Biosafety is an important consideration when designing engineered bacteria. Our engineered bacteria is expected to work in the soil, so we need to consider whether the product synthesis can be easily controlled and whether there are any potential risks to soil structure, crop growth, the balance of soil microbiota and human health. So a "suicide system" is designed to ensure the biosafety of our engineered bacteria the environment and human.

Bacteria suicide is a common phenomenon in nature, which is a programmed cell death of prokaryotes. A quorum-sensing (QS) based suicide gene circuit has been studied. The quorum sensing system allows cell-cell communication by some auto-inducing signal molecules like AHL. The cell behavior will be synchronized with the concentration of AHL and thus precise regulation of cell population will be achieved. And the result in the QS-based suicide circuit showed about 10% of cells will retained after one round of suicide, and these cells starts another round of growth. The cell behavior and cell density will repeat after that.

In addition to QS-based suicide system, suicide systems with other regulatory parts can also be designed by synthetic biology. Here, we designed a temperature-responsive suicide system to achieve temperature-controlled lysis process by a novel lysis genes (Gene ID: IF654_RS00240) (Figure 1).

The lysis gene is from Enterobacteria phage KleenX174. To better express the lysis gene in engineered bacteria, the codon optimization of the lysis gene was conducted according to the codon preference of Escherichia coli. Figure 3 shows the number of codons we optimized to make our codons more in line with Escherichia coli preference. The modified lysis gene is shown in BBa_K4182007.

Initially, we planned to construct our suicide circuit (Plasmid 5) using vector backbone pSB1K3 by Golden Gate assembly, and we can obtain several clones. However, it was found that after colony PCR verification, only weak target bands could be observed (Figure 4), and the plasmids extracted from the recombinant DH5α cells was at very low concentration, and sequencing could not be completed.

We supposed that due to the low copy number of pSB1K3 vector, it is difficult to extract sufficient plasmid from the engineered bacteria for further validation. Therefore, we replaced the backbone to pSEVA341, a higher-copy-number vector and re-constructed the plasmid (Figure 5). As shown in Figure 6, the obvious target bands were observed, and the plasmid correctness was further confirmed by sequencing.

The engineered cell harboring plasmid 5 and blank vector respectively, were culture at 30℃ overnight, and then the temperature was shift to 42℃. The OD600 of each group was detected every 1 h, and the growth curve of these strains were determined as follows.

The results clearly demonstrated the cell growth was significantly inhibited after heat at 42℃ compared to the strain without lysis gene. And about 9% of whole cell population was retained after heat. Compared to the commonly used suicide protein MazF (BBa_K302033), the lysis protein in our study is also efficient but shorter and easy to be manipulated, which can be used as an alternative and update for MazF.

AA, aspartic acid, is a novel natural herbicide that can be synthesized by fungi (Yan Y et al, 2018). In the situation of increasing tolerance to existing herbicide of glufosinate (APHTHINE), AA offers another environmentally effective and low-tolerance option with significant results (See the results of Yan Y et al). AA targets dihydroxylation dehydrase (DHAD) in the synthesis pathway branched-chain amino acid and leads to the growth inhibition of plants. Branched-chain amino acids (BCAAs), including leucine, isoleucine, and valine, are essential nutrients for plant growth, and the key point of their biosynthetic pathways are dihydroxydehydrase (DHAD) which catalyzes αβ-dihydroxylation dehydration reaction to form the precursor α-ketoacid. DHAD is highly conserved in different plant species and DHAD with its BCAA biosynthetic pathway does not exist in mammals, making it an ideal target for herbicides. The biosynthetic pathway of AA is shown as follows. The precursor pGPP is synthetized via MVA pathway from glucose, which will be catalyzed by FPPS to generate FPP, and eventually to AA by astABC gene cluster.

fpps and astABC (from the soil fungus Aspergillus terreus) were codon-optimized based on E. coli and chemically synthesized. And the synthetized astAB and astC are cloned into two separate plasmids as shown in Figure 3. In order to avoid the metabolic stress caused by high-copy plasmids, the AA synthesis circuit (Plasmid 3) was constructed based on the medium-copy number backbone pBBRMCS1. It contains the astABC gene cluster regulated by the lac promoter and the specific transcription terminator of E.coli rrnB gene, as well as several high-efficient RBS (RBS1-3) (Figure 4). The astABC gene cluster, LacI-Plac regulatory sequence, and linear pMCS1 plasmid backbone were obtained by PCR respectively, and final plasmid 3 was constructed one-step Golden Gate assembly. The plasmid 3 was confirmed by colony PCR verification and gene sequencing (Figure 5).

Due to the long cycle of plant experiments and the limited time, we did not conduct plant experiments. However according to the paper "Resistance gene-directed discovery of a natural-product herbicide with a new mode of action" (Yan Yan et al, 2018), the activity of AA was extensively studied and showed that 250 μM AA exhibit an efficient activity to kill plants. And transgenic plants containing AA-inhibitor protein DHAD has an obvious resistance to AA, indicating the potential of our herbicide to kill weeds. Our primary study on the novel herbicide will promote its further research and applications in the future.

Therefore, we successfully constructed and provided a novel termperature-controlled suicide circuit with lysis gene, and a novel herbicide synthesis circuit. The parts are as follows: