Dynamic transcriptome and GWAS uncover that a hydroxyproline-rich glycoprotein suppresses Agrobacterium-mediated transformation in maize

Genetic transformation is a crucial tool for investigating gene function and advancing molecular breeding in crops, withAgrobacterium tumefaciens-mediated transformation being the primary method for plant genetic modification. However, this approach exhibits significant genotypic dependence in maize. Therefore, to overcome these limitations, we herein combined dynamic transcriptome analysis and genome-wide association study (GWAS) to identify the key genes controllingAgrobacteriuminfection frequency (AIF) in immature maize embryos. Transcriptome analysis ofAgrobacterium-infected embryos uncovered 8483 and 1580 genotype-specific response genes in 18-599R (low AIF) and A188 (high AIF), respectively. A weighted gene co-expression network analysis (WGCNA) further revealed five and seven stage-specific co-expression modules in each corresponding line. Basd on a self-developed AIF quantitation method, the GWAS revealed 30 AIF-associated single nucleotide polymorphisms and 315 candidate genes under multiple environments. Integration of GWAS and WGCNA further identified 12 key genes associated with high AIF in A188, among which,ZmHRGP, encoding a hydroxyproline-rich glycoprotein, was functionally validated as a key factor of AIF in immature embryos. Knockout ofZmHRGPfurther enabled to establish a high-efficiency genetic transformation system for the 18-599R line, with the transformation frequency being approximately 80%. Moreover, transient reduction ofZmHRGPexpression significantly enhanced the AIF of maize calluses and leaves. Overall, these findings advance our understanding of plant factors controllingAgrobacteriuminfection and contribute to the development of more efficientAgrobacterium-mediated transformation systems in crops.