Metabolomic profiling of buffalo grasses (Stenotaphrum Secundatum L.) expressing a novel glufosinate resistant gene

Boonchaisri, S 2019, Metabolomic profiling of buffalo grasses (Stenotaphrum Secundatum L.) expressing a novel glufosinate resistant gene, Doctor of Philosophy (PhD), Science, RMIT University.

Document type: Thesis
Collection: Theses

Attached Files
Name Description MIMEType Size
Boonchaisri.pdf Thesis application/pdf 11.29MB
Title Metabolomic profiling of buffalo grasses (Stenotaphrum Secundatum L.) expressing a novel glufosinate resistant gene
Author(s) Boonchaisri, S
Year 2019
Abstract The aim of this research was to determine whether any perturbations occur in the metabolism of herbicide-resistant (HR) buffalo grasses (Stenotaphrum secundatum L.) compared to their wild type counterparts. The HR buffalo grasses were genetically modified by inserting the pat gene from the soil bacterium (Alcaligenes faecalis). This gene encodes phosphinothricin acetyltransferase (PAT) which is responsible for N-acetylation of the broad-spectrum herbicide, glufosinate. Four HR buffalo grasses (93-1A, 93-2B, 93-3C and 93-5A) were clearly resistant to glufosinate treatment above a commercial dose (5% v:v). Varying degrees of resistances among the 4 HR cultivars were observed and seemed to correlate with expression levels of the inserted pat gene.

The molecular phenotypes of 4 HR buffalo grasses (93-1A, 93-2B, 93-3C and 93-5A) were evaluated from gas chromatography-mass spectrometry (GC-MS) based metabolomics. Three wild type cultivars (WT 8-4A, WT 9-1B and WT 9-2B) were also analyzed to provide references for a comparison. Such approach detected a total 199 metabolites, of which 95 metabolites were identified and categorized into several classes of chemistries including amino acids, organic acids, fatty acids, sugars, sugar alcohols, sugar phosphates, sterols and other classes of compounds, most of which belong to carbon central metabolism. At baseline conditions (without glufosinate treatment), pair-wise comparisons between each of the 4 HR line (93-1A, 93-2B, 93-3C and 93-5A) and the parental wild type (WT 9-1B) did not detect consistent differences in the relative concentrations of metabolites across every sample. After glufosinate treatment, pair-wise comparisons of each HR line (93-1A, 93-2B, 93-3C and 93-5A) with WT 9-1B detected a significant difference in the relative abundance of 11, 4, 13 and 24 metabolites, respectively. Two groups were separated, using Principal Component Analysis (PCA). The first group included all 3 wild type cultivars and the weakest HR line (93-2B), after all cultivars were exposed to glufosinate treatments. The second group comprised of all baseline samples and the glufosinate-treated samples of 3 stronger HR lines (93-1A, 93-3C and 93-5A). Phenylalanine, isoleucine and galacturonate were the 3 strongest contributors to the observed separation in the PCA. The enrichment pathway analyses did not identify any significantly enriched pathways at baseline conditions when the 4 HR lines were compared with the WT 9-1B. However, when glufosinate treatments were applied, 5 metabolic pathways were significantly enriched (p < 0.05) [4 nitrogen metabolisms related pathways (aminoacyl-tRNA biosynthesis, alanine, aspartate and glutamate metabolism, valine, leucine and isoleucine biosynthesis and nitrogen metabolism) and 1 carbon metabolism-related pathway (glyoxylate and dicarboxylate metabolism)].

The same 4 HR and 3 WT buffalo grasses were further characterized from metabolomics profiles, using high resolution-liquid chromatography-mass spectrometry (HR-LC-MS). A total of 34,052 clusters were detected from mass spectra, representing metabolites and adducts from both primary and secondary metabolisms. Due to the extremely large amounts of the features detected, PCA was applied to reduce the dimension of the dataset. The results showed that samples were divided into 3 groups comprised of 1) all baseline samples, 2) 3 stronger HR lines (93-1A, 93-3C and 93-5A) exposed to glufosinate treatment and 3) 3 wild type cultivars as well as 93-2B exposed to glufosinate treatment. At baseline conditions, mummichog based pathway analyses revealed no significantly enriched pathways when comparing WT 9-1B to either the weakest HR line (93-2B) or the strongest HR line (93-5A). In contrast, glufosinate treatment caused phenylalanine metabolism and the TCA cycle to be significantly enriched (p < 0.05) as several metabolites (e.g. phenylpyruvic acid, trans-cinnamic acid, 4-coumaric acid, 2¿oxoglutaric acid, succinic acid, isocitric aicd, cis-aconitic acid, citric acid, pyruvic acid and oxaloacetic acid) were considerably accumulated in the glufosinate-sensitive WT 9-1B compared to the glufosinate-resistant 93-5A.   

Top-down and bottom-up proteomics analysis (TDP and BUP, respectively) of the strongest HR cultivar (93-5A) and the WT cultivar (WT 9-1B) were compared, using HR-LC-MS. A total 111 protein clusters (representing relatively small size intact proteins) were detected in TDP. PCA plots demonstrated that the proteins belonging to glufosinate-treated WT-9-1B were separated from other groups of samples (baseline 93-5A and WT 9-1B as well as glufosinate-treated 93-5A). TDP clustering patterns from PCA corresponded well with the PCA constructed from the glufosinate-induced senescence scores, suggesting a relationship between senescence mechanisms and TDP compositions. For BUP, a total of 28,095 peptide clusters (representing proteins fragments from numerous proteins in every isoform) were detected. Based on the relative abundance of clusters, minimal numbers of peptide clusters (3.45%) were significantly different (p < 0.01) between 93-5A and WT 9-1B at baseline conditions, resulting in these 2 cultivars clustering together in the PCA scores plot. With glufosinate treatment, 5.46% of peptide clusters showed significant differences in the relative abundances between 93-5A and WT 9-1B. Although the PCA of the glufosinate-treated WT9-1B line showed a distinct separation from glufosinate-treated 93-5A, some degree of overlapping between the 2 cultivars was noticed due to high variations in BUP composition resulted from variability in senescence stage between different replications. Analyses of the relationships between significantly different metabolites and proteins obtained from a pair-wise comparison between 93-5A and WT 9-1B detect neither protein-protein nor metabolite-protein interactions under baseline conditions. The same analyses of interaction between metabolites and proteins from glufosinate-treated 93-5A and WT 9-1B revealed 5 protein-protein interactions and 3 metabolite-protein interactions, thus indicating alterations in both protein and central carbon metabolisms of the WT 9-1B in response to glufosinate treatment.

In summary, this thesis shows that the detectable metabolome and proteome of 4 HR lines (93-1A, 93-2B, 93-3C and 93-5A) were highly similar to that of 3 wild type cultivars (WT 8-4A, WT 9-1B and WT 9-2B) at baseline conditions. These findings support the conclusion that these HR buffalo grasses are `substantially equivalent' to the parental wild type comparators.  Expected changes in both the metabolome and proteome in the herbicide sensitive wild type cultivars were associated with glufosinate-induced senescence. Hence, the results produced in this research provide additional knowledge necessary for informing regulatory bodies, by providing them with a framework of evidence which strongly recommends the inclusions of multi-omics analyses in the assessment of transgenic plants.
Degree Doctor of Philosophy (PhD)
Institution RMIT University
School, Department or Centre Science
Subjects Plant Biology not elsewhere classified
Keyword(s) Metabolomics
Stenotaphrum secundatum
Herbicide resistance
Version Filter Type
Access Statistics: 45 Abstract Views, 30 File Downloads  -  Detailed Statistics
Created: Tue, 24 Mar 2020, 08:54:27 EST by Keely Chapman
© 2014 RMIT Research Repository • Powered by Fez SoftwareContact us