Amaranthus retroflexus, L.

Weston, Paul A., Gurusinghe, Saliya, Birckhead, Emily, Skoneczny, Dominik, Quinn, Jane C. & Weston, Leslie A., 2019, Chemometric analysis of Amaranthus retroflexus in relation to livestock toxicity in southern Australia, Phytochemistry 161, pp. 1-10 : 4

2.2. Cytotoxicity of methanolic extracts of A. retroflexus

Exposure to A. retroflexus leaf extracts induced rapid cell death in NIH3T3 cells within two hours following treatment, with cells exhibiting typical apoptotic features including nuclear condensation and cell shrinkage ( Fig. 5 ). Control cultures treated with equivalent volumes of methanol (MeOH) showed very limited evidence of apoptosis, displaying characteristic fibroblast morphology and active cell division.

Exposure of NIH3T3 cells to extracts of A. retroflexus leaves from all three plant collections resulted in significantly higher cytotoxicity than extracts of other tissues obtained from the same plants. Cytotoxicity of leaf tissue extracts was observed at both concentrations examined (0.25 and 0.50 mg /ml) ( Fig. 6 ).

2.3. Identification of the toxic constituents of Amarathus retroflexus extracts

Despite the predominance of numerous nitrogen-containing compounds and chlorophyll derivatives in methanolic extracts of A. retroflexus , these compounds were not significantly or positively correlated with cytotoxicity. Stepwise linear regression using the relative abundance of compounds against in vitro cytotoxicity revealed no appreciable correlation between any components separated via HILIC chromatography and bioactivity. However, the cytotoxic activity in A. retroflexus leaf extracts was highly correlated at both concentrations examined [0.25 mg /mL (r 2 = 0.928, P <0.001) and 0.5 mg /mL (r 2 = 0.840, P <0.001)] with the presence of a single compound eluting on C 18 with m/z of 824.3300 ( Fig. 7 ). Given the very high coefficient of determination (r 2) and the statistical significance of this match, we attempted to determine the structure of this potential toxin or toxicity biomarker by examining MS-MS spectra. The predicted molecular formula of this constituent was C 35 H 49 N 7 O 16. The use of Agilent Molecular Structure Correlator (MSC) applied to MS-MS spectra of this compound yielded a predicted 83% match with a small modified peptide, N-acetyl-L-α- aspartyl-L-alanyl-L-α- aspartyl-L-α- glutamyl-O- (carboxymethyl)-L-tyrosyl-L-leucinamide (structure 1a, Fig. 8 ). Two related modified peptides (1b and 1c, Fig. 8 ) also produced high scores via MSC, but their scores were somewhat lower (81 and 79%, respectively). The various software packages and libraries used for assistance with structural elucidation indicated that all candidate molecules as potential matches for the putatively active compound contained nitrogen and were small modified peptides. The MS-MS spectra of the putatively active compound under collision energies of 10, 20 and 40 eV is shown in Fig. S1.

Figures

Fig. 5. Typical morphology of mouse fibroblast cell line NIH3T3 48 h after treatment with Amaranthus retroflexus extract. Methanol-only treatments served as a negative control for each concentration. Scale bar = 50 μm.

Fig. 6. Cell viability relative to methanol-only vehicle control following treatment with extracts of Amaranthus retroflexus collected on three separate occasions in a) Wagga Wagga, 2014; b) Wagga Wagga, 2015 and c) Uranquinty, 2015. Values are means ± SEM. Cell viability was significantly lower in leaf extracts at both 0.25 mg/ml and 0.50 mg/ml compared to other tissue types and vehicle controls (*, P <0.05).

Fig. 7. Relationship between the abundance of the putative active compound from Amaranthus retroflexus extracts (m/z 824) and cytotoxicity as measured by inhibition of growth of murine fibroblast cells.

Fig. 8. Structure of the best three matches to the MS-MS spectrum of the putative cytotoxic compound from Amaranthus retroflexus extracts. The best match was with 1a (N-acetyl-L-α-aspartyl-L-alanyl-L-α-aspartyl-L-α-glutamyl-O-(carboxymethyl)-L-tyrosyl-L-leucinamide).

Fig. 1. Analysis of Amaranthus retroflexus plant material for crude protein (%), total N (%) and nitrate N (mg/kg) in inflorescence/seeds, leaves and stems/ shoots.

Fig. 2. Nitrate content of Amaranthus retroflexus inflorescence, stem, leaf and root tissues collected from two field locations in Wagga Wagga and Uranquinty, NSW in 2015 as measured by ion exchange and HPLC. Nitrate concentration is expressed in units of mg per g of extract; values are means ± SD. Results of factorial ANOVA is shown in the figure; * = significance at P <0.05, *** = significance at P <0.001.

Fig. 3. Relative abundance of nitrogen (N)-containing (black bars), chlorophyll (Chl) derivatives (medium grey bars) and other plant metabolites (light grey bars) in crude extracts of four tissue types of Amaranthus retroflexus analysed by HILIC chromatography interfaced with quadrupole time-of-flight mass spectrometry (HPLC-MS-QTOF).