ScGAI is a key regulator of culm development in sugarcane

Abstract

Sugarcane contributes more than 70% of sugar production and is the second largest feedstock for ethanol production globally. Since sugar accumulates in sugarcane culms, culm biomass and sucrose content are the most commercially important traits. Despite extensive breeding, progress in both cane yield and sugar content remains very slow in most countries. We hypothesize that manipulating the genetic elements controlling culm growth will alter source-sink regulation and help break down the yield barriers. In this study, we investigate the role of sugarcane ScGAI, an ortholog of SLR1/D8/RHT1/GAI, on culm development and source-sink regulation through a combination of molecular techniques and transgenic strategies. We show that ScGAI is a key molecular regulator of culm growth and development. Changing ScGAI activity created substantial culm growth and carbon allocation changes for structural molecules and storage. ScGAI regulates spatio-temporal growth of sugarcane culm and leaf by interacting with ScPIF3/PIF4 and ethylene signaling elements ScEIN3/ScEIL1, and its action appears to be regulated by SUMOylation in leaf but not in the culm. Collectively, the remarkable culm growth variation observed suggests that ScGAI could be used as an effective molecular breeding target for breaking the slow yield gain in sugarcane.

Fig. 1.

ScGAI encodes the nuclear DELLA protein in sugarcane. (A) Top, schematic drawing of ScGAI protein showing all the conserved domain along the sequence; bottom, protein alignment of DELLA domain highlighting the identical amino acids among the sequences. DELLA and TVHYNP amino acids are underlined. (B) Cartoon (top) and surface (bottom) representation of the predicted tertiary structure of DELLA domain from ScGAI. The electrostatic surface is represented by regions negatively charged (red), positively charged (blue), polar (dark gray) and hydrophobic (light gray). The overlap between predicted and native structure has a root-mean-square deviation (RMSD) value of 0.091. (C) Phylogenetic tree of DELLA family in Arabidopsis, tomato, pea, grape, barley, wheat, rice, maize, sorghum, and sugarcane. The red branch of the tree is the conserved DELLA family in monocotyledonous plants. (D) Subcellular localization of ScGAI expressed in Arabidopsis mesophyll protoplast. The construct AtPARP3:mCHERRY (mCHERRY) was used as nuclear control. DIC, differential interference contrast; YFP, yellow fluorescent protein. Scale bars, 20 µm.

Fig. 2.

ScGAI is SUMOylated in sugarcane leaves. (A) Expression profile of native ScGAI in different tissues of 10-month-old sugarcane; bar plots show means ±SD of three biological replicates. (B) Immunoblotting of ScGAI protein in sugarcane tissues. (C) Sequence alignment of the non-canonical SUMOylation motif in DELLA proteins. Asterisk represents the conserved SUMOylation site lysine residue. (D) Sequence alignment of GID1 from rice, wheat, maize, sorghum, Arabidopsis, and sugarcane displaying the SUMO-interacting motif (SIM). Light gray depicts the conserved amino acids among the sequences. (E) Expression profile analysis of ScSIZ1, ScSUMO1, and ScOTS1 transcripts in leaf +1 (L+1), apical shoot, and fith and ninth internodes. FPKM, fragment per kilobase of exon per million fragments mapped. Bars show means ±SD of three biological replicates. (F) Immunoprecipitation using anti-SUMO1 antibodies in crude extract of leaf +1. (G) The leaf numbering system proposed by Kuijper (1915). The first fully expanded leaf with visible dewlap (indicated by an arrow) and photosynthetically active was considered as leaf +1. (H) Close-up view of juvenile leaf L−2. (I) Immunoblotting of the ScGAI protein in different sections of juvenile (L0, L−1 and L−2) and fully expanded (L+1) leaves of Q208 (1 month old). The arrow indicates the non-SUMOylated ScGAI protein. Equal amounts of protein samples (10 µg) were loaded. CB, Coomassie Blue-stained membrane as loading control. B, base; M, middle; T, tip.

Fig. 3.

ScGAI-misexpressing sugarcane lines. (A) Phenotype of ScGAIOE and HpScGAI lines showing the stunted and taller stems, respectively. The earlier onset of elongated internodes in HpScGAI is numbered from the soil to the top. Arrows indicate the first visible dewlap. (B) Height of 3-month-old sugarcane plants. The data points represent means ±SD of three biological replicates. Red lines U1 and U3 represent untransformed control plants. Untransformed control plants were produced through all the tissue culture and transformation steps used for generating transgenic plants but without the introduction of transgene. The group comprising all HpScGAI lines exhibited significantly higher values for height compared with the group of ScGAIOE lines (unpaired one-tailed t-test, P<0.01). The gray asterisks indicate significance (P<0.05) for unpaired one-tailed t-test between the group comprising ScGAIOE lines FR28 to FR2 and untransformed control plants; the black asterisks indicate significance (P<0.05) for the comparison between the group comprising HpScGAI lines HR26 to HR37 and untransformed control plants.

Fig. 4.

ScGAI regulates tillering and culm development in sugarcane plants. (A) 3-month-old transgenic FR10 (ScGAIOE; dwarf), HR1 (HpScGAI; tallest), and untransformed lines. Height, internode number, and elongation and tiller number; bars show means ±SD of eight biological replicates. (B) Zoomed-in detailed view of 3-month-old plants. Internode numbers counted from soil to top. Arrows indicate the first visible dewlap. (C) Immunoblotting using a sugarcane anti-DELLA (anti-ScGAI) antibody. Each lane was loaded with 20 µg of total protein from leaf +1, apical shoot (shoot) and fifth internode (5th int) tissues of untransformed control, FR10, and HR1 lines of 6-month-old plants; CB, Coomassie Blue.

Fig. 5.

Carbon balance is severely altered in transgenic plants. Metabolite-based clustering of leaves (L+1) and fifth and ninth internodes in ScGAIOE (FR10 line) and HpScGAI (HR1 line) compared with untransformed control. The intensities are color-coded. Red represents high and blue represents low intensities. The statistical significance of metabolite variation was determined by comparing the data from a given tissue from all genotypes by Tukey’s test (Supplementary Dataset S1).

Fig. 6.

ScGAI interacts with ScPIF3/4 and ScEIN3/EIL1 in sugarcane. (A) Structure of the sugarcane DELLA ScGAI and its truncated versions used in the screening. Protein schematic comparison between AtEIN3 and ScEIN3 sequences and the protein truncation ScEIN3(233–522) used in the yeast two-hybrid assay. (B) Auto-activation activity of the different bait constructs in yeast cells. Full-length DELLA was capable of activating the transcription of reporter genes in the absence of prey proteins and also was toxic upon expression in yeast cells. (C, D) Co-transformations with different combinations were performed. On SD-Leu-Trp medium, diploid yeast cells were confirmed. On SD-Leu-Trp-Ade-His medium, only positive yeast cells for protein–protein interaction grew. AD, activation domain; BD, binding domain; pGBD and pGAD, empty vectors. Positive controls: 53-BD encodes murine p53 and T-AD encodes the SV40 large T-antigen. Negative control: Lam-BD encodes lamin. (E, F) BIFC assay was performed in Arabidopsis protoplasts. YFP^N and YFP^C, N-terminal and C-terminal yellow fluorescent protein, respectively. (G) Subcellular localization of the truncated protein, namely gaiΔNterminal:VENUS, used as negative control in the BIFC assay. AtPARP3:mCHERRY was used as nuclear control. Scale bars, 20 µm.