The genetics of glyphosate resistance in barnyard grass, fleabane, windmill grass and feathertop Rhodes grass

Take home message

Three out of four species of glyphosate-resistant northern-region weeds have non-target site resistance mechanisms. Only feathertop Rhodes grass (which is not considered resistant) has the Pro106Ser target site mutation.

Introduction

Herbicide resistance mechanisms are broadly categorised as either “target site” or “non-target site” mechanisms. Target site resistance mechanisms are caused by changes to the gene/protein that is targeted by a particular herbicide mode of action (MoA). These targets are well characterised and therefore target site resistance mechanisms are relatively easy to detect genetically. Target site mechanisms are also generally single gene traits so their inheritance is relatively easy to predict. Further, target site mechanisms are, by definition, specific to one MoA. This type of resistance therefore cannot result in cross resistance between herbicide MoA’s. On the other hand; every other type of resistance mechanism is categorised as “non-target site resistance” (NTSR). This can include anything from additional waxiness of the plant cuticle through to detoxification pathways and various transportation pathways. “Non-target site” resistance can involve multiple genes, have complex inheritance, and many NTSR mechanisms of small effect can be present in one plant line. This makes the identification of the genetic changes underlying NTSR mechanisms difficult to elucidate by traditional methods. NTSR mechanisms can also potentially confer cross-resistance to more than one herbicide MoA, and are therefore undesirable from a management perspective.

From a theoretical perspective high rates of herbicide application are more likely to select for target site resistance, and low rates are more likely to select for NTSR. There is evidence to support this prediction in metabolism-based NTSR to diclofop in annual ryegrass. Glyphosate can be metabolised by some plants, but the by-products are toxic and no field resistance has been explained by metabolism to date. The structure of the target site (EPSPS gene) of glyphosate also means that relatively few mutations are able to cause target site resistance in the field. The structure of the EPSPS protein may therefore partially explain the relatively long time it has taken for target site resistance to Roundup to develop, as it reduces the chance of developing target site resistance regardless of herbicide rate.

This paper will discuss the ways new gene sequencing technology has been used to assess glyphosate resistance in four species of weeds; barnyard grass, fleabane, windmill grass and feathertop Rhodes grass.

Results and discussion

The only species found where resistance was fully explained by a target site mutation was feathertop Rhodes grass. This species is regarded to be tolerant rather than resistant, however, from a genetic point of view it is resistant as the target site mutation would have only spread throughout the population in response to the selection pressure imposed by glyphosate. Barnyard grass also had a target site mutation in one population (Table 2). This mutation was only present in one of the three genomes (the species is polyploid), and only in one of the two lines that showed strong resistance. This species therefore shows a mix of target site and NTSR resistance mechanisms across different populations. Fleabane and windmill grass showed no evidence of target site mutations, indicating that both are resistant to glyphosate through NTSR mechanisms. Fleabane has an ancestral duplication of the EPSPS gene; however, this does not explain resistance as it is present in both resistant and susceptible individuals, and across multiple species in the genus. Windmill grass shows some preliminary evidence of additional copies (4x) of the EPSPS gene, this could partially explain resistance but requires further investigation.

Table 1. Mutations present at the EPSPS target site gene for each of the 12 individual plants sequenced, the two mutations that confer glyphosate resistance were scored and all lines were susceptible at the target site with the exception of QBG41.

Gene expression changes in response to glyphosate treatment in fleabane and barnyard grass were also assessed through two transcriptome experiments. In barnyard grass many more genes show different expression patterns in the strong resistant lines compared to the lines with intermediate resistance (table 2.). This indicates that many genes could be involved with NTSR mechanisms in barnyard grass. In fleabane relatively few genes were differentially expressed (DE) following removal of genes that were DE in susceptible lines, indicating that NTSR in fleabane could be caused by few genes. The genes that were up-regulated in fleabane mostly match to transporter genes.

Table 2. Number of differentially expressed genes in each of the resistant lines of barnyard grass

Conclusions

Non-target site resistance mechanisms to glyphosate are widespread in the northern region. Understanding the risk of cross resistance requires that the specific NTSR mechanism is known. This used to be very difficult with traditional genetic approaches but new technology and methods are now allowing us to unravel the genetic basis of NTSR mechanisms.

Acknowledgements

This work is supported by the CRDC (Cotton Research and Development Corporation).

Contact details

The University of Queensland 
Goddard building, UQ St. Lucia, 4072, Brisbane.
Ph: 07 3365 2769
Email: j.hereward@uq.edu.au