News

The AFSSAPS (French Agency for Health Product Safety) has again suspended gene therapy trials for X-linked severe combined immunodeficiency (X-SCID) led by Alain Fischer and Marina Cavazzana-Calvo at the Necker hospital in Paris. Treatment of children with this rare genetic deficiency began in 1999, but the trial was suspended between 2002 and 2004 when two of eleven children treated at that point developed the same form of leukaemia caused by insertion of the retroviral vector used for the gene therapy (see journal club article). One of these children has subsequently died; however, the trials were resumed due to the severity of the disease; without therapy, affected children will die anyway due to overwhelming infections. Now the trials have been halted again because a child treated in 2002 has developed abnormal T-cell lymphoproliferation (an overgrowth of white blood cells), which could be another form of leukaemia. This complication is under investigation.

Natallie Evans, who has been denied the right to use her frozen embryos by the High Court and Court of Appeal, has decided to take her case to the European Court of Human Rights (ECHR) (see BBC news story). Ms Evans wants to have implanted embryos created and frozen when she and her then-partner Howard Johnston were undergoing IVF treatment. She is unable to have children naturally as she was left infertile by treatment for ovarian cancer. However, her relationship with Mr Johnston has ended; he has refused his consent to allow Ms Evans to use the embryos and requested that they be destroyed.

Ms Evans has argued before the courts that as Mr Johnston had already consented to the creation, storage and use of the embryos, he should not be able to veto her right to use them. She believes she is being discriminated against because if she had become pregnant naturally, her partner would not have been able to force her not to have the child. In the previous decisions, the High Court and the Court of Appeal ruled that, according to UK law (the Human Fertilisation and Embryology Act 1990), consent from both the biological parents was required to enable embryos to be used. As Mr Johnston had withdrawn his consent, approval could not be given to Ms Evans. Permission to appeal to the House of Lords was refused as her petition did not raise any new issues in law and had already been the subject of judicial determination.

The ECHR will only hear cases brought by individuals who feel they have been the direct victim of a breach of one of their fundamental rights. Ms Evans will ask the ECHR to decide if the UK law breaches her right to a private and family life. The point will also be raised whether the embryos have a qualified right to life of their own. The ECHR has not as yet considered these issues and their decision could have wide-reaching legal implications.

The International HapMap Consortium, a public-private partnership of scientists and funding agencies from Canada, China, Japan, Nigeria, the United Kingdom and the United States, has announced plans to use additional funding of $3.3 million to create an improved version of the map of human genetic variation. The first draft of the human haplotype map, comprising 1 million single nucleotide polymorphisms (SNPs) as markers of genetic variation, is due to be completed by the end of this month. The improved ‘Phase II’ HapMap will be much more information-rich than the first version, incorporating many more SNPs and increasing the density of SNP markers from the present average of one for every 3,000 bases in the human genome to around one for every 600 bases. The second phase of the HapMap project will enable researchers to refine gene discovery to specific regions of the genome in the search for genes related to common diseases such as asthma, cancer, diabetes and heart disease.

NHGRI Director Francis Collins commented: "This will help us create a far more powerful HapMap than we ever imagined. We sincerely thank all those who are giving their time, technology and money to help turn this dream into reality. The payoff will be a better understanding of the genetic risk factors underlying a wide range of diseases and conditions". The additional HapMap funding came from the National Human Genome Research Institute, the Wellcome Trust, Genome Canada/Genome Quebec, Bristol-Myers Squibb and Pfizer. Perlegen Sciences is making its rapid, high-throughput genotyping capacity available for the genotyping of additional SNPs from publicly available databases to add to the HapMap.

