MiR396-GRF module associates with switchgrass biomass yield and feedstock quality

Abstract

Improving plant biomass yield and/or feedstock quality for highly efficient lignocellulose conversion has been the main research focus in genetic modification of switchgrass (Panicum virgatum L.), a dedicated model plant for biofuel production. Here, we proved that overexpression of miR396 (OE-miR396) leads to reduced plant height and lignin content mainly by reducing G-lignin monomer content. We identified nineteen PvGRFs in switchgrass and proved thirteen of them were cleaved by miR396. MiR396-targeted PvGRF1, PvGRF9 and PvGRF3 showed significantly higher expression in stem. By separately overexpressing rPvGRF1, 3 and 9, in which synonymous mutations abolished the miR396 target sites, and suppression of PvGRF1/3/9 activity via PvGRF1/3/9-SRDX overexpression in switchgrass, we confirmed PvGRF1 and PvGRF9 played positive roles in improving plant height and G-lignin content. Overexpression of PvGRF9 was sufficient to complement the defective phenotype of OE-miR396 plants. MiR396-PvGRF9 modulates these traits partly by interfering GA and auxin biosynthesis and signalling transduction and cell wall lignin, glucose and xylan biosynthesis pathways. Moreover, by enzymatic hydrolysis analyses, we found that overexpression of rPvGRF9 significantly enhanced per plant sugar yield. Our results suggest that PvGRF9 can be utilized as a candidate molecular tool in modifying plant biomass yield and feedstock quality.

Keywords: GRF; biomass yield; feedstock quality; miR396; switchgrass.

Figure 1

Phenotype comparison of the R3 stage switchgrass WT and OE‐miR396 plants. (a) Photograph of typical WT and OE‐miR396 plants. (b) Relative expression of miR396 revealed by quantitative RT‐PCR (n = 3). A nuclear switchgrass small RNA gene, U6 was used as an internal control. (c) Comparison of plant height (n = 4). (d, e) The average internode length (d) and number (e) (n = 4). (f) The bottom internode diameter of stem (n = 4). The data shown in (c), (d), (e) and (f) are the means of four biological replicates (with twenty technical repeats each) ± SD. (g) Comparison of dry biomass yield of WT and OE‐miR396 plants (n = 4). The error bar indicates standard deviation. The different letters indicate statistically significant differences determined by Duncan’s multiple range test (P < 0.05).

Figure 2

Stem cell wall components analysis of WT and OE‐miR396 lines. (a) Phloroglucinol‐HCl staining assay of lignin in 1NE3 cross sections of WT and OEs. The red coloration indicates the presence of lignin. c, collenchyma; vb, vascular bundle cells. Scale bar = 50 μm. (b) The Klason lignin of CWR of WT and OEs (n = 3) with five technical repeats each. (c, d) AcBr lignin content (c) and the lignin unit content (d) of CWR of WT and OEs (n = 2). G, guaiacyl; S, syringyl; H, p‐hydroxyphenyl. (e, f) The glucose (e) and xylose (f) content of CWR of WT and OEs (n = 3) with five technical repeats each. The error bar indicates standard deviation. Different letters and asterisks represent significant differences (P < 0.05).

Figure 3

Validation and expression patterns of the miR396 target genes. (a) 5’ RLM‐RACE experimental validation of the miR396a cleavage sites (arrows) of the miR396 putative target genes. (b) The relative expression of the miR396‐target PvGRFs in the stem cells (S1) of WT and OE‐miR396 lines revealed by qRT‐PCR analysis (n = 3). (c) The expression patterns of the miR396‐target PvGRFs in different tissues of switchgrass (n = 3). An illustrated photograph of the sampling site is shown in Figure S1b. R, rachilla; M, lemma; P, pistil; T, stamen; stem (S1, S2); leaf (L1, L2); S1 and S2 were sampled from the second internode of the E3 stage tiller; L1 and L2 were sampled 1 cm from the base of the leaves. The error bar indicates standard deviation. Different letters represent significant differences (P < 0.05).

Figure 4

PvGRF9 positively regulates plant height and lignin content. (a,b) Photograph of typical WT and PvGRF9‐SRDX transgenic (9sr) plants (a) and rPvGRF9 transgenic (r9ox) plants (b). (c) Relative expression of PvGRF9 in WT, 9sr and r9ox lines (n = 3). Asterisk indicates statistically significant differences between WT and TG plants (P < 0.05). (d, e) Comparison of the plant height (d), average internode length (e) of WT, 9sr and r9ox lines (n = 4). (f) The dry biomass yield of per plant stems of WT, 9sr and r9ox lines (n = 4). (g) Comparison of Klason lignin content (n = 3) with five technical repeats each. (h) Comparison of lignin monomer content (n = 2). (i) Comparison of the glucose yield of the stem cell wall residues of WT, 9sr and r9ox lines (n = 3) with five technical repeats. The error bar indicates standard deviation. The different letters indicate statistically significant differences determined by Duncan’s multiple range test (P < 0.05).

Figure 5

Overexpression of PvGRF9 or rPvGRF9 in OE‐miR396 line, OE17 rescued impaired plant phenotype. (a) A photograph of the R3 stage tillers of WT, OE17 and eight representative OE17 plants overexpressing PvGRF9 (9ox/OE17), or rPvGRF9 (r9ox/OE17), respectively. The white arrow point to the inflorescence node position. (b) Relative expression of PvGRF9 in WT, OE17, 9ox/OE17 and r9ox/OE17 by RT‐PCR (n = 3). (c) Comparison of plant height of WT, OE17 and (r)9ox/OE17 (n = 4) with twenty technical replicates each. (d) The stem dry biomass yield per tiller after culturing six months in greenhouse (n = 4). (e, f) Klason lignin content (e) and glucose content (f) of cell wall residues of WT, OE17 and (r)9ox/OE17 (n = 3) with five technical repeats each. The error bar indicates standard deviation. The different letters indicate statistically significant differences (P < 0.05).

Figure 6

Expression analyses of miR396, miR396‐targeted PvGRFs and other related genes in WT and transgenic plants. (a) Relative expression analysis of miR396 and miR396‐targeted PvGRFs in WT, OE17 and (r)9ox/OE17 plants (n = 3). The error bar indicates standard deviation. The different letters indicate statistically significant differences (P < 0.05). (b) Heat map of relative expression levels of the genes in GA and auxin biosynthesis and signalling transduction, lignin biosynthesis, glucose and xylan biosynthesis in WT and transgenic plants. The qRT‐PCR tested data were subjected to log₂‐fold change (n = 2) with three technical repeats for each.

Figure 7

Effects of PvGRF9‐SRDX and rPvGRF9 overexpression on cell wall saccharification. (a) Enzymatic hydrolysis efficiency of cell wall residues of the transgenic plants. (b) The per plant contents of glucose and xylan released from WT and transgenic plants after pretreatment. The data are shown as the means of three biological replicates (with five technical repeats each) ± SD. The different letters and asterisks indicate statistically significant differences (P < 0.05).