Sir

Your informative News Feature on germline engineering, “Biology's last taboo” (Nature 413, 12–15; 2001), suggests that the presence of donated mitochondria in a human embryo created with donor cytoplasm does not constitute germline engineering. Specifically, the feature states: “Mixing mitochondria does not risk damaging important chromosomal genes, nor could it endow children with eugenic traits.” Gregory Stock of the University of California, Los Angeles, is quoted as saying that applying the term 'germline engineering' to this situation is “playing games with words”.

Unfortunately, we know that devastating human diseases — heritable through the female line — can be caused by mutations occurring in the handful of genes carried on the mitochondrial genome and required for oxidative phosphorylation. It is not true to suggest that harm to the germline cannot come from mitochondrial manipulation. It's less clear whether mitochondrial DNA manipulation could result in inheritance of positive traits, but one might speculate that the rate or efficiency of oxidative phosphorylation could affect highly energetic tissues, say of muscle and nervous system, with unknown consequences.

You mention the work of Jacques Cohen and colleagues, using donor cytoplasm and mitochondria to 'rejuvenate' the eggs of women with fertility problems. This is not inconsequential to women of child-bearing age who have been diagnosed with a mitochondrial DNA disease. A process such as this 'rejuvenation' may hold the possibility of allowing these women to bear children who do not carry the mutant mitochondrial DNA in their genomes. Thus it represents exactly the same promise of germline engineering discussed in relation to chromosomal DNA defects. This is not to support or condone the work conducted by Cohen, but merely to point out its significance.