5-Hydroxytryptophan (5-HTP) is a direct biosynthetic precursor of serotonin, a critical neurotransmitter in mood regulation. Our project aims to revolutionize serotonin supplementation by genetically modifying the probiotic Escherichia coli Nissle 1917 (EcN) to produce 5-HTP in the intestines. This approach enhances and regulates 5-HTP production, incorporating protective mechanisms to ensure the probiotics successfully pass through the stomach's acidic environment. The resulting "SERENE" probiotics could offer a reliable and accessible option for improving mental well-being.
▲ Figure 1: The human Tryptophan - 5-HTP - Serotonin pathway.
Escherichia coli Nissle 1917 (EcN) has been widely used as a probiotic since its discovery in the second half of the 19th century by Alfred Nissle1. It is one of the most extensively studied non-pathogenic E. coli strains. It is an active ingredient in approved medications, making it a reliable probiotic option and a robust chassis for our project.
In the serotonin biosynthesis pathway, tryptophan hydroxylase 1 (TPH1) catalyzes the formation of 5-HTP from L-tryptophan. Monomeric-human TPH1 (m-hTPH1) is a truncated form of TPH genes (TPH1 mutant with a deletion of the first 99 N-terminal and last 24 C-terminal amino acids, NΔ99/CΔ24), has been validated for successful expression in E. coli.
TPH1 requires tetrahydrobiopterin (BH4) as a cofactor, which is oxidized during the reaction. BH4 is regenerated from its oxidized form, BH4-4a-carbinolamine, by pterin-4a-carbinolamine dehydratase (PCBD1) and quinoid dihydropteridine reductase (QDPR)(Wang et al., 2018). In our project, we applied human PCBD1 (hPCBD1) and human QDPR (hQDPR) genes. The hPCBD1 dehydrates the BH3OH to the qBH2, and hQDPR reduces the qBH2 to the BH4, acting as an enzyme in the BH4 regeneration system.
▲ Figure 2: The enzyme m-hTPH1 and the BH4 regeneration system.
We have engineered EcN to express the m-hTPH1, hPCBD1, and hQDPR proteins using a polycistronic expression vector, pST39. The hTPH1, hPCBD1, and hQDPR genes were inserted into plasmid pST39 using Gibson Assembly, resulting in the construction of plasmids pST39-m-hTPH1, pST39-hPCBD1, pST39-hQDPR, and pST39-m-TPH1-hPCBD-hQDPR. These plasmids were initially constructed in E. coli DH5α and expressed in E. coli BL21 for protein production.
To further enhance 5-HTP production, we employed a DNA scaffold strategy to bring the m-hTPH1, hPCBD1, and hQDPR enzymes into close physical proximity. By utilizing zinc finger protein (ZNF) binding domains, we positioned the scaffold using the Rolling Circle Replication (RCR) mechanism to generate circular single-stranded DNA (ssDNA) with complementary sequences, forming the ZNF binding motifs. This approach ensures sufficient scaffold formation in EcN, promoting efficient enzyme clustering.
Zinc finger proteins (ZNFs) are widely expressed and perform diverse functions, including interactions with DNA, RNA, and proteins. The classic zinc finger structure is stabilized by zinc ions, which coordinate with cysteine and histidine residues in various combinations. Importantly, zinc fingers exhibit DNA binding specificity. In our design, we utilize the zinc finger binding domains Zif268 (Bulyk et al., 2001), PBSII (Conrado et al., 2012), and ZFa to locate the DNA scaffold accurately.
To facilitate enzyme assembly on the DNA scaffold, we fused the Zif268, PBSII, and ZFa zinc finger binding domains with the hTPH1, hPCBD1, and hQDPR enzymes, respectively (Giuseppe Raschellà et al., 2017). The binding motifs corresponding to each zinc finger binding domain are strategically placed on the DNA scaffold. As the order of domains and the types of linkers used can significantly impact the folding and function of fusion proteins, we conducted Protein Modeling to predict the optimal structure of each fusion protein. This modeling allowed us to determine the appropriate linkers required for the correct folding and functionality of the three fusion proteins.
▲ Figure 3: Protein structure simulations using AlphaFold2 and iTASSER.
Rolling Circle Replication (RCR) is a critical molecular process in DNA replication and various molecular biology applications. It enables the one-way amplification of circular DNA, resulting in long single-stranded or double-stranded DNA products with repetitive sequences.
During RCR, a replication initiator enzyme, such as RepA, recognizes the target plasmid's double-stranded origin (DSO) and introduces a nick in one of the DNA strands. DNA polymerase then utilizes the intact strand as a template to elongate the nicked strand. Once the elongated strand is fully synthesized, the replication initiator enzyme ligates it into a circular single-stranded DNA (ssDNA). This circular ssDNA can then serve as a template for secondary strand synthesis (Ruiz-Masó et al., 2015).
In our project, we inserted the DNA scaffold template with complementary ZNF binding motifs between RCORI-105 and RCORI-65 sequences for RCR. The circular ssDNA produced through RCR will naturally assemble into a DNA scaffold via these sequence complementarities, enabling the spatial organization of the enzymes involved in 5-HTP production.
Our full circuit involves two plasmids: one encoding the three fusion proteins (Zif268-m-hTPH1, PBSII-hPCBD1, and ZFa-hQDPR) and another controlling the PCA-regulated DNA scaffold formation.
▲ Figure 4: Full circuit of two plasmids, utilizing polycistronic expression vector, pST39.
To protect the probiotics from the acidic conditions of the gastrointestinal tract (e.g., gastric acid), we employed a layer-by-layer (LbL) assembly method using chitosan and sodium alginate on the surface of EcN cells. We enhanced the stability of these multilayers by adding calcium chloride to form a gel layer through ionic cross-linking.