Article Text

Happy families
  1. BILL OLLIER,
  2. JANE WORTHINGTON
  1. ARC Epidemiology Research Unit, Manchester University Medical School, Stopford Building, Oxford Road, Manchester M13 9PT

    Statistics from Altmetric.com

    Barely a century has passed since Gregor Mendel provided the explanation for inheritance and a foundation for what we now understand as genetics. Looking back it is perhaps even harder to believe that it is only just over 40 years on from Crick and Watson, who with coworkers determined the structure of DNA and cracked the genetic code. Given this comparatively short period of time, progress has been dramatic in all areas of genetics, but particularly so in medicine where we are now in an exponential phase of development and the profession is gradually realising that it is caught up in the spiral of a new genetic revolution.

    Until recently medical genetics was relatively circumscribed and mainly concerned with rare monogenic conditions where the disease phenotype was often obvious or dramatic and the mode of inheritance known. Phenomenal recent progress in molecular biology has spawned molecular genetics and a new strain of geneticist interested in common conditions with a complex and polygenic aetiology. These include such conditions as schizophrenia, asthma, hypertension, heart disease, malignancies, and a wide range of autoimmune and arthritic conditions. This enthusiasm has been catalysed by major initiatives such as the Human Genome Project where the objective is to characterise all human DNA in terms of genes and sequence. With parallel advances in information technology such data are readily accessible and high density genetic maps of markers, such as polymorphic microsatellite repeats, are now available for linkage and association studies. Further technological developments have also been occurring at a pace and the methods for both DNA sequencing and the genotyping of DNA polymorphisms have been largely automated leading to a potential for vast throughput of samples and the development of ‘factory genetics’. Many centres world wide have already invested heavily in setting up major research programmes for complex disease and this is now being mirrored by the private sector and the pharmaceutical industry.

    Current strategies

    The major incentive has usually been to identify all the genes making a significant contribution towards aetiopathogenesis of a condition. (The precise definition as to where the lower threshold lies and what constitutes a significant contribution is often difficult to judge). This has been possible by adopting whole genome screening where no a priori hypothesis is made for any known or candidate gene. Instead consecutive regions of DNA (10-20 centimorgans) are tested for linkage within the condition, starting at one end of the human genome and finishing at the other. This is often referred to as exclusion mapping and an initial screen can usually be achieved using 300-400 polymorphic markers. Microsatellite repeat markers are ideal for this purpose and many thousands have been identified and placed on the genetic map at a resolution of one centimorgan.

    This approach has already been successful in a variety of conditions notably in type I diabetes where 11 possible susceptibility regions were identified1 although a growing list of success stories is now emerging.2-4 Such studies are based on the investigation of genetic linkage and necessitate the collection of families with multiple affected cases.

    The traditional approach to linkage studies has been to investigate multi-generation and extended pedigrees using LOD score (logarithm of odds) analysis. Using this method, assumptions have to be made about the mode of inheritance and estimates of both recombination (the level of cross over between the marker gene and the disease gene) and gene penetrance (the proportion of people carrying the disease genotype who express the disease phenotype). Furthermore the disease status of every family member analysed has to be decided (this is sometimes impossible especially in conditions such as rheumatoid arthritis or osteoarthritis where disease onset may be late).

    As such, LOD score analysis is generally not applicable to the investigation of polygenic disease and a more popular approach has been to use affected sibling pair (ASP) methods. By analysing only affected pairs the problems of gene penetrance, mode of inheritance, and recombination can be ignored. ASP analysis is based on the simple rules discovered by Mendel that for any given pair of siblings, the chances of inheriting alleles identical by descent (IBD) at any given locus are 25%, 50%, 25% for sharing 2, 1, 0 alleles respectively. Given sufficient numbers of affected sibling pairs, if the genetic marker under test is linked with the condition, there should be a significant distortion away from this expected distribution towards greater sharing of alleles. This method of analysis (and a number of variations based on it) is a robust and rigorous way of conducting exclusion mapping. Unfortunately it is relatively insensitive to weak genetic effects and many hundreds (if not thousands) of families may be required to detect small genetic contributions.5

    Currently whole genome screening of ASP families represents the best approach to the investigation of complex polygenic diseases. This has meant that a large collection of ASP families have been required and this has usually only been possible through concerted national or international initiatives.