88% of GPs surveyed on behalf of the insurance company Norwich Union Healthcare said patients should know more about hereditary conditions; even when patients did provide basic family histories, less than half of the doctors felt confident about this information (see BBC news item). Another survey of members of the public found that a third knew little about their family's health history and over three quarters did not feel they needed to know. This is an interesting contrast to a recent US survey, in which over 95% of respondents did consider their family history to be important for their own health (see CDC report). UK GP Dr Ann Robinson said: "If you know you have a family tendency to develop Alzheimer's, heart disease, diabetes or some types of cancer, you can get specific advice about how to prevent the disease yourself or at least pick it up in the early stages". However, cancer and Alzheimer's disease experts have stressed that a family history of disease does not necessarily mean a person is at risk. – 9/2/04 Dr Philippa Brice

The European Commission has announced that a consortium has received 14.5 million euros to study how diet and genetic factors can combine to cause obesity. Consisting of individuals from 30 organisations across 15 European countries, the DiOGenes (diet, obesity and genes) Integrated Project will “…carry out the most comprehensive study yet of dietary components and the genetic and behavioural factors influencing weight gain.” The project will have different parts, beginning with an investigation of the role carbohydrates and high dietary protein play in making individuals feel satiated or ‘full.’ In parallel, research partners will conduct a dietary intervention study involving normal weight and obese families in eight European countries. Researchers will analyse the genetic variation in candidate genes in order to identify sets of DNA polymorphisms, adipose tissue mRNA's and plasma peptides that would help predict an individual’s response to nutrients involved in weight change. In addition, the psychological and behavioural factors of research participants that influence weight gain will be analysed. Food technology studies will aim to develop foods that consumers like but which also provide ‘enhanced satiety signals’ that would help limit intake.

The ultimate goal of the consortium is “…to develop a method of predicting an individual’s weight change in response to different dietary nutrients, which in turn should lead to personalised diet-based treatments.” A demonstration project will show how newly developed concepts related to dietary advice and individual risk analysis in weight management can be used in practice. As it is estimated that obesity and co-related morbidities account for five percent of the health budgets of European countries, interested parties such as policy makers, the food industry and consumer groups will no doubt eagerly await the results from this project.

The Organisation for Economic Co-operation and Development (OECD) has launched a consultation on its Draft Guidelines for the Licensing of Genetic Inventions. The need for these guidelines grew out of the discussions held at a November 2002 meeting of the Working Party on Biotechnology, the senior intergovernmental body of the OECD that deals with biotechnology. At that meeting, member countries agreed that a set of guidelines should be written and disseminated. At an earlier OECD workshop on these issues some concerns had been raised regarding intellectual property as it has been applied to genetic inventions used for human health care purposes. While it was agreed that in general the current intellectual property system is working, it was felt that guidelines were needed, especially in the area of access to diagnostic genetic tests. The draft guidelines therefore focus on principles and best practice for licensing inventions used in human health care. In particular, they “…seek to foster the objectives of stimulating genetic research and innovation while maintaining access to health products and services.” The consultation period ends on 11 March 2005. The OECD hopes to endorse the final guidelines, revised based on responses to this consultation and further expert meetings, by the end of 2005 or beginning of 2006.

The Consortium for the Barcode of Life (CBOL) announced at the International Barcode for Life conference in London that it will create a database containing the DNA sequence of every species on earth, according to the BBC (see BBC news report). The international consortium aims to catalogue our estimated 10-30 million species, at a cost of about £1 per genetic test, by storing a short stretch of mitochondrial DNA from each species. The information will supplement existing taxonomic data, enabling researchers to link into those resources. Zoos, botanic gardens, government agencies and private sector companies from around the world make up the CBOL. British members include the Royal Botanic Gardens, Kew; the Natural History Museum, London and the University of Bangor. The data gathered will be publicly available; the National Center for Biotechnology Information at the US National Institutes of Health is creating a barcodes section in its GenBank database.

Three specific initial projects were announced. First, all known fish types, approximately 15,000 marine and 8,000 freshwater species, will be coded in the ‘Fish-Bol’ project. The bird project, Birds of the World, will catalogue the 10,000 known avian species and 8,000 plants in Costa Rica will be catalogued in a plant project. It is hoped that accessing this data will help researchers identify species more easily and more reliably. As well, the knowledge collected may help in conservation efforts.

Doubters fear that this effort might further damage the already dwindling art of taxonomy, a discipline that is being pursued by fewer people, while the CBOL sees expanding interest in taxonomy as one of its goals. Others are concerned that the barcode system will not be reliable in distinguishing closely related species, for example when a new or hybrid species appears in the population. Those who believe in barcoding had difficulty getting initial support for their idea [Wade, N. NY Times 14/12/04]. However, enthusiasm for the project is growing and CBOL expects to see the 10,000 avian barcodes available by 2010.

