London (April 30th, 2004) – A few days ago, there was a flurry of interest in the press about a suggestion that a particular human gene incorporated into rice could help weed control.

Until now, GM weed control has used bacterial genes introduced into crop plants, so enabling them to tolerate herbicides sprayed to kill weeds.

One such herbicide blocks the synthesis of essential chemicals for plant growth; the bacterial enzyme in the transformed plant provides an alternative route. Another weedkiller stops a step in plant metabolism, causing the toxic accumulation of a particular chemical; a different bacterial enzyme removes the block. However, this sort of tolerance is normally conferred just towards a single herbicide, the continued use of which might allow resistance to be built up among the weeds by random chance mutation. Resistance of this sort does sometimes occur in the wild if a particular weedkiller is used for too long a period.

Varying herbicide usage is a way of avoiding potential resistance build-up but then a whole range of tolerant crops would have to be developed, each resistant to a particular herbicide. Would it be possible to confer resistance to several herbicides simultaneously so they could be alternated in the field without having to use different crop strains?

A possible way of doing so might be to use a multi-purpose detoxifying enzyme which would destroy a range of herbicides. Some progress has been made in this direction: a gene coding for a human liver enzyme, which is particularly good at breaking down harmful chemicals in the body, had been expressed in rice. Plants modified in this way this might also be useful for removing toxic chemicals from contaminated land; the chemicals from the soil would be absorbed by the plants and then destroyed by the new enzyme they contained.

Obvious questions arise which must to be resolved before such a technique is used: will it work, will the gene migrate to other plants (including weeds) and what might happen if it did?

Another point brought up in the press articles was that no one will want to eat the partially human-derived food because it will smack of cannibalism. Is that an argument that makes sense?

What exactly constitutes “cannibalism”? A nice juicy human steak or a boiled missionary would, to most people, be clear examples. But this case is rather different. The only human aspect of the proposal is the use of human-derived information.

The physical gene itself (the appropriate stretch of DNA) can be obtained from a single human hair or from a cheek swab. It is then multiplied billions of times in the test tube; the original bit that came from a person would long since have been diluted and effectively vanished in the mass of material synthesized in the laboratory. Moreover, we already use such human genetic material for medical purposes without anybody seeming to worry: most, if not all, insulin used nowadays to treat diabetic patients is manufactured by microorganisms carrying the human insulin gene. Is that being cannibalistic?

While the combination and collective interactive functioning of our genes makes us uniquely human, the individual genes themselves do not. Around 98% of our genes are similar to those in chimpanzees, 95% to pigs, 90% or so to mice and so on down the evolutionary tree. Genetically we probably overlap 30-40% with bananas, with which we share a good deal of basic biochemistry. Yet nobody suggests that we are being cannibals when we eat a pork chop or a banana. So would it be cannibalistic for us to eat that proposed rice?


1. Rice with a Human Touch. Science News (16 April 2005) (

2. GM industry puts human gene into rice. Independent on Sunday (24 April, 2005) (

3. GM fears as human liver gene is put into rice. Daily Telegraph (25 April, 2005) (

4. S. Hirose et al. (2005). Transgenic rice containing human CYP2B6 detoxifies various classes of herbicides. Journal of Agricultural and Food Chemistry, 53, 3461-3467. (


  Human rice?