The identification of the gene regulatory network that controls the formation of xylem — the major component of wood — opens up new avenues for manipulating plant biomass. See Article p.571
Cellulose, hemicelluloses and lignin are key natural polymers that make up the bulk of plant biomass1. These biopolymers are also renewable resources for the production of dietary fibre, paper and biofuels2. On page 571 of this issue, Taylor-Teeples et al.3 report the identification of the gene regulatory network that controls the synthesis of these biopolymers in root xylem cells of the model plant Arabidopsis thaliana.
Xylem is a plant tissue that provides mechanical support and the main mechanism for transporting water and nutrients from root to shoot tissues. To perform these important functions, xylem cells deposit a specially reinforced structure termed the secondary cell wall1 (Fig. 1). Xylem secondary cell walls are composed mainly of cellulose, hemicelluloses and lignin. The cellulose forms a network of load-bearing fibres coated in hemicelluloses and embedded in lignin, providing mechanical strength and rigidity (akin to steel rods set in reinforced concrete). However, the presence of lignin is a major impediment to efficient extraction of the sugars in cellulose and hemicelluloses for their conversion to biofuels2. Hence, understanding how the relative proportions of these biopolymers are controlled in plant tissue would open up opportunities to redesign plants for biofuel use.
Xylem cells control the relative abundance of biopolymers in part by regulating expression of the genes that encode the enzymes for polymer synthesis4. Expression is controlled by transcription-factor proteins that bind DNA sequences, termed promoters, close to the genes. A handful of transcription factors have been identified that control the expression of individual genes regulating the production of cellulose, hemicelluloses and lignin. But this small-scale, gene-by-gene approach has provided a highly fragmented picture of the potential regulatory interactions between xylem-associated transcription factors and their gene targets.
Taylor-Teeples et al. instead adopted a network approach — screening more than 460 transcription factors expressed in the root xylem of A. thaliana for their ability to bind the promoters of around 50 previously characterized genes that encode cell-wall components or other transcription factors involved in xylem formation. This large-scale analysis provided a remarkable overview of the regulatory process, revealing a highly interconnected network composed of some 240 genes and more than 600 new protein–DNA interactions.
The xylem regulatory network shows that each cell-wall gene is bound, on average, by 5 different transcription factors, each belonging to one of 35 distinct families of regulatory proteins. This regulatory arrangement provides a huge number of combinatorial possibilities, which Taylor-Teeples and colleagues show is crucial for integrating environmental signals such as salt or iron stress — alterations in the expression of certain transcription factors allowed different sub-networks to be used to adapt the cellular response to these conditions.
The network also reveals that many of the transcription factors are not part of simple linear pathways, but form a series of feed-forward loops (FFLs). Such regulatory systems are well recognized in systems biology, and typically involve a transcription factor that controls the expression of other transcription factors, which then collectively co-regulate their target genes. For example, the authors find that the transcription factor E2Fc binds to more than 20 promoters, including those for the genes encoding the transcription factors VND6, VND7 and MYB46, as well as genes associated with cellulose, hemicellulose and lignin production. Although FFLs are common in biological systems5, they are remarkably numerous in the xylem network, occurring close to 100 times. They are also frequently embedded within one another, creating FFL cascades. For example, the network shows that VND7 and MYB46 also bind to the promoters of many E2Fc target genes.
So why are there so many FFLs? Not only are there many possible components, but even for simple systems with only three components, there are many possible ways to wire them6. Common to all FFLs is a direct path (in which a source transcription factor regulates a target gene) and an indirect path (in which the same source factor regulates an intermediate transcription factor that regulates the same target). For 'coherent' FFLs, the direct and indirect paths have the same overall effect on the target gene (both activate or repress its expression), whereas for incoherent FFLs, one path activates and the other represses. Mathematical modelling of these loops has revealed that different arrangements can produce a range of responses from target genes. For example, coherent FFLs can protect against unwanted responses to fluctuations in inputs, whereas incoherent FFLs can speed up transcriptional responses6. In the case of the xylem network, a coherent FFL could result in tight regulation of cell-wall gene expression, thereby promoting secondary cell-wall synthesis in a switch-like manner to prevent the deposition of secondary cell-wall material in non-xylem cells.
It is not yet possible to determine exactly what types of FFL are present in the xylem regulatory network described by Taylor-Teeples et al., because although the technology used by the authors identifies interactive nodes, it cannot predict whether they relate to transcriptional activation or repression. However, these nodes provide a framework for future research to characterize key interactions in a targeted, gene-by-gene manner, and to determine the precise regulatory structure. This will allow the identification of ways to manipulate this network to engineer different cellular properties and develop new plant varieties for biofuel use. The description of the network also helps to explain why plant transcription factors have so far largely eluded identification by genetic screens, owing to functional redundancy among regulators of secondary-cell-wall biosynthesis. This knowledge can now be used to perform more-precisely targeted screens of gene function, by creating combinations of mutations that overcome this genetic redundancy.