The Exported Chorismate Mutase from Bacterial Pathogens of Mammals: From Residues Important for Protein Export to the Quest for the Biological Function

Abstract

Protein export in bacteria is the translocation of proteins out of the cytoplasm. The majority of proteins destined for export are equipped with an N-terminal leader peptide to be recognized by the transport machinery. Protein export can be exploited for the industrial production of biomolecules with bacteria as the production strains, simplifying downstream purification processes. Moreover, many exported proteins act as virulence factors to help pathogenic bacteria to invade and persist in the host. Therefore, deciphering factors important for the bacterial protein export mechanism is crucial both for the optimization of biotechnological production as well as the definition of new targets for antibiotic treatment. Here, we used the exported chorismate mutase (*CM) from a pathogenic bacterium as a test case to study protein export. Chorismate mutase (CM) is an enzyme from the intracellular shikimate pathway converting chorismate to prephenate and leading to the synthesis of the aromatic amino acids L-tyrosine and L-phenylalanine. The shikimate pathway is ubiquitous in bacteria, fungi, and plants, but absent in mammals. CMs of the AroQ class are α-helical and can be divided into four subclasses from AroQγ to AroQγ. The *CMs studied in this work belong to the AroQγ subclass enzymes, which possess an N-terminal leader peptide. Their role in the extracytoplasmic space is not known. However, they occur in many pathogenic bacteria, like Mycobacterium tuberculosis, Pseudomonas aeruginosa, and Salmonella enterica serovar Typhimurium, but are absent, for example, in Escherichia coli. In contrast, AroQα, AroQβ, and AroQδ are intracellular and essential housekeeping CMs. Directed evolution approaches targeting intracellular CMs in bacteria were previously established. On media without Tyr and Phe only cells possessing a functional intracellular CM are able to grow. We applied this directed evolution system to study protein export in bacteria by using the AroQγ subclass CM of S. Typhimurium (*StCM) as a model. The enzyme without the leader peptide can complement an intracellular CM deficiency of a Tyr and Phe auxotrophic E. coli strain whereas the wild-type *StCM containing its leader peptide is not able to do so. Therefore, we could actively select for export-defective variants upon subjecting the full-length protein to directed evolution. We directly identified sequence patterns in the hydrophobic core region of the heterologous bacterial leader peptide that are detrimental for efficient export in E. coli. The discovered amino acid changes, which drastically hampered the export of the enzyme from the cytoplasm, where substitutions from hydrophobic into charged residues, with high occurrence of Leu to His (L9H) and Met to Lys (M14K) exchanges. These results are in agreement with earlier findings that the hydrophobic core of a leader peptide is the most crucial region required for successful export. They also prove the suitability of our *CM model system to study protein export using directed evolution. Additionally, we were interested to find residues in the mature portion of the protein that affect efficient export despite the presence of a functional leader peptide. So far, we discovered residues which, upon mutation, strongly hamper protein export. Possible functions of the discovered residues in inter- and intramolecular interactions important for export are discussed. Additional experiments are required to confirm these findings and explain the underlying mechanisms, including testing a broader range of bacterial *CMs to allow for generalized conclusions about efficient protein export. Since the shikimate pathway takes place intracellularly, the role of a CM outside of the cytoplasm is puzzling. We confirmed that *CMs are very active enzymes. However, outside of the cytoplasm there seems to be no source of chorismate nor use for prephenate. For several phytopathogenic organisms it has been shown that the exported chorismate mutase is involved in their virulence and important for host invasion. We have previously reported that also in pathogenic bacteria of mammals the AroQγ enzyme might act as a virulence factor, even though their hosts do not produce or metabolize chorismate. Here, we addressed the elucidation of the natural function of *CMs and investigated their influence on the phenotype of pathogenic bacteria of animals. For this we focused on two pathogenic bacteria, S. Typhimurium and P. aeruginosa. The pathogen S. Typhimurium inhabits the gut of its host, which can lead to inflammation of the intestine resulting in severe diarrheal disease. We explored whether *StCM is required under nutrient-limiting conditions or whether it could be involved in resistance to growth-inhibiting agents using a previously established set of aroQγ S. Typhimurium strains, including an aroQγ knock-out, a complemented knock-out, and the wild-type strain. Together with the company Biolog Inc. we performed a high-throughput approach to test nutrient usage and sensitivity to chemicals by measuring metabolic activity under different growth conditions. Differences in the metabolic activity of the three aroQγ S. Typhimurium strains were assessed through pairwise comparison. Under nutrient-limiting conditions, no significant difference in metabolic activity was observed comparing the three strains. However, some antimicrobial conditions did reveal differences, prompting further investigations to clearly establish whether *CM confers a significant advantage for the pathogen. Additionally, we sought to reveal the nature of the signals that induce aroQγ gene expression. We could show that deployment of *CMs occurs in response to environmental signals that indicate host interactions. An S. Typhimurium reporter strain with a transcriptional fusion of lacZ to the aroQγ gene has been constructed previously and it has been shown that a low Mg2+ concentration led to 30-fold increase of aroQγ promoter activity. Since low Mg2+ concentration mimics an environment in a host cell, we were looking for other host-specific factors sensed by S. Typhimurium during the infection process. During host colonization, S. Typhimurium has to overcome the harsh conditions of the intestinal tract. We found that the addition of bile salts to the growth medium enhances aroQγ promoter activity about 3-fold, while an increase of 100-fold was observed if combined with low Mg2+ concentration. However, we could demonstrate in a growth experiment that the enzyme does not seem to be involved in conferring resistance of the organism to bile salts. It is therefore still unclear how the enzyme is needed in the intestine of a host. Potential involvements in signaling cascades are discussed. Bacteria possess tightly regulated signaling mechanisms recognizing environmental cues leading to the expression of required virulence factors to successfully adapt to and persevere in a host environment. This is typically not established by a single bacterium alone but rather by a group, which communicates via chemical signals to establish group behaviors (e.g. biofilm formation), ultimately resulting in a response, such as virulence. We hypothesized that the *CM might be involved in such interactions either with the environment or with other bacteria from the community. Therefore, another model organism, P. aeruginosa, was considered. This opportunistic pathogen has the ability to invade a broad range of habitats including organs of mammals. When colonizing the lungs of cystic fibrosis patients deadly infections can result. This species has already been thoroughly used to study bacterial group behaviors, including biofilm formation and swarming. Two P. aeruginosa mutants were acquired from a transposon mutant library containing insertions in the aroQγ gene. Biofilm and swarming behaviors were tested, comparing the mutants to the wild-type parental strain. Under biofilm-promoting conditions, the wild type showed increased biofilm formation whereas the aroQγ mutants did not increase their biofilm mass. Furthermore, on a semi-solid surface the aroQγ mutants exhibited a strong swarming phenotype compared to the more weakly swarming wild type. This is compatible with the notion that biofilm formation and swarming ability are oppositely regulated. In fact, the differential phenotypes of the mutants and the wild type suggest an involvement of aroQγ in the regulatory mechanism controlling swarming and biofilm production. Future experiments will involve genetic complementation of the phenotypes to establish this hypothesis further. In conclusion, amino acid residues important for protein export were discovered in this work employing directed evolution. Besides exploiting the *CM to investigate protein export, the natural function of this group of exported enzyme was studied. Using the pathogenic bacteria S. Typhimurium and P. aeruginosa as model organisms and comparing wild type and aroQγ mutants in a variety of experiments, our results suggest that *CMs are important under specific environmental conditions and that they might trigger particular phenotypes upon recognition of environmental cues. Further experiments are proposed to elucidate the biological role of *CMs in pathogens of mammals.