    The British Diabetic Association were first off the mark and well over 400 type I diabetic sibling pair families have been collected and analysed. Late onset diabetic families have also been collected. The Arthritis and Rheumatism Council (ARC) were quick to follow with their National Repository of genetic material from arthritis families. This has particularly concentrated on adult rheumatoid arthritis (RA) and over 300 ASP families are presently being analysed. As has been shown in diabetes research, it is important to verify results in families collected in different populations and over 200 RA ASP families have already been collected as part of a French initiative. A further NIH sponsored collection of RA ASP families in the USA has also recently started.

    Rheumatology is proving to be a particularly rich area for analysis of complex genetic diseases and family material is being additionally collected for such conditions as; osteoarthritis, ankylosing spondylitis, osteoporosis, systemic lupus erythematosus, primary Raynauds phenomenon, and juvenile chronic arthritis.

    The advantages of centrally collecting affected families are many and it is clearly better to have one large collection with the power to answer the important questions rather than disparate smaller collections. It is critical that the epidemiological and clinical data collected are standardised and of the highest quality, and that diagnostic uncertainty and clinical heterogeneity are well documented. No matter how sophisticated molecular technology becomes, nothing sensible will emerge from analysis if the material being studied is too heterogeneous and contributes excessive ‘noise’ to the system.

    Centralised collection makes financial as well as scientific sense. If a good collection of family material exists, it can also be used in subsequent projects. These studies can be achieved immediately without delay or additional expense incurred for the repetition of sample collection.

    Some initiatives are going ‘the extra mile’ (and expense) of producing immortalised EBV cell lines for the material collected. This is clearly worthwhile if tremendous time and effort has been expended in collecting the families, if the families are rare or if many groups want access to the material. The amount of DNA extracted from a peripheral blood sample is highly variable and if DNA supplies become exhausted for a key family member, that family can cease to be of any practical use. Even in the age of the polymerase chain reaction, it is surprising how much DNA is consumed by active research groups. For example over 12 groups have been using DNA samples from the RA families within the ARC National Repository and in excess of 300 μg of DNA has been used on most samples. Where only limited samples of DNA are available, more care has to be taken as to who has access to this material.

    Ethical issues and ownership

    Ideally, so long as patient confidentiality is maintained, it is in the best interest of everyone for genetic material to be distributed to as many research groups as possible, thus maximising the chances of an important discovery and new avenues for treatment. Indeed this is the altruistic spirit in which patients willingly donate their samples and clinicians their time and expertise.

    The issue of who owns such material and the ethics of who financially benefits from intellectual property rights as discoveries are made and patents filed, remains a grey and uncomfortable area. This becomes an increasing anxiety as more biotechnology companies move into this area and either want to use such material themselves or collaborate with academic groups.

    Ownership of the National Repository of Multicase Arthritis Families remains with the ARC and material is only released on the proviso that any discoveries reflect this agreement. In doing so at least a proportion of any financial gain can be ploughed back into further arthritis research. The British Diabetic Association have already followed this course and by making family material available to American groups they have been able to finance further family collection.

    As often happens in new areas of research, ethical considerations tend to lag someway behind the developments in technologies and results that necessitate them. Family collection for genetic research has quickly become commonplace and the area continues to expand as the genetics of normal complex traits and behavioural genetics are tackled. Clearly as these studies progress more ‘cans of worms’ will continue to be opened and ethical issues relating to how genetic data are used will have to be tackled. For example, confidentiality of results and the requirements of health care and life insurance companies, the provision of counselling for patients who would not previously have thought of their condition as having a genetic component.

    As a preliminary response a discussion document on complex genetics in medicine6 has recently been published. More immediate are the practical issues of how collections of family genetic material are handled in an ethical way and how this can be made available to interested groups. The Wellcome Trust have recently sponsored a forum for all parties collecting family material, to discuss many of the issues involved in this complex area.

    As has happened in diabetes research, the analogy of ‘happy families’ may be appropriate, as the game can only be won if players are prepared to exchange their cards.

    References

    View Abstract

    Request permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.