15 February 2005A new five-year collaborative project between six EU countries and Iceland is to look at the genetic basis of addiction. It has been recognised for some time that susceptibility to forms of addiction (abnormal dependency on substances such as tobacco, alcohol or drugs, or habit-forming behaviours such as gambling or eating disorders) has a genetic component, although environmental influences play a major part in the development of addictions. The new European Integrated Project GENADDICT (genomics, mechanisms and treatment of addiction) aims to identify these genetic factors and unravel the biological mechanisms underlying the condition, in the hope of developing new strategies to treat affected individuals. Project co-ordinator Professor Ian Kitchen of the University of Surrey, said: “In the short term, it could lead to the diagnosis of people that are prone to addiction, while in the long term, our research could help to identify novel targets for pharmaceutical companies to produce treatments that can fight drug cravings and relapse” (see interview).

The project has received €8.1 million EU Sixth Research Framework Programme (FP6) funding, and brings together eight leading public and private research organisations including Reykjavik-based deCODE genetics and Iceland’s National Centre of Addiction Medicine, which will lead the human population genetics arm of the research. This will be combined with animal genetics studies and gene-expression strategies, and it is hoped that the research will lead to the identification of genes that are common in addiction processes, as well as genes that may be specific for predisposition towards particular forms of addiction. However, such research highlights potential ethical issues that could arise if genetic associations are successfully identified, such as the problem of testing for genetic susceptibility to addiction.

10 February 2005A range of scientists and business people launched an initiative this week to establish a foundation in the UK to support stem cell research and its applications. The group are seeking £100million to create a charitable foundation that will aim to accelerate translational research and the use of stem cell medicines in the clinic, bringing closer the prospect of treatments for conditions such as Parkinson’s Disease and spinal injury.

The key figure behind the plans for the foundation is Professor Sir Christopher Evans, the former biochemist and Chairman of venture capital group Merlin Biosciences. Following a distinguished research career, Sir Chris established over twenty science-based companies, including Chirsoscience plc, which is now part of the Celltech group and one of Europe’s largest biotechnology firms. Merlin Biosciences also have investments in a number of life science companies, including ReNeuron whose products are based on stem cell technologies. Sir Chris is now on personal quest to ensure that the clinical and commercial benefits of stem cell research are felt in the UK, and not taken overseas due to the lack of support for essential work such as clinical trials. In an interview with BBC News, he commented, "[Britain] pioneered this entire field but now we are sliding backwards somewhat, as others accelerate ahead. You see big breakthroughs from China, Korea, Japan and in Germany; and there is a wall of money surfacing in the USA". He added, "We need to take potential stem cell medicines, regenerative medicines, to the clinic, into patients and do proper clinical trials and prove the benefits. That's the delivery everybody wants, but to get there it's a few more years and a lot of money and I think that there's a gap - I know that there's a gap - and I think it's quite substantial and I want to do something about it."

The proposed foundation has already drawn a number of high profile supporters including, Virgin Group chairman, Sir Richard Branson, scientist Lord Robert Winston, geneticist and author Professor Steve Jones, and the former Chairman of GlaxoSmithKline, Sir Richard Sykes. The foundation is seeking private as well as public funding and is currently reported to be in talks with Government on this issue. –

8 February 2005Professor Ian Wilmut of the Roslin Institute in Edinburgh has been granted a licence to clone human embryos. The Human Fertilisation and Embryology Authority (HFEA) has given permission for the cloning as part of research into motor neurone disease (MND) led by Professor Wilmut and Professor Chris Shaw of the Department of Neurology at King's College Hospital, London; this is only the second such license awarded by the HFEA to date. The first license, awarded to the Newcastle Centre for Life in August 2004 for cloning to create embryonic stem cells genetically identical to donor cells, is currently subject to a legal challenge, brought on the grounds that the HFEA was not forthcoming and transparent in its reasons for granting the license, and that it could not lawfully grant the license (see previous newsletter item).

MND is a severe and progressive condition affecting around 5000 people in the UK, caused by the loss of motor neurones that control muscle movement via the brain and spinal cord. The researchers plan to clone embryos from MND patients in order to create embryos with the disease, with a view to studying the mechanism of motor neurone degeneration and investigating how cells may respond to new drugs. This is in contrast to most forms of therapeutic cloning, where the aim is to use embryos to create healthy tissue for the treatment of people with diseases. Angela McNab, Chief Executive of the HFEA said: “We recognise that Motor Neuron Disease is a serious congenital condition. Following careful review of the medical, scientific, legal and ethical aspects of this application, we felt it was appropriate to grant the Roslin Institute a one-year licence for this research into the disease” (see press release). Patient groups have reportedly supported the research as potentially providing more information than animal experimentation (see BBC news report) whilst critics said that the procedures are immoral or of questionable benefit.

24 February 2005The United Nations Sixth Committee has recommended to the General Assembly adoption of a non-binding declaration calling for a ban on human cloning. The United Nations Declaration on Human Cloning calls for the ban, as it is incompatible with human dignity and the protection of human life. The vote was 71 for the measure, 35 against, with 43 abstentions. This declaration was agreed only after a decision on a binding treaty on this issue could not be reached (see PHGU newsletter November 2004). Originally, Costa Rica, with allies such as the United States, had fought for a treaty banning all types of human cloning. Belgium and the United Kingdom, as well as other countries, disagreed, stating that there should be a ban on reproductive cloning but that countries should be left to legislate on therapeutic cloning at a national level. When a decision could not be reached, Italy proposed a weaker non-binding declaration. Representatives debated the wording in declaration fiercely over several days but no consensus could be reached. The United Kingdom voted against the resolution and, together with other dissenters, has stated that they will not abide by it as the wording could be interpreted to include therapeutic cloning, which is legal in this country. The Costa Rican delegate praised the decision, noting that ‘Human dignity should always prevail over the interests of science.’

18 February 2005The US Senate has approved, by a vote of 98-0, ‘A bill to prohibit discrimination on the basis of genetic information with respect to health insurance and employment.’ The bill, S. 603, would amend current employment and medical legislation to prevent genetic information from being used to discriminate against individuals. Amongst other provisions, health insurance premiums could not be raised because an individual has sought genetic services, such as testing or counselling. The cost of group health insurance plans could not be raised based on the genetic information of an individual member of the plan and individuals could not be required to undergo a genetic test. Health insurers would not be able to set rules for eligibility that are based on genetic information. Employers, employment agencies, labour organisations or training programmes would not be able to discriminate against an individual or deprive them from an employment opportunity based on genetic information. It would also be unlawful for an employer to “…request, require, or purchase genetic information with respect to an employee or a family member of an employee…” Exceptions would include the collection of family history information, if the employee provides written authorisation or if the organisation conducts genetic monitoring of the biological effects of toxic substances in the workplace. In addition, any genetic information that is in the hands of the employees would have to be kept confidential, except in cases such as responding to court orders. The bill lays out penalties for offenders.

The approval of this bill, however, does not signal the end of the legislative process on this issue. It must be sent to the House of Representatives for agreement or amendment. The bill might have difficulty in the House where, according to the Washington Post, “…business groups that oppose it hold more sway.” The Senate passed a bill almost identical to this one in 2003, however, the House never reached a decision on it. Supporters hope that this bill will have a smoother passage; both President Bush and Francis Collins, Director of the National Human Genome Research Institute, have signalled their support for the bill.

The use of DNA microarrays in tumour profiling is a rapidly growing area of research; gene expression profiles identified using microarrays are being correlated with specific molecular tumour signatures, and it is thought that this information can be applied to aid diagnosis, prognosis and choice of treatment. The first clinical trials using gene-expression profiling to stratify breast cancer patients into good and poor prognosis groups (the latter group being chosen to receive chemotherapy) are already underway in Europe and the USA.

DNA microarray data requires complex analysis, which may be approached in different ways. A new paper in the Lancet reports the results of an analysis of data from published studies of cancer gene expression as a predictor of outcome; the authors took advantage of the system of recording microarray data in an international repository according to set MIAME (Minimum Information About a Microarray Experiment) guidelines. This study used a training-validation strategy, whereby a ‘training set’ of microarray data is used to identify the molecular signature associated with a given outcome, and an independent ‘validation set’ of microarray data is used to estimate the proportion of misclassifications produced by that molecular signature [Michiels S, Koscielny S and Hill C (2005) Lancet 365, 488-492].

From a review of all microarray studies of cancer prognosis published between January 1995 and April 2003 [Ntzani EE and Ioannidis JP (2003) Lancet 362, 1439-1444], the authors selected all studies on survival-related outcomes (with rates between 14% and 58%) that included between 60 and 240 patients, a total of seven different studies. For each study, the dataset was divided into 500 training sets, within which equal numbers of patients had each of the two alternative survival-related outcomes, and 500 associated validation sets. This was intended to maximise the power of the comparison between average gene expressions in the favourable and unfavourable outcome groups. For each training set, a molecular signature was identified, comprising the 50 genes for which expression was most highly correlated with prognosis, and two average gene expression profiles (favourable and unfavourable) were defined. Each patient in the corresponding training set was classified as belonging to the favourable or unfavourable outcome group based on which average gene expression profile correlated most closely with their molecular signature. The proportion of misclassifications in the validation set for each of the studies was then estimated.

The authors found that the outcome-related molecular signatures were very unstable. When compared with the tumour signatures defined in the original studies, a significant degree of variation was observed. Typically, only some of the ‘key genes’ present in the molecular signatures defined in the original studies were found to be present in more than half of the training set signatures, whilst other genes absent from the original signature were present in over 250 of the 500 training set signatures.

The proportion of misclassifications was found to be related to the size of the training set used for analysis of the microarray data. With a training set of ten patients, the proportion of misclassifications for the seven studies varied between 40% and 50%, but in six of these studies the proportion of misclassifications decreased as the training-set size increased (down to 31% in the best case, for a training set of 90 patients). Hence, the use of larger training-set sizes would reduce the incidence of misclassifications and improve the predictive ability of molecular signatures. Michiels et al. also found that the use of only a single validation set for microarray data analysis greatly increased the instability of the resulting molecular signature as a predictor of outcome. Tellingly, five of the seven studies did not classify patients to outcome groups better than chance.

Comment: The authors of this paper advocate their chosen method of validation for microarray data, and stress the necessity for much larger studies before gene expression profiling can be used in a clinical setting. These results lend weight to existing concerns that validation of microarray studies is presently inadequate, and that much larger sample sizes are needed. A commentary accompanying the report calls for multiple independent studies, each comprising several thousand patients [Ionnnidis JP (2005) Lancet 365, 454-455]. The policy implications of this conclusion are significant. Notably, the need for allocation of public funding and resources for future research and development in cancer genomics is shown to be urgent. Normal practice in research funding is to grant limited funding for small, proof-of-concept trials, which may subsequently be considered for further funding to permit large-scale trials. For DNA microarray research, this approach may be wholly inappropriate, since to generate reliable prognostic signatures necessitates large sample sizes. Such trials would be made vastly easier by the use of a national or international open access data repository with required terms of standardisation (such as that set down in the MIAME guidelines).

In the UK, an NHS-based resource with appropriate infrastructure for the storage and use of tissue samples and related information including clinical data, may be the only sensible way forward in terms of realising the clinical potential of DNA microarray technology to transform cancer management (see 2004 Cancer Genomics Workshop report and Brice P (2004) Health Serv J. 114, 24-25). The National Cancer Tissue Resource, an initiative funded by the National Cancer Research Institute (NCRI) and co-ordinated by NTRAC, is the obvious candidate. This is intended to be a national NHS-embedded network of tissue banks linked to clinical trials and follow-up data, underpinned by a central bioinformatics hub. Researchers will be able to remove tissue samples and to deposit data, which will in turn be nationally accessible. But the urgency to establish resources and fund large-scale trials to generate genuinely useful gene expression profiles is clearly great, and may well call for much greater investment than is currently available for the National Cancer Tissue Resource.

