Overexpression of PvWOX3a in switchgrass promotes stem development and increases plant height

Abstract

Switchgrass (Panicum virgatum L.) is an important perennial, noninvasive, tall ornamental grass that adds color and texture to gardens and landscapes. Moreover, switchgrass has been considered a forage and bioenergy crop because of its vigorous growth, low-input requirements, and broad geography. Here, we identified PvWOX3a from switchgrass, which encodes a WUSCHEL-related homeobox transcription factor. Transgenic overexpression of PvWOX3a in switchgrass increased stem length, internode diameter, and leaf blade length and width, all of which contributed to a 95% average increase in dry weight biomass compared with control plants. Yeast one-hybrid and transient dual-luciferase assays showed that PvWOX3a can repress the expression of gibberellin 2-oxidase and cytokinin oxidase/dehydrogenase through apparently direct interaction with their promoter sequences. These results suggested that overexpression of PvWOX3a could increase gibberellin and cytokinin levels in transgenic switchgrass plants, which promotes cell division, elongation, and vascular bundle development. We also overexpressed PvWOX3a in a transgenic miR156-overexpressing switchgrass line that characteristically exhibited more tillers, thinner internodes, and narrower leaf blades. Double transgenic switchgrass plants displayed significant increases in internode length and diameter, leaf blade width, and plant height but retained a tiller number comparable to that of plants expressing miR156 alone. Ultimately, the double transgenic switchgrass plants produced 174% more dry-weight biomass and 162% more solubilized sugars on average than control plants. These findings indicated that PvWOX3a is a viable potential genetic target for engineering improved shoot architecture and biomass yield of horticulture, fodder, and biofuel crops.

Fig. 1. Molecular characterization of PvWOX3a.

a Neighbor-joining phylogenetic tree of WOX3a-related proteins from Panicum virgatum, Arabidopsis thaliana, Medicago truncatula, and Oryza sativa. The tree was constructed from an alignment conducted using MEGA X with 1000 bootstrap replicates, and the bootstrap values of each node are shown on the tree. b Subcellular localization of the WOX3a-eGFP fusion reporter in N. benthamiana cells by confocal laser microscopy. GFP, bright field, and merged images are shown. Scale bar = 50 μm. c Expression patterns of PvWOX3a in switchgrass. E4I2 Internode 2 at the E4 stage, E4L2 Leaf 2 at the E4 stage, E3I2 Internode 2 at the E3 stage, E3L2 Leaf 2 at the E3 stage, Inflorescence; and Crown buds. qRT–PCR was normalized to the expression of switchgrass PvUbq2. Values are means ± SEs (n = 3)

Fig. 2. Morphological characterization of PvWOX3a-overexpressing transgenic plants.

a The expression levels of PvWOX3a in transgenic lines revealed by qRT–PCR. Switchgrass PvUbq2 was used for normalization. Values are means ± SEs (n = 3). b Gross phenotypic characterization of switchgrass plants overexpressing PvWOX3a (WOX3aOE). Control plants carried the pANIC6B empty vector. Scale bar = 5 cm. Leaf 3 at the E4 stage (c) and Internode 3 at the E4 stage (d) of control and WOX3aOE transgenic plants are shown. Scale bar = 5 cm. Three-month-old tillers were used to measure plant height (e), diameter of Internode 3 (f), length of Leaf 3 (g), and width of Leaf 3 (h). Three tillers from the same plant were measured for each replicate. Fresh weight biomass yield (i) and dry-weight biomass yield (j) of transgenic switchgrass plants. The control plants and WOX3aOE transgenic plants were harvested after 4 months of growth in the greenhouse. Values are means ± SEs (n = 6). Asterisks represent significant differences determined by Student’s t test. ****p < 0.0001; ***p < 0.0002

Fig. 3. Overexpression of PvWOX3a in switchgrass promotes cell proliferation and vascular development and affects cell wall composition.

