When designing live bacterial therapeutics, preventing the escape of engineered bacteria into the environment is a crucial element in the biocontainment strategy. Regular immigration of organisms can wreak havoc on local ecosystems and this damage may only be exacerbated by releasing genetically modified organisms. This is particularly problematic in the human body and amongst human-associated microorganisms where horizontal gene-transfer is even more widespread due to increased cell-cell contact such as in biofilms . The transfer of genes such as those conveying antibiotic resistance, virulence factors, toxins can render our chassis potentially dangerous for use in living therapeutics.
Previous iGEM teams have integrated kill switches within their design, some of which we took inspiration from to design our kill switch. Team TU_Darmstadt 2020 created a cre-recombinase and quorum-sensing-based kill switch that triggers cell death if cell density levels fall under a certain threshold. Wageningen 2019 and Munich 2012 have incorporated the mazEF toxin/antitoxin system to trigger cell death, while the NCKU Tainan 2015 team knocked out the dapA gene to create an auxotrophic system, meaning that the bacteria were engineered to become incapable of synthesizing one or more compounds required for survival . In such systems, cell death is triggered by the absence of the auxotