Sir

Barritt et al.1 have reported the birth of several 'transmitochondrial babies' following ooplasmic transplantation at in-vitro fertilization (IVF) centres around the world. In this technique, about 10% of the ooplasm is transferred from a healthy donor egg into a compromised patient's egg2. Because normal human oocytes contain between 200,000 and 5 million mitochondria, the transferred ooplasm probably contains about 20,000–500,000 mitochondria and mitochondrial (mt) DNAs, together with other cytoplasmic components.

Ooplasmic transplantation involves the mixing of mtDNAs in the cytoplasmic germ line. Because this may produce durable and unknown genetic effects that transcend single generations, this new technology must be scrutinized both publicly and scientifically. The social, medical and legal questions about the birth of children who bear the DNA of three parents — mtDNA from the biological mother and the ooplasm donor, and nuclear DNA from both biological parents — cannot be fully anticipated without public discussion and timely scientific publication of results. Cohen and his colleagues are to be commended for their open reporting of rare cases of aneuploidy and other complications associated with ooplasmic transplantation3.

We agree with Parens and Juengst4 that increased public supervision is needed. Along these lines, the US Food and Drug Administration (FDA) issued a letter on 6 July 2001 (http://www.fda.gov/cber/ltr/cytotrans070601.htm) to investigators using ooplasmic transplantation and other forms of gene transfer in human gametes. Investigators using these technologies are now required to submit an 'investigational new drug' (IND) application to the FDA, outlining the research methods and objectives of the clinical study.

We wish to make two additional recommendations. First, we suggest that the National Institutes of Health (NIH) expands its Recombinant DNA Advisory Committee, or creates a new regulatory body, to encompass all ooplasm and nuclear-transfer technologies, to allow informed regulation and public reporting. Until a more solid public and scientific database is established, the number of licences issued should be limited to a handful of highly qualified, excellent IVF facilities. Each of these should be required to follow certain protocol-driven data collection and publication procedures.

Data collection would include haplotype analysis of donor and recipient mtDNAs, screening for the most common pathogenic mtDNA mutations, documentation of gestational complications and maternal–fetal immunological and metabolic parameters, and careful neurodevelopmental testing of the infants and toddlers. The NIH regulatory body should include: people with expertise in a wide range of IVF technologies, including ooplasmic transplantation; specialists in medical ethics, nursing, law, mitochondrial cell biology, mitochondrial medicine and genetics; and informed members of the general public.

Our second recommendation is that the US Congress set aside a budget for a request for applications, issued by the NIH, to solicit proposals to study animal models of the ooplasmic transfer technology. Although significant biological differences make direct comparison of the details of reproductive cell biology between species difficult, these studies would provide the basic data needed to understand the pros and cons of ooplasmic transfer technology more fully5. It is only with carefully regulated progress, timely and systematic publication, public supervision and public debate that we can maintain public trust in science, while moving forward in the genomic and post-genomic ages of medicine that lie ahead.