Modern molecular methods in human disease

Studies and treatments of human disease have already been transformed by applications of new molecular techniques. Cloning has proved invaluable not only in the detection and amplification of DNA, as described in the previous section. but also in the safe generation of large amounts of purified proteins encoded by the DNA-for example, recombinant clotting factor VIII. The next section highlights three specific areas at the forefront of current medical research.

Historical overview

The earliest disease genes to be identified were analysed because insights into the pathological biochemical defects allowed analysis of the implicated gene, e.g. the globin genes in haemoglobinopathies. Such an approach would now be known as candidate gene analysis. The advances in molecular technologies have also allowed us to explore human diseases that had previously yielded few clues as to their cause. Instead, the disease endpoint can be correlated with inheritance of a particular chromosomal region, as a means of identifying the precise location of a disease gene before proceeding to its ultimate identification and isolation. Since this type of mapping approach leads to the characterisation of genes of unknown function, the process was originally referred to by Frank Ruddle as 'reverse genetics'.

The most efficient method of detecting the chromosomal segment bearing a disease gene is to identify a causative structural disruption which can be observed microscopically by karyotype analysis . Identification of numerous disease genes has been facilitated in this way. For example, the chromosomal locations of many tumour suppressor oncogenes were identified because a common mechanism for loss of the 'normal' second copy is by deletion of a chromosomal segment sufficiently large to be recognised by chromosomal analysis (the loss of heterozygosity' approach). A second approach is to identify a co-segregating genetic disease for which the disease gene position is already known. Usually, neither of these approaches is possible, so inheritance of the disease is compared to the inheritance patterns of anonymous markers of DNA for which the genetic location has been determined. These markers exploit polymorphisms that distinguish alleles on different chromosomes. Early markers were based on restriction fragment length polymorphisms (RFLPs), but these have been superseded by short tandem repeat polymorphisms in which repeats of 2–5 bp such as (CA)n and (AGAT)n occur, with the number of repeat units varying between different chromosomes . Allele-sharing by individuals with the disease may then be assessed in:

• large families (linkage studies)
• family subsections such as affected siblings (sib-pair analyses)
• isolated populations with a small pool of genes (and often a shared environment)
• laboratory animal populations which transmit a disease similar to that of humans.

To date, the most consistently successful approach has been linkage analysis in monogenic diseases, exploiting Mendelian inheritance principles through large families. The principles of linkage analysis.

Linkage studies are less useful in polygenic and multifactorial diseases when the contribution of each mutation to the disease is small. In this case, allele-sharing may be assessed between affected individuals who do not belong to the same family. A discussion of the advantages and limitations of these methods is beyond the scope of this text. However, their strengths may be illustrated by successes such as asthma, in which susceptibility loci have been determined using many of the approaches listed above-allele-sharing between siblings, in a geographically remote population (Tristan da Cunha), and in mouse models.

The methods outlined above identify the sites of novel disease genes. Two complementary approaches can then be used to identify the disease gene from the identified region.

Positional cloning strategies characterise the disease gene interval in detail by defining distances between genetic markers in the region. The approaches have been assiste by the availability of a number of libraries containing segments of human genomic DNA cloned into a variety viral, bacterial or yeast-based vectors. These clones may be required to isolate genes in the region; alternatively, they may aid the fine placement of genes already known to map to the general area. Finally, a suitable gene located within the disease interval can be analysed for mutations in individuals with the disease. Particularly with the advances of the HGP, once a disease gene region has been identified, it is often found to encompass the map placement of previously identified genes, some of which may be considered as potential candidate disease genes by virtue of their molecular structure, or data concerning the encoded protein. Such a 'positional candidate approach, combining mapping and functional data, has become the most frequent means of identification of inherited disease genes.

Essentially, the sequence of the suspected gene in affected individuals needs to be determined and compared to the normal sequence. However, even with advances in automated sequencing technologies, this can still be a daunting task, particularly for large genes with numerous exons, and genes for which limited sequence information is available at the outset. Disease-causing mutations often occur in coding DNA sequence or intron-exon boundaries, but may reside in the promoter region or within introns, or may have deleted the mutant gene entirely so that in autosomal genes only the normal allele is detected. The exact methods used vary according to the nature of the suspected disease gene, but usually include the following elements:

• Assessment of the region by genomic Southern blot analysis to exclude large deletions and rearrangements.
• Sequence of DNA (or cDNA reverse transcribed from mRNA) which has been amplified by PCR. Initial strategies are likely to include screening methods for mutations such as altered gel mobility or mismatchgenerated susceptibility to digesting enzymes.

In the course of these studies, a number of variations are likely to be found in the DNA sequence in affected individuals. All will have been co-inherited with the diseasecausing mutation--that is, the variation and disease gene will be in linkage disequilibrium. Distinguishing between incidental variations in sequence and disease-causing mutations in that gene, or an adjacent gene, is not always easy. Definition of a disease-causing mutation should be based on evidence that:

• The variation is genuine, and not due to artefact gener ated by a single method of analysis. This is helpfully addressed by supporting data from at least two of genomic DNA, mRNA or protein.
• There is no other potential causative variation within the gene, i.e. the gene should usually be sequenced entirely.
• Only individuals with the abnormality have the disease or disease predisposition, and all affected individuals within the family share the same abnormality. This is relatively simple for a monogenic disease but more difficult in polygenic diseases.

The goal of molecular medicine is to generate new therapies for disease to accompany the goal of molecular biology directed towards the understanding of biological processes and disease. Unfortunately,to date molecular techniques have tended to generate information about disease susceptibility in the absence of any therapeutic advances.

However, several approaches are likely to lead to better understanding of disease pathogenesis and the generation of new therapies. This is particularly true of attempts to create a model of a human disease in a system which can be manipulated. Biochemical effects can be studied in an individual cell type by in vitro culture experiments, either studying cells derived at biopsy from affected individuals or transforming cells with abnormal DNA to mimic the human disease state. While crucial, the potential of such studies is limited to selected conditions and devoid of the types of cellular, hormonal and immune-mediated responses critical in the whole organism.

A particularly powerful technique is to recapitulate the disease state in a laboratory animal, allowing assessment of the natural history of disease, effects of risk factors, and analysis of potential therapeutic approaches .