It seems clear that using transgenic plants (Hg model and Cd model at this site) is not only a possibility but can provide tremendous potential for enhancing phytoremediation of heavy metals.

New genes and their functions are currently being investigated.  It is important to note that engineering plants should not be the only strategy utilized in a phytoremediation project but complement strategies implemented.  Moreover, there are potential risks in using genetically modified plants as phytoremediators.  Introduced DNA and the vectors that contain them should be well characterized.  Extensive field trials using transgenics in phytoremediation have not been reported and must be done to investigate their potential in the field.

Two important considerations for any transgenic plant desired for use are i) potential weediness and ii) the ability to outcross

Potential Weediness:

•  Some plants are subject to weediness by single gene changes although usually several gene changes are necessary.  Extensive agriculturally bred plants are less likely to exhibit this phenomenon.  However, most potential species for phytoremediation are not agriculturally significant.  Therefore, one must investigate in the literature, field, and laboratory, whether single gene changes do provide a growth advantage or weediness. It is the opinion of this author that transgenic plants used in phytoremediation would not offer any advantage outside of the contaminated site containing the pollutant the said plant is engineered to remediate.

• The ability to outcross is dependent on several factors.  Are there any relatives near the plant of interest?  Is it possible to control for pollination by harvesting plants before flowering or by removing pollen-producing organs.  Is the contaminated area isolated from related species?  There is a possibility transgenic plants could cross with near-by food crops conferring undesirable accumulation in the food crop species.  However, as David Glass points out, "the proximity of such food crops to hazardous waste would be more troubling than the remote possibility of a single gene transfer across species."