28 February 2005Human embryonic stem (ES) cell lines are currently the focus of intensive research with a view to developing therapeutic interventions against a range of diseases, because these cells are pluripotent, having the potential to differentiate into any type of specialised cell. The hope is that they could therefore be stimulated to develop into the appropriate sort of cell to replace damaged or dysfunctional cells. Besides concerns over the ethical issues associated with the use of cells derived from human embryos, the safety of using such cells for transplantation is also an important consideration. A new paper in Nature Medicine reports that human ES cell lines are effectively contaminated with animal cell surface molecules and would therefore induce an immune rejection reaction if transplanted into human recipients (see also BBC news report).

ES cells are initially cultured on mouse embryonic fibroblast ‘feeder cells’ until the reach the embryoid body stage. At all stages of culture the stem cell lines are grown and maintained in nutrient media containing fetal calf serum, or alternative sera that may also contain animal products. The authors of the new study looked at specific sialic acids; these are a group of cell surface molecules present in mammals of which the most common are N-glycolylneuraminic acid (Neu5Gc) and its metabolic precursor N-acetylneuraminic acid (Neu5Ac). Human cells express primarily Neu5Ac, being unable to convert it into Neu5Gc. However, the research team led by Ajit Varki of the University of California, San Diego, had previously shown that human embryonic stem cells are capable of the uptake and cell surface expression of animal Neu5Gc derived from feeder cells or nutrient media. This poses a potential barrier to transplantation of such stem cells, since most humans are sensitised to Neu5Gc as a ‘foreign’ antigen, possibly as a result of eating meat and dairy products.

In the new publication the authors examine the effects of this expression of Neu5Gc on human ES cells [Martin MJ et al. (2004) Nature Medicine advance online publication, 25th January 2004]. Cells expressing Neu5Gc were identified using fluorescent antibody labelling, and separated from non-human feeder cells. Chemical analysis showed that Neu5Gc represented up to 10% of the total sialic acid composition of these cells, and the most probable source was found to be the commercial serum component of the cell culture medium, in which Neu5Gc represented over 50% of the total sialic acid content.The researchers experimented with the conditions of human ES cell culture, and found that Neu5Gc could not be wholly eliminated due to the essential initial requirement for mouse derived ‘feeder cells’ to grow the ES cells. However, the use of heat-treated human serum (selected to contain low levels of antibodies against Neu5Gc) rather than commercial serum replacement was found to significantly reduce the levels of Neu5Gc expression on the ES cells, by around eight-fold.

Exposure of human ES cells grown under standard conditions to human serum containing high levels of anti-Neu5Gc immunoglobulins (antibodies) resulted in binding of the antibodies to the ES cells, although when serum containing low levels of immunoglobulins was used, no significant binding was observed. The authors then asked whether this antibody binding would trigger the normal immune response to foreign antigens, by stimulating the binding of another immune mediator molecule called complement C3b, which in turn can initiate destruction of the foreign body by immune system cells. They found that 37% of ES cells grown under standard conditions and exposed to high levels of antibodies bound to C3b, as compared with 22% of such cells exposed to low levels on antibodies. Only 13% of ES cells grown in human serum and exposed to high levels of antibodies bound to C3b.

The authors conclude that human ES cells grown under standard conditions incorporate animal-derived Neu5Gc molecules to an extent that could well cause immune-mediated destruction on transplantation into humans. They warn that this includes all stem cell lines currently available to state-funded researchers in the US, which were derived using mouse feeder cells and are maintained under standard culture conditions. They also note that the presence of Neu5Gc could have additional adverse consequences, because the cell surface sialic acids are important components of cellular self-recognition; normal function and tissue development of transplanted cells derived from ES cells might be impaired.

Comment: This paper reinforces existing concerns about the use of animal derived products in human stem cell culture; efforts to create ‘feeder-free’ culture processes have been unsuccessful, although other groups are reportedly attempting to develop methods that wholly eliminate the use of non-human components. It would seem that there are valid reasons to persist with these efforts. However, the additional expense of using solely human-derived feeder cells and sera would be enormous (as would the authors’ proposed idea of using heat-inactivated serum from the individual who is to receive the cell transplant) and could potentially create problems in terms of quality control (maintaining consistent standards and conditions for use) as well as raising alternative safety concerns, such as the potential for transmission of infectious disease agents. The authors of the study suggest that the safest approach to human ES cell use would be to derive wholly new cell lines grown in conditions completely free of non-human components, which seems a sensible proposal, although subject to the caveats outlined above. However, as they note, this approach would not be available to state-funded researchers in the US under current regulations.

