Proteomic Profiling, Transcription Factor Modeling, and Genomics of Evolved Tolerant Strains Elucidate Mechanisms of Vanillin Toxicity in Escherichia coli

Abstract

A particular problem for the biotechnological production of many of the valuable chemicals that we are now able to manufacture in bacterial cells is that these products often poison the cells producing them. Solutions to improve product yields or alleviate such toxicity using the techniques of modern molecular biology first require a detailed understanding of the mechanisms of product toxicity. Here we have studied the economically important flavor compound vanillin, an aromatic aldehyde that exerts significant toxic effects on bacterial cells. We used high-resolution protein abundance analysis as a starting point to determine which proteins are upregulated and which are downregulated by growth with vanillin, followed by gene expression and mutant studies to understand the mechanism of the response. In a second approach, we evolved bacterial strains with higher vanillin tolerance. Their genome sequences have yielded novel insights into vanillin tolerance that are complementary to the proteomics data set. Vanillin (4-hydroxy-3-methoxybenzaldehyde) is an economically important flavor compound that can be made in bacterial cell factories, but toxicity is a major problem for cells producing this aromatic aldehyde. Using (i) a global proteomic analysis supported by multiple physiological experiments, mutant analyses, and inferred transcription factor modeling and (ii) adaptive laboratory evolution (ALE) of vanillin tolerance combined with genome-wide analysis of the underlying mutations, mechanisms of vanillin toxicity in Escherichia coli have been elucidated. We identified 147 proteins that exhibited a significant change in abundance in response to vanillin, giving the first detailed insight into the cellular response to this aldehyde. Vanillin caused accumulation of reactive oxygen species invoking adaptations coordinated by a MarA, OxyR, and SoxS regulatory network and increased RpoS/DksA-dependent gene expression. Differential fumarase C upregulation was found to prevent oxidative damage to FumA and FumB during growth with vanillin. Surprisingly, vanillin-dependent reduction pf copper (II) to copper (I) led to upregulation of the copA gene and growth in the presence of vanillin was shown to be hypersensitive to inhibition by copper ions. AcrD and AaeAB were identified as potential vanillin efflux systems. Vanillin-tolerant strains isolated by ALE had distinct nonsynonymous single nucleotide polymorphisms (SNPs) in gltA that led to increased citrate synthase activity. Strain-specific mutations in cpdA , rob , and marC were also present. One strain had a large (∼10-kb) deletion that included the marRAB region. Our data provide new understanding of bacterial vanillin toxicity and identify novel gene targets for future engineering of vanillin-tolerant strains of E. coli . IMPORTANCE A particular problem for the biotechnological production of many of the valuable chemicals that we are now able to manufacture in bacterial cells is that these products often poison the cells producing them. Solutions to improve product yields or alleviate such toxicity using the techniques of modern molecular biology first require a detailed understanding of the mechanisms of product toxicity. Here we have studied the economically important flavor compound vanillin, an aromatic aldehyde that exerts significant toxic effects on bacterial cells. We used high-resolution protein abundance analysis as a starting point to determine which proteins are upregulated and which are downregulated by growth with vanillin, followed by gene expression and mutant studies to understand the mechanism of the response. In a second approach, we evolved bacterial strains with higher vanillin tolerance. Their genome sequences have yielded novel insights into vanillin tolerance that are complementary to the proteomics data set.