168 VOLUME 25 NUMBER 2 FEBRUARY 2007 NATURE BIOTECHNOLOGY Genetically modified lignin below ground To the editor: Correspondence in Nature Biotechnology 1–3 in response to a News Feature on the biotechnological potential and environmental risks of releasing transgenic trees 4 has drawn attention to our earlier paper reporting the first field trials of trees with modified lignin 5 . The focus of this debate on the below-ground effects of genetically modified (GM) lignin is highly appropriate because lignin is one of the most abundant biopolymers on the planet, an important regulator of the decomposition of plant residues and the precursor of much of the stable organic matter in soils. The amount of carbon in soils outweighs that in vegetation and the atmosphere combined 6 and has the potential to moderate future climate change. Talukder 3 raises concerns that a low-lignin phenotype may pose an environmental risk by promoting faster decomposition of litter and increased CO 2 emission because microbial enzymes will reach their target polysaccharides more easily when the physical barrier presented by the degradation-resistant lignin is reduced. This proposed influence of altered lignin on decomposition is not speculation, as we have shown in both poplar 5 and tobacco 7 that short-term (<100 days) decomposition of plant material with modified or, in the case of tobacco, low-lignin, is significantly faster than that of corresponding wild-type material and that this difference is largely explained by reduced protection from microbial attack afforded to labile components by the modified lignin 8 . But before this evidence is wildly extrapolated to global scenarios of climate change, however, we would urge consideration of the longer term—after all, the complete breakdown of plant residues can take decades or more. A study from our group spanning 18 months found no significant differences in the extent of decomposition between field-grown wild-type and lignin-modified poplar wood; indeed, the variation between replicates of each genotype was greater than the variation between genotypes 9 . This suggests that changing environmental conditions during growth in the field have a greater influence on wood decomposition than the genetic modifications to lignin in these genotypes. Similar work we have carried out on lignin-modified tobacco grown and allowed to decompose under controlled laboratory conditions demonstrates that once the accessible polysaccharides have decayed, the decomposition of the remaining lignin- rich material is either no different or actually slightly slower for modified plants than for the wild type (Fig. 1). There could be several explanations for this. Lignin- biodegrading fungi rely in part on the energetic supplement from polysaccharides because of the low-energy yield from lignin itself. When these polysaccharides are more rapidly depleted, as in the modified material, subsequent decay of the lignin could be retarded. Alternatively, the more condensed structure of the modified lignin may make it more difficult to degrade. Investigating these hypotheses fully will require long-term studies and will be challenging because of the difficulty, for example, of growing isotopically labeled trees to maturity. Following the same reasoning as Talukder, the consequence of modified lignin plants decomposing more slowly over the longer term would be that carbon would be held up in soils for a longer period during its passage around the biogeochemical carbon cycle, thereby potentially reducing the flux of CO 2 into the atmosphere. Before advocating the growth of trees with modified lignin as part of an atmospheric CO 2 mitigation strategy, we should recognize that the effects of lignin modification on decomposition have been fairly subtle, short-lived and detected only in controlled experiments in the laboratory and field. In the natural or semi-natural environment of forest plantations, the environmental variability of soils, hydrology and climate, the species and genotype of trees grown and their age at harvest, and other ‘natural’ variables, will have a much greater effect on decomposition and soil properties. Indeed, for the four-year field experiment of lignin-modified poplars, we showed that the differences in soil organic carbon and microbial biomass between samples taken from the experimental plots and those from the surrounding grassland were larger than those between soil samples from beneath wild-type and modified trees 5 . This is not surprising as it has been repeatedly demonstrated that plant species differ in the composition of the microbial communities around their roots and support different abundances of soil microbes 10 . Thus, introducing any new crop or tree to a soil is likely to have some effect on the local soil ecosystem. If this is not a concern for conventionally bred plants, why should Reduced COMT R 2 = 0.964 Wild type R 2 = 0.974 Reduced CAD R 2 = 0.977 0.003 0.004 0.005 0.006 0.007 0.008 0.009 100 150 200 250 300 350 400 450 Time (d) CO 2 production (mg C g –1 soil) Figure 1 CO 2 production from soil amended with stem material from unmodified tobacco and tobacco plants with modifications that reduce expression of either cinnamyl alcohol dehydrogenase (CAD) or caffeic acid O-methyl transferase (COMT). Soil was amended with 1% (wt/wt) air-dried and chopped plant material and incubated at 20 °C. Over the period from 112–403 days, the accumulating CO 2 was measured by gas chromatography 7 . Regression analyses are shown and the means (s.d.) for the slopes (rates) are as follows: tobacco plants with unmodified lignin (■) = 0.00518 (0.00032) μg C g –1 soil day –1 ; tobacco plants with reduced CAD (♦) = 0.00581 (0.00034) μg C g –1 soil day –1 ; and tobacco plants with reduced COMT (▲) = 0.00441 (0.00032) μg C g –1 soil day –1 . CORRESPONDENCE © 2007 Nature Publishing Group http://www.nature.com/naturebiotechnology