Development of database infrastructure

Musterer, the program developed by Greg Quinn to record, track and organize all information about pedigrees of mice in the APF libraries and phenome bank, continues to acquire new enhancements. A major effort has successfully yielded systems for connecting samples from individual mice in the library, such as DNA or blood, with their results, with personnel performing the tests, and with the other information on the mice. At the core of this critical development is an interface to Musterer developed by Greg Quinn that allows barcode scanners to record the precise identity of each mouse and the corresponding sample in a 96-well plate, as the sample is collected. A separate set of barcodes register each plate of samples and the subplate replicas that are tested by PCR, ELISA, or flow cytometry. These plate barcodes interface with recording equipment such as the flow cytometer. Shirine Chaudhry, Katherine Sullivan and their team have worked through logistics and training of staff to come up with a successful solution to this challenge in the holding rooms. Michelle Townsend, Aisling Murtagh, Suzanne Ewing and their team established plate-based genotyping and sample recording. Geoff Osborne and Sara Dawson worked closely with Greg to develop the subplate concepts and implement the connection with the flow cytometer equipment. Edyta Kucharska, Heather Domaschenz, and Carola Garcia de Vinuesa have developed parallel systems for ELISA data.

A time-consuming problem for all staff and users to date has been the job of manually drawing family trees to visualize complex breeding relationships and inheritance of related phenotypes/genotypes. In the last year, Adrian Gee developed an elegant addition to Musterer that instantly displays a customized three-generation genogram chart for any mouse or pedigree of interest. Each individual in the genogram is marked with up-to four different genotypes or observations selected by the user. Solving this data visualization problem involved considerable consultation with staff, and innovative coding, in order to fulfil the diverse needs of users. All that have clicked the new tree icon button on the Musterer have made the same response when the data comes up instantly. "Wow!"

With the advent of serving as a Major National Research Facility, a key challenge lies in planning and coordinating many parallel projects using the same library of mice. Carola Garcia de Vinuesa, Heather Domaschenz, Edyta Kucharska and Gerard Hoyne have worked intensely with Greg Quinn to develop tools to coordinate different screens and activities on parallel pedigrees and screens in serial generations. These have initially been piloted in Excel, and successful solutions then written into Musterer. A key tool resulting from this effort is a Musterer interface showing which animals are due for which tests, and which have already been tested. We anticipate these tools will continue to evolve as the user base grows and we build on unique experience and ideas across the facility network.

APF Mutant Libraries

The first year of operations as a Major National Research Facility saw our first experience coordinating five parallel screens on large shared libraries of mice. These screens were diverse: a Wellcome Trust program between the John Curtin School and Oxford University searching for new immunology genes and models; a Juvenile Diabetes Research Foundation/NH&MRC program seeking genes and pathways in diabetes; a Monash Institute of Reproduction and Development screen of sperm and DNA for genes controlling male fertility; a Murdoch Children's Research Institute screen for genes and models for understanding hearing and deafness; and a screen for regulators of blood, metabolism and obesity by Phenomix Aust Pty Ltd. Each of these efforts has been successful, with new heritable sublines proceeding to detailed phenotype analysis and mapping.

Aside from delivering specific results for each of these projects, a number of important general advances have been driven by this experience. The new libraries and screens involved complex genetic sensitisation with transgenes that could not be fixed in homozygous state. Consequently, hundreds of genotypes needed to be determined each week. Efficient 96-well plate procedures have been developed and implemented to test genotypes by PCR on small skin samples, or on residual cells from blood screens. The use of challenge screens has led us to evolve strategies to enhance the specificity of screens to differentiate true-breeding heritable changes from background variability, such as modification to pedigree structures and family sizes. Procedures for scheduling sampling, challenges, testing, breeding, and culling have been developed and refined.

The growth of these libraries and users needing access to them has placed the current facilities under tremendous pressure. The Facility Manager Adrienne McKenzie, Colony Coordinator Katherine Sullivan, and Phenotyping and Mapping Coordinator Shirine Chaudhry, and their team, have achieved the impossible by delivering these year-one achievements in the absence of the necessary space and facilities. We look forward to moving to the new building.

Phenotype and Mapping Resources

Each of the participants in the APF have continued to develop expertise in comparative analysis of key mammalian processes and phenotypes. In the John Curtin School at the ANU Carola Garcia de Vinuesa, Edyta Kucharska, Heather Domaschenz, Sara Dawson and a team of colleagues have developed and validated new screens to interrogate a wide range of processes in the immune system, ranging from flow cytometric subsets in blood, immunohistologic tests, and autoantibodies, to different types of T cell and antibody responses. The screens have been validated by isolating a large set of new strains with immune response abnormalities or autoantibodies, and by applying them to strains that form the nucleus of the Phenome Bank (eg Miosge et al J Exp Med 2002; J Jun et al Immunity 2003).

The mapping resource has also improved strategies for the fine mapping stage of identifying novel mutations, which reduces the initial chromosomal interval containing the mutation from 5-10 cM (~10-20 Mb of DNA) down to a target of less than 1Mb. A limiting step is availability of polymorphic markers at this density. Belinda Whittle and Connie Angelucci have evaluated two solutions in parallel. One employs single nucleotide polymorphisms (SNPs) listed in the Ensembl database as confirmed between C57BL/6 and three other strains (129/Sv, C3H and Balb/c). These have been tested by amplifying the surrounding 200bp and resequencing. In the case of C3H vs B6, ~80% of these SNPs are also informative in B6xCBA crosses. In the case of Balb/c vs B6 SNPs, ~50% are also informative in B6 x NOD crosses. The other strategy employs simple dinucleotide or trinucleotide repeats annotated in the Ensembl database. Approximately 50% of these have typable length polymorphisms between the mapping strains. Between the two methods, marker density is not likely to be limiting for future mapping by users. We favour the SSLP markers in the first instance because they are less work and expense to type, and will be less susceptible to the "SNP deserts" that exist in the common mouse strains.

Phenome Bank

Gabriel Sanchez-Partida, Claire Kennedy and their Monash team have evaluated and established more efficient methods for freezing and storing sperm from large numbers of mice.  A key development has been the establishment of snap freezing in pellets on dry ice, and storing multiple replica pellets in a single cryovial.  The team has validated the method in freezing up to 500 mice per week.  Methods for shipping large numbers of samples between Canberrra and Monash sites have also been established.  Optimal systems for thawing B6 sperm for IVF or ICSI, and suitable inventory systems are being evaluated and this continues to be a major focus.

A formidable collection of mutant strains has already been assembled, and started to be accessed by users. These strains are currently stored frozen and maintained as viable breeding nuclei.  A key goal for the next year will be to rationalize the number of strains maintained as breeding animals.

Current Phenome Bank collection

Facility Staff

Phenomix Australia Pty Ltd