human genome project and the future of diagnostics treatment and pre - PDF document

human genome project and the future of diagnostics, treatment, and prevention LANCET J B van Ommen, E Bakker, J T den Dunnen the past 5 years, the Human Genome Project has had a tremendous influence on the field of genetics. This influence will soon become extended across the whole of biology and medicine. After centuries of description, we are now on the eve of truly understanding the cellular processes in the human body. Up until the late 1980s, tracing the causes of genetic disease involved time- consuming linkage analysis in families, by limited sets of genetic markers. At the end of that decade a more comprehensive approach was tabled--the Human future of diagnostics first area benefiting directly from the current developments is that of diagnostics. During the past decade, we have already seen a large improvement in diagnostic resolution: 354 (suppl I): 5-10 Human and Clinical Genetics, Leiden University Medical Center, Leiden, Netherlands G J B van Ommen PhD, E Bakker PhD, J T den Dunnen PhD) to: G J B van Incidence Gene Mutation detection rate* Monogenic fibrosis 1:40003 Duchenne muscular dystrophy 1:40001- 12 X syndrome '4 Huntington's disease 1:5000-10 000 Hemophilia A 1:10 O00q~ 13 1:10 000 14 kidney disease 1:15005 15% 15 cancers:::: cancer 1:4000 9 50-65% 9'16 9 35% LFFraumeni-syndrome § Ataxia-telangiectasia ¶'1 17 polyposis coil 18 Hereditary non-polypoeis coil 1:2000 1°'1133% disorders hypercholesterolaemia 1:500 Hypedipidaemia §'2 *If no mutation is detected and the pedigree is large enough, haplotype analysis can include or exclude a gene as potentially mutation-bearing. disorders, frequency in males. Lifetime risk of colorectal cancer among caucasians is about 4%. Among women, the lifetime risk of breast cancer is 10-14%. Both these risks increase with age and positive family history, for hereditary non-p

olyposis colon cancer, the so-called Amsterdam criteria and MSI tests are used to preselect for gene testing. 11 §Varying from 1 in 120 in Ashkenazi Jews to 1 in 4000 in other populations. ¶Rare familial clustering. For general reference DNA diagnosis is commonly thought to have been widely available for years, such as Duchenne muscular dystrophy (DMD), haemophilia A, and cystic fibrosis (CF), the mutation detection rate is only 60-90%. 7 This low rate is due to a combination of factors such as gene complexity Molecular medicine • 354 • July ° 1999 sz5 LANCET 1: Status and advances in mutation detection (scanning) technology Classic technologie~f which some variants are frequently being developed gene sequencing The gold standard and increasingly an option, especially for smal er genes (5-10 kb), since high-throughput 96-channel cap I ary sequencers have become available. variants when several different run conditions are used. arising structureand detection of the product. Protein-truncation test 22~32 Specifically designed to detect stop mutations. Based on thecoupled transcr ationof the mutated sequence, to which causesa truncated protein product when a stop mutation is present3 ~ Uses cDNAcopies of the rnRNA of interest. In some cases it copy. technology probably available in the near future spectrometry ssDNA fragments of up to a few hundred bp can be ionised and their time-of-flight in an electrical field measured (MALDI-TOF). 34 Provides such precise molecular-mass data that the exact composition of the fragment can be derived from known tables and usually unambiguously correlated with the possible sequence. Widespread use currentlydepends on the availability of sufficiently extensive and precise tables34 DNA chips Examples have been published of mutation detection with very-high-density DNA chips (upto 400 000 datapo

ints per cm2), developed by Affymetrix. ~5 These Gene Chips (Affymetrix, Santa Clara, CA, USA) contain the complete coding sequence for the CF gene. 3" the human mtDNA sequence2~a segment of the BRCA1, ~ thep53 gene, and a few others. For negative and positive controls, most potential var ants of the sequences need to be included as wel DNA ch p techno ogy especially after further system development to reduce unit cost and increase robustness in diagnostic settings, is thought to be one of the more powerful and scalable future technologies. ds=double stranded_ss=single stranded. cancer is a case in point, with BRCA1 and BRCA2 identified but possibly more genes still to discover. Polycystic kidney disease is another example; with a frequency of 1 in 1500, it is one of the most common autosomal-dominant monogenic diseases. 85% of cases are caused by PKDI mutations on chromosome 16, and have a very poor mutation detection rate of 10%, ~s and 15% are caused by PKD2 mutations on chromosome 4. 8 The heterogeneity score is led by vision loss due to retinitis pigmentosa with, so far, 24 loci identified. 19 Unfortunately, molecular geneticists see themselves defeated by their success. The public as well as primary health-care providers and policy-makers expect wonders from DNA diagnostics, while asking increasingly difficult diagnostic questions. At the same time, amazingly little funding is being specifically devoted to improving the technology. The latter is often deemed a problem already solved, and the resolution of low detection rates, even for the common monogenic diseases, is seen as marginal cosmetics. If the current high expectations of genetic diagnosis are not to end in disillusionment, health-care policy bodies should stimulate further academic involvement in developing powerful, low-cost diagnostic technology and its applicatio

