Identification of ferulic acid tolerance targets in engineered Komagataella phaffii by transcriptomics-integrated genome-scale metabolic modeling

Ferulic acid (FA) is an abundant aromatic precursor in lignocellulosic biomass and a promising substrate for vanillin biosynthesis, but its toxicity limits microbial conversion at elevated concentrations. Komagataella phaffii is a widely used non-conventional yeast cell factory with favorable properties for aromatic compound production, yet the cellular mechanisms underlying FA tolerance in this host remain poorly understood. An engineered K. phaffii strain carrying a heterologous FA-to-vanillin pathway was exposed to 20 mM FA under non-induced (FE–NI) and methanol-induced (FE-I) conditions. Both conditions exhibited higher growth rates and final cell densities compared with the wild-type, with FE-I exhibiting active FA catabolism and transient vanillin accumulation. Notably, FE–NI achieved comparable tolerance without detectable FA consumption, indicating that metabolic reprogramming of the host cell contributes to FA tolerance. Transcriptomic profiling revealed condition-dependent stress responses: FE–NI activated a host-level program comprising cell surface reinforcement, mitochondrial energy and redox support, and central carbon remodeling, whereas FE-I coupled active FA conversion with enhanced aldehyde detoxification and sugar-metabolism remodeling. Candidate tolerance targets were prioritized by integrating transcriptome data with a genome-scale metabolic model using an E-Flux approach. Functional validation by overexpression in the wild-type strain identified AGC1 and PAS_chr3_0834 as the most effective targets under non-induced and methanol-induced conditions, with colony number increasing 7.4-fold and 4.0-fold relative to the control, respectively. Co-expression of AGC1 and PAS_chr3_0834 increased biomass accumulation relative to the wild-type but did not exceed the effect of AGC1 alone under the tested condition. This study reveals condition-dependent FA tolerance programs in engineered K. phaffii and shows that integrating transcriptomics with genome-scale metabolic modeling can facilitate the identification of tolerance targets. The validated targets represent valuable engineering options for improving the robustness of K. phaffii cell factories employed in FA-based aromatic compound bioconversion.