CRISPR/Cas9-mediated mutation of GhCAD decreases the gossypol content of cottonseed

Abstract

Cottonseed is the most important byproduct of cotton production. However, high free gossypol contents limit the application of cottonseed in the food or feed industry. In this study, CRISPR/Cas9 technology was used to knock out the (+)-δ-cadinene synthase gene (GhCAD) to decrease the gossypol content. Gossypol levels decreased approximately 64% in cottonseeds and leaves following the targeted mutation of GhCAD. If only GhCAD1-A was edited, the seed gossypol content decreased by approximately 46%, but there were no major changes in the leaf gossypol content. In addition, the protein and fatty acid (C16:0, C18:1, and C18:2) profiles of the transgenic cotton seeds were similar to those of the control cotton seeds. Furthermore, transcriptome analysis revealed that the jasmonic acid signal transduction pathway was significantly enriched among the DEGs, and GhMYC2-D09 expression was down-regulated. Silencing of GhMYC2-D09 via virus-induced gene silencing decreased the expression of gossypol biosynthesis-related genes, ultimately restricting the accumulation of gossypol in cotton leaves. In contrast, the overexpression of GhMYC2-D09 in hairy roots had the opposite effect. Dual-luciferase assays revealed that GhMYC2-D09 can activate the expression of GhCAD1-A and GhCAD1-C, but Y1H assays revealed that GhMYC2-D09 cannot bind directly to GhCAD promoters. In conclusion, we used CRISPR/Cas9 technology to silence GhCAD expression and developed new genetic resources for generating low-gossypol cotton materials. Furthermore, we characterized GhMYC2-D09 as a transcription factor that increases gossypol biosynthesis. These findings may provide new insights to further elucidate the regulatory network of gossypol biosynthesis.

Keywords: GhCAD; CRISPR/cas9; Cotton; Gossypol.

Fig. 1

Construction of vectors for transformation and mutation analysis of T₀ plants. (A) GhCAD1-A and GhCAD1-C expression profiles in the TM-1 calycle, leaf, petal, stem, torus, ovule, root, and seed determined via transcriptome sequencing. (B) GhCAD1-A and GhCAD1-C gene structure and target sites as well as the T-DNA region of the CRISPR/Cas9 vector used for targeted mutations in cotton. (C) Editing efficiency of nine sgRNAs in hairy roots using the CRISPR/Cas9 system. Twenty hairy roots were used for each target site. (D) Mutation analysis of T₀ gene-edited plants

Fig. 2

Gossypol contents and agronomic traits of gene-edited cotton plants and wild-type (WT) R15. (A) Comparison of seed gossypol contents of WT and T₁ mutants (n = 4, P < 0.05, Tukey’s HSD test). (B) Comparison of leaf gossypol contents of WT and T₁ mutants (n = 4, P < 0.05, Tukey’s HSD test). (C–F) Gossypol levels in cotton lines determined by LC‒MS. (C, D) Seeds of WT and T₂ mutants (n = 4, P < 0.05, Tukey’s HSD test). Black arrows indicate gossypol peaks for WT and mutant plants. (E, F) WT and T₂ mutant leaves. (G) Protein and fatty acid contents in WT and gene-edited plants. (H) Germination of seeds from WT and gene-edited plants. (I) Plant height and fiber quality (upper half mean length, breaking tenacity, uniformity index, and micronaire) of WT and gene-edited plants. (J) PCR analysis of Cas9 in T₂ cotton plants. Cas9 was detected in 22 sgAC2-1 progeny plants. Cas9-free plants are indicated in red. Negative control: R15; positive control: plasmid containing Cas9

Fig. 3

Comparative analysis of sgAC2-1, sgC1-2, and R15 leaves. (A, B) Venn diagrams showing shared and unique (A) down-regulated and (B) up-regulated DEGs in CAD1-C (sgC1-2) and CAD1-AC (sgAC2-1). (C, D) KEGG enrichment analysis of (C) down-regulated and (D) up-regulated DEGs. (E) qRT‒PCR analysis of expression levels of 12 genes selected according to RNA-seq data. Values are presented as the mean ± SD of three independent biological replicates. GhUBQ7 was used as a internal control. ^*p < 0.05, ^**p < 0.01, Student’s t-test. (F) ABA and JA signal transduction pathway and a heatmap of the transcript levels of selected genes (according to transcriptome data)

Fig. 4

Effect of silencing GhMYC2 on gossypol content. (A) Conserved domain analysis of MYC2 proteins from nine species. Six GhMYC2 genes were indicated by a red frame. (B) Gossypol contents after three GhMYC2 gene groups were silenced by VIGS. (C) Expression level of GhMYC2-A09 and GhMYC2-D09 after silencing. (D) Gossypol contents after GhMYC2-A09 and GhMYC2-D09 were silenced. (E) Expression level of gossypol biosynthesis-related genes in GhMYC2-D09-silenced plants. Gene accession numbers: CAD (Gh_D05G3506), CYP706B1 (Gh_D03G176100), CYP82D113 (Gh_D05G199000), CYP71BE79 (Gh_A13G153200), DH1 (Gh_A01G228800), 2-ODD-1 (Gh_D13G235700). Values are presented as the mean ± SD of three independent biological replicates. Statistical significance: *P ˂ 0.05 (Student’s t test)

Fig. 5

Validation of GhMYC2-D09 as a transcription factor. (A) Subcellular localization of GhMYC2-D09 in Arabidopsis protoplasts. (B) GhMYC2-D09 expression profiles in different tissues (root, leaf, stem, petal, bract, sepal, torus, and ovule at different stages). (C) CAD (Gh_D05G3506) and GhMYC2-D09 expression at specific time points (0, 1, 3, 6, and 12 h) after a 100 µM MeJA treatment. Values are presented as the mean ± SD of three independent biological replicates

Fig. 6

Gossypol contents of OE-GhMYC2-D09 plants. (A) Overexpression of GhMYC2-D09 in hairy roots, with the empty pCAMBIA1302 vector serving as a control (CK). (B) Expression of GhMYC2-D09 in three OE-GhMYC2-D09 hairy root lines. (C) Expression of gossypol biosynthesis-related genes in OE-GhMYC2-D09 hairy roots. Gene accession numbers: CAD (Gh_D05G3506), CYP706B1 (Gh_D03G176100), CYP82D113 (Gh_D05G199000), CYP71BE79 (Gh_A13G153200), DH1 (Gh_A01G228800), 2-ODD-1 (Gh_D13G235700). (D) Gossypol content in OE-GhMYC2-D09 hairy roots. Bar: 1 cm. Values are presented as the mean ± SD of three independent biological replicates. Statistical significance: *P ˂ 0.05 (Student’s t test)

Fig. 7

Regulatory effect of GhMYC2-D09 on GhCAD expression. (A) Transcriptional activation assay in yeast. SDO (SD/−Trp medium); TDO (SD/−His/−Trp/−Ade medium). (B) Schematic representation of the reporter and effector. GhCAD1-A and GhCAD1-C promoter fragments (2,000 bp) were fused with the LUC gene. The effector plasmid contained the full-length GhMYC2-D09 sequence under the control of the CaMV 35 S promoter. (C) Effect of GhMYC2-D09 on the GhCAD promoter. LUC/REN, Luciferase/Renilla ratio. Values are presented as the mean ± SD of three independent biological replicates. Statistical significance: *P ˂ 0.05 (Student’s t test). (C) Effect of GhMYC2-D09 on GhCAD1-A and GhCAD1-C promoters in N. benthamiana leaves (E) Binding of GhMYC2-D09 to GhCAD1-A and GhCAD1-C promoters assessed in a Y1H assay