22 February 2005Prenatal testing for hereditary diseases and genetic abnormalities generally requires the use of amniocentesis or chorionic villus sampling (CVS), invasive procedures with a small but significant associated risk of miscarriage. Recently, there have been moves towards developing non-invasive methods of testing using fetal cells present in the maternal blood or plasma. Analysis of cell-free enriched samples of fetal DNA derived from maternal plasma has already been shown to be effective for the identification of fetal genetic traits that differ from the mother, such as the presence of Y-chromosome specific sequences, or the RhD gene in Rhesus negative mothers. However, this approach is not suitable for the analysis of fetal genetic traits that do not differ largely from the maternal alleles. A paper in the latest edition of the Journal of the American Medical Association reports on the use of a novel technique to detect fetal point mutations associated with the hereditary disorder beta thalassaemia.

The Swiss researchers have previously shown that fetal DNA present in maternal blood samples can be selectively enriched based on their size, which is typically smaller than maternal DNA sequences also present (an average of around 300 base pairs, rather than 500 base pairs for the maternal DNA). They then set out to investigate whether this would permit the detection of fetal point mutations [Li Y et al. (2005) JAMA 293, 843-849]. These most subtle forms of mutation underlie many monogenic disorders, including ß-thalassaemia; in this study, samples were taken from a population of 32 pregnant Italian women where the babies’ fathers were known to carry one of four disease-associated ß-globin gene point mutations common among the Mediterranean population (IVSI-1, IVSI-6, IVSI-110, and codon 39), whilst the mothers carried different ß-globin gene mutations.

Maternal blood samples were collected at prior to chorionic villous sampling. Fetal DNA sequences were separated out by size and amplified using the polymerase chain reaction (PCR), using peptide nucleic acids (PNAs) specific for the maternal ß-globin allele to suppress amplification. PNAs bind with high affinity and specificity to DNA sequences, and are capable of distinguishing between sequences that differ by as little as a single base pair; by suppressing amplification of the maternally derived fetal allele by at least 1000-fold, the paternally inherited mutant sequences were effectively enriched. An allele-specific form of quantitative PCR amplification was then used to determine the ratio of wild-type allele to each of the four mutant paternal alleles. This allowed the researchers to clearly distinguish between samples that contained wild-type alleles from those heterozygous for a mutant allele, even where the mutant allele represented a small proportion of the total DNA present. Results were compared with those obtained by CVS.

The technique correctly indicated the presence or absence of the four mutant sequences in 28 out of 32 samples (88%); there was one false result for the IVSI-1 allele and three for the codon 39 mutation. The authors attribute these inaccuracies to an inadequate concentration of template fetal DNA, and for the purposes of further analyses classed results as ‘uncertain’ where the total cell-free DNA levels were below a certain level, proposed to be a reasonable level to obtain from a standard maternal blood sample. This exclusion alters the calculated sensitivity and specificity of the overall analysis from 92.8% and 88.2% respectively, to a sensitivity of 100% and specificity of 93.8%.

The authors conclude that their two-stage procedure is effective for the detection of paternally inherited fetal mutant alleles for four common ß-thalassemia mutations, and propose that it would also be suitable for the detection of other fetal single-gene mutations, such as those associated with achondroplasia. They estimate that samples could be analysed using this technique for US$8 (around £4 / €6) each, and say it could be applicable for prenatal screening in developing countries. They also suggest that similar approaches could provide an alternative approach for non-invasive prenatal assessment of fetal single-gene disorders involving compound heterozygous mutations (for example, where both parents are carriers of different disease-associated alleles for cystic fibrosis).

Comment: This report represents an interesting development in progress towards non-invasive screening methods, but it would be premature to conclude that the technique will be useful in the clinical setting. For one thing, it would be of no use where parents carry the same disease-associated alleles. Moreover, although no screening test is likely to perform with 100% sensitivity and specificity, the results of this analysis were not compelling, given that the number of samples analysed was very small. It would be of value to see how well the technique performs with much larger sample sizes. However, if the technique proved sufficiently effective and could be automated, the relatively low cost might indeed make it of value for screening programmes in some parts of the world.