n to common diseases as well as rare diseases. Clearly, the latter diseases will be low-priority targets for commercial development, but it will be increasingly ethically unacceptable to send a significant proportion of patients home without an answer. This uncertainty will reflect badly on the field of medical genetics and cause substantial psychological damage. Studies of applicants for predictive tests for Huntington's disease 2° and breast cancer 2°,21 convincingly show that people deal far better with either positive or negative results than with receiving no result at all--especially since this often concludes a long, anxious, and infringing decision process. In other words, molecular diagnostics has a great future in supporting people who will make decisions that affect personal life and family planning and allow substantial and voluntary savings in health-care costs, but the task is far from finished. To realise its potential, the high expectations should be reduced and today's complacency replaced by active investment to improve cumbersome technology. There is still ample room for technological innovation. st6 Molecular medicine ° 354 • July ° 1999 LANCET oligonucleotide blot hybridisation (ASO) 2z39 Olfgonucleotide-Iigation assay 22,4° fuyv~l ol~l Jal-~v-i Ivl~ I auu L! IO! I I lyUI lUI~aUUI I-UclbeU tt~UI If tlLILIB5 ~flL1 T.ILI$, oe~er OIsIIngulsrllrlg and mismatched samp e sequences. Adaptations to a solid-state array ng format are in progress. minisequencin~ z~.4~ techniques." Some of the scanning techniques, especially the DNA-chip technologiesfl ~ can also be used to screen for known mutations. detection molecular diagnosis, distinction is made between mutation scanning and screening technologies. Scanning technologies aim to find unknown mutations in candidate or known disease genes. Screening techniques aim

to find known mutations, preferably with high throughput. Scanning techniques that are available and emerging are shown in panel 1. Screening technologies are shown in panel 2, and those that involve fluorescent-energy transfer (FRET), currently becoming available, are in panel 3. An extensive treatise on existing technologies and future expectations has recently been published by Cotton. 22 In the panels, the sensitivities of the different techniques are expressed as a percentage of the mutations that could be confirmed in the DNA segment studied with any other technique. These detection rates should not be confused with the low diagnostic detection rates mentioned above owing to as yet unidentified types of mutations or genetic heterogeneity. analysis rapidly increasing power of bioinformatics (databases and internet), automation (laboratory robotics), and nanotechnology (DNA and laboratory chips) has resulted in a substantial improvement in information gathering and interpretation. On one hand, this directly improves diagnostic sensitivity and precision. On the other hand, this will allow us to fathom the more complex interactions and functional networks operating in living organisms. A structured description of the data from the genome analysis of human beings and other organisms, is but a first step in this process. The next step, more challenging but also more rewarding, will entail many forms of large-scale parallel comparisons. These comparisons will include those between related genes in one organism, variants of the same gene in different populations living under different circumstances (hence the value of the "Human Genome Diversity" research), genes and genomes of different organisms, genes in health and disease states in human beings and animal models--natural or designed--and above all, between gene expression in rest

and growth, in development and ageing, in cancer and degeneration. This type of research, sometimes reviled as non- hypothesis-driven, will shortly yield a bounty of new patterns of data. Since pattern recognition is the time- honoured basis par excellence to develop and test new hypotheses, there is little doubt that the genome project will increase our understanding and applications far beyond what is currently conceivable. "Expression profiling", the expression levels of hundreds or thousands of genes from high-density microarrays, will enable monitoring of the expression status of cells and tissue on a global basis and highlight induced and suppressed pathways by comparison with normal and diseased states. This technique will not only provide a strong boost to the study of more complex genotype-phenotype correlations, but also elucidate the events between the primary mutation and the dysfunction of the cell or organism. It is expected that expression profiling will itself soon become a powerful diagnostic tool. Most disease states will be characterised by a typical gene expression pattern in one or more tissues of relevance, a gene expression "signature". Once these signatures are well described for each disease, they can be reduced to their basic patterns, which can help in differential diagnosis or help define stages of progression in a disease. Thus, a new format of "functional-defect diagnosis" will make its entry. This approach is by no means limited to genetic diseases: it will be equally applicable to many forms of dysfunction and cancer. Moreover, this approach will be a potent ally in reducing Molecular medicine • 354 • July • 1999 THE LANCET expression distortions. Gene-product diagnosis assisted if diagnostically relevant Therapy and prevention prevention. After of better order" and be replaced by Molecular medicine