To harness the capabilites of L. lactis for potential use as a probiotic treatment, we have developed a model for L. lactis that has been engineered for the treatment of homocystinuria. We hope to incorporate the RBS sequences obtained from our RBS optimization tool as a part of this model.

Homocystinuria is a rare genetic disorder characterized by the inability of the individual to metabolize methionine. This causes an increased accumulation of homocysteine, an intermediate. This is known to cause thrombotic events like strokes and heart attacks. We have attempted to engineer L. lactis to convert methionine to homocysteine. As it is a GRAS organism, L. lactis is an ideal chassis for developing probiotics.

However, L. lactis does not possess the genes necessary to convert methionine to homocysteine and the gene which converts cystathionine to cysteine. We aim to engineer these genes into L. lactis and use an RBS sequence which can increase the expression of these genes.

Current methionine metabolic capacities of L. lactis:

Conversion of methionine to cysteine involves the following steps:

• Conversion of methionine to homocysteine
• Conversion of homocysteine to L-cystathionine
• Conversion of L-cystanthionine to L-cysteine
• Metabolism of L-cysteine

1. Conversion of methionine to homocysteine

The image below shows the metabolic network of L. lactis involved in the conversion of methionine to homocysteine. The gene which converts an intermediate of methionine (S-adenosyl methionine) to S-adenosyl-L-homocysteine needs to be engineered.

2. Conversion of homocysteine to cystathione
Wild type L. lactis has the ability to convert L-homocysteine to L-cystathionine. However, L. lactis also has the ability to synthesize an enzyme cystathionine-gamma synthase. This enzyme converts L-cysathionine back to L-homocysteine. We aim to knock out this gene.

3. Conversion of L-cystathionine to L-cysteine

The enzyme which converts L-cystathionine to L-cysteine is natively absent in L. lactis and must be incorporated into it. This enzyme however, is produced in yeast naturally, hence we propose engineering this enzyme into L. lactis. The image below shows the absence of enzyme in L. lactis and presence of the same enzyme in S. cerevisiae.

4. Metabolism of L-cysteine

This occurs naturally in L. lactis. However, we plan to use the RBS Optimization tool to increase the expression of the carrier proteins which export L-cysteine from L. lactis to the environment. Based on the genes which need to be inserted and knocked out in L. lactis for the successful conversion of methionine to cysteine, we have designed a genetic circuit.

To portray the application of our web tool, we modelled the transcription and translation of the inserted enzyme, and we aim to show that with the right RBS, we can increase the production of our enzyme of interest to the optimal rate.
We have modeled the transcription and translation as follows:

1. Binding of the RNA polymerase to the promoter to form a promoter-RNAP complex:
   RNAP + promoter → promoter-RNAP
2. The promoter transcription complex dissociates to give back the promoter, RNA polymerase and the mRNA sequence.
3. Binding of the Ribosome to the mRNA at the ribosome binding site to form an mRNA-ribosome complex.
4. The conversion of the mRNA-ribosome complex to form the protein and return back the mRNA and ribosome.

After modeling this on SimBiology, we plotted the production of protein, mRNA with respect to time.

Constants Considered:

• Binding of RNAP to the promoter - [1]
• Dissociation of RNAP-promoter complex to form mRNA - taken approximately
• Binding of the mRNA to the Ribosome to form a mRNA-Ribosome complex - [2]
• mRNA-ribosome complex dissociates to form the protein, mRNA and ribosome. - [2]

Initial Concentrations Considered:

• Concentration of promoter - 2.6x10^-9
• Concentration of RNAP - 1.9x10^-5
• Concentration of ribosome - 4.4x10^-5

We aim to use constraint-based modeling techniques to understand the optimal concentration of CGL enzyme that needs to be produced. We aim to then use the RBS calculator to engineer the necessary RBS, to obtain optimal protein production rate.

