In order to genetically engineer a tree, one must be able to insert the desired gene (transgene) into the chromosome of an individual cell, and reproduce a whole plant from the cell containing the transgene. The former is referred to as transformation; the latter is called regeneration. Because poplar is amenable to both of these processes, it is frequently used as a model tree species. For selected poplar genotypes, transformation and regeneration can be accomplished in about six months.

Most genetic engineers capitalize on the ability of a common plant pathogen, Agrobacterium tumefaciens, to transform plant cells. However, Agrobacterium inserts genes in a random fashion, and the surrounding DNA affects the efficiency with which an inserted gene is expressed. Thus, numerous lines (i.e., resulting from independent insertion events) need to be produced and field-tested over a number of years and sites to ensure that the transgene is stably expressed at appropriate levels, and that the trees are otherwise unaffected.

To date, poplars have been genetically engineered for a diverse set of scientific and economic goals. The genes controlling these traits may come from a variety of sources, including bacteria and other plant species. Commercially relevant traits include: insect and disease resistance, herbicide tolerance, wood properties, early flowering, and reproductive sterility. The latter is done both for transgene containment purposes and to maintain rapid rates of vegetative growth after trees mature.

In order to introduce a trait through conventional breeding, the responsible gene must be available in a sexually compatible gene pool. For many desirable traits, this is not the case. Moreover, some of these traits are under polygenic control. Trees also have relatively long juvenile periods and extensive genetic segregation occurs during sexual reproduction. These problems are largely obviated through genetic engineering. However, once a transgene has been stably integrated, it can be introgressed into genotypes that have been improved through traditional breeding.

We have conducted numerous short-term, small-scale field studies with transgenic poplars in six states over the past several years. Data from these trials reveal that a large majority of transgenic trees show no evidence of damage due to genetic rearrangements, and that inserted genes are expressed stably from year to year, after vegetative propagation, and in a variety of environments. A number of genes that are expected to impart sterility have been used to produce transgenic lines, most of which are now or will soon be field-tested. Finally, we have been studying gene dispersal from wild and planted cottonwoods. Results from this work have been used to generate a landscape model for predicting the degree of sterility needed to provide an extremely low level of ecological impact from transgene spread into wild populations in the Pacific Northwest.