a Cross sections of control and WOX3aOE internode bases of Internode 2 at the E4 stage. Scale bar = 200 μm. b Cell numbers were determined by counting along the radius at the 12 o’clock position from the outer edge of the pith to the epidermis in cross sections excised from the base of Internode 2 at the E4 stage of control and WOX3aOE transgenic plants. Values are means ± SEs (n = 9). c Acid-insoluble lignin content of three independent WOX3aOE lines and control plants. Hemicellulose content (d) and cellulose content (e) of three independent WOX3aOE lines and control plants. Values are means ± SEs (n = 3). Asterisks represent significant differences determined by Student’s t test. ****p < 0.0001; **p < 0.0021; *p < 0.0332; ns means no significance

Fig. 4. Overexpression of PvWOX3a downregulated GA2ox and promoted cell elongation in switchgrass.

a Longitudinal section of Internode 2 at the E3 stage of control and WOX3aOE transgenic plants. Scale bar = 100 μm. b Cell lengths based on longitudinal sections of E3I2 of control and WOX3aOE transgenic plants. Values are means ± SEs (n = 18). c The expression levels of PvGA2ox3 and PvGA2ox7 in three WOX3aOE lines were revealed by qRT–PCR. Switchgrass PvUbq2 was used for normalization. Values are means ± SEs (n = 3). d Growth of yeast cells on SD/-Trp-Leu-His supplemented with 100 mM 3-AT. pHIS2.1-ProGA2ox3 plus pGADT7 and pHIS2.1-ProGA2ox7 plus pGADT7 served as the negative controls. e Dual-luciferase assay showing the repression of PvGA2ox3 and PvGA2ox7 by the PvWOX3a effector construct compared to the control effector construct. Values are means ± SEs (n = 3). qRT–PCR analysis of PvWOX3a (f), PvGA2ox3 (g), and PvGA2ox7 (h) expression levels in wild-type switchgrass plants treated with 200 μm GA₃. Switchgrass PvUbq2 was used for normalization. Values are means ± SEs (n = 3). i qRT–PCR analysis of endogenous PvWOX3a expression levels in control plants and WOX3aOE lines. Values are means ± SEs (n = 3). The asterisks represent significant differences in b and e, as determined by Student’s t test. ****p < 0.0001; **p < 0.0021. Asterisks represent significant differences in c, f, g, and h as determined by one-way ANOVA. *p < 0.0332; ***p < 0.0002; ****p < 0.0001

Fig. 5. Overexpression of PvWOX3a in a miR156-overexpressing transgenic line.

a Gross phenotypic characterization of control plants, miR156-overexpressing transgenic plants (miR156OE-27), and overexpression of WOX3a in miR156OE-27 transgenic plants (miR156OE_WOX3aOE). Scale bar = 5 cm. Internode 3 at the E4 stage (b) and Leaf 3 at the E4 stage (c) of control, transgenic miR156OE-27, and transgenic miR156OE_WOX3aOE plants are shown. Scale bar = 5 cm. Three-month-old tillers were used to measure plant height (d), internode number (e), tiller number (f), and internode diameter (g). Three tillers per plant were measured for each replicate. Values are means ± SEs (n = 3–7). The letters above the error bars indicate significant differences determined by one-way ANOVA (p < 0.05, Duncan’s multiple-range test)

Fig. 6. The double transgenic lines showed higher biomass and solubilized sugar yields than the control plants.

Comparison of postharvest fresh (a) and dry (b) weights of total above-ground biomass of control and miR156OE_WOX3aOE lines after four months of growth in the greenhouse. Values are means ± SEs (n = 4–9). Enzymatic hydrolysis efficiency (c) and solubilized sugar yields (d) of three miR156OE_WOX3aOE independent lines compared to control plants. Values are means ± SEs (n = 3). The asterisks represent significant differences determined by Student’s t test. **p < 0.0021; ****p < 0.0001