[1] Újvári A, Martin C. Thermodynamic and Kinetic Measurements of Promoter Binding by T7 RNA Polymerase. Biochemistry. 1996;35(46):14574-14582.
[2] Kierzek A, Zaim J, Zielenkiewicz P. The Effect of Transcription and Translation Initiation Frequencies on the Stochastic Fluctuations in Prokaryotic Gene Expression. Journal of Biological Chemistry. 2001;276(11):8165-8171.
[3]Xu C, Shi Z, Shao J, Yu C, Xu Z. Metabolic engineering of Lactococcus lactis for high level accumulation of glutathione and S-adenosyl-L-methionine. World J Microbiol Biotechnol. 2019 Nov 14;35(12):185. doi: 10.1007/s11274-019-2759-x. PMID: 31728760.
[4]Alcaide P, Krijt J, Ruiz-Sala P, Ješina P, Ugarte M, Kožich V, Merinero B. Enzymatic diagnosis of homocystinuria by determination of cystathionine-ß-synthase activity in plasma using LC-MS/MS. Clin Chim Acta. 2015;438:261–5
KEGG Pathways: https://www.kegg.jp/pathway/map=lla00270&keyword=Methionine

One of the core concerns in synthetic biology is if Genetically Engineered Organisms (GEOs) pose a danger to their environment. Although precautions are taken, it is impossible to predict the interactions of a novel organism in a system as complex as an ecosystem. We chose to address the need for bio-contaiment with a kill switch, a genetic circuit designed to eliminate the GEOs in a controlled manner when certain conditions are met. The kill switch designed here is activated in the low-glucose environment of the distal part of the human digestive system.

The kill switch works via the well-known mazEF toxin-antitoxin (TA) system activated by the carbohydrate response element binding protein (CHREBP) glucose sensing system. We used a glucose sensing system from the 2019 NUDT China iGEM team. The mazEF toxin-antitoxin system is comprized of the toxin gene MazE and its antitoxin MazF. The antitoxin works by forming a complex with the toxin that deactivates it. mazF inhibits translation thus halting protein production in the cell[https://doi.org/10.1242/jcs.02619]. As the kill switch is supposed to activate in the deficit of glucose, the expression of mazE upwards regulated by the glucose sesnsing system in the abundance of glucose. As is generally the case in TA systems [https://doi.org/10.4161%2Fmge.26219], The antitoxin is degraded quickly by the clPAP protease if it’s production ceases. The expression of the toxin, MazF is by a constitutional promoter selected of such strength as to have the expression of the toxin between the expected high and low level of the antitoxin corresponding to glucose levels.

CHREBP is a transcriptional factor of the glucose sensing system. It is de-phosphorylated from CHREBP P by Protein phosphatase 2 (PP2ase) and then binds to the hybrid promoter controlling the expression of the mazE protein. The expression of PP2A is affected positively glucose, mediated by a circuit consisting of Glucose-6-Phosphate (G6P), 6-Phosphate-Gluconolactone (6PGL), 6-Phosphoglucanoate (6PGLc), Ribulose-5-Phosphate (R5P), Xylulose-5-Phosphate (X5P). The kinetic parameters for the model are from [https://doi.org/10.1007/BF00928908].

The MATLAB add-on SimBiology was used to create the kinetic model, and SimBiology Analyzer was used to simulate it. These applications offer tools for automatic ODEs-based modeling, simulation, and analysis of kinetic systems. The ODEs that correspond to the chemical processes mentioned below can be used to characterize the kinetics of our system.

We have assumed a 1000-fold difference in glucose levels in the stomach and dis- tal digestive tract, following data from [https://doi.org/10.1152/ajpgi.1990.259.5.g822]. As seen in figure 2, the antitoxin mazE is greater in concentration than the toxin mazF in high glucose conditions. This corresponds to the cell continuing to be alive. From figure 4, illustrating the low glucose condition, we can see that the concentration of mazF is greater than mazE, corresponding to cell death.

One of the greatest hurdles of engineered therapeutic mechanisms is integrating an efficient solution into a living organism in all its complexity, with minimal loss of product function. In our approach to resolving homocystinuria, it is to be noted that the approach is not to maximize the production of the required enzymes but to keep it, and other metabolites at optimum levels. Achieving this level of finesse in modulating enzyme activity while minimizing perturbations to other integral systems is a direct application of our prediction and optimization tools. Using an array of RBS sequences characterized by expression levels and the predictor-optimizer tool, we are able to easily add additional features like kill switches, whose core toxin-antitoxin system requires fine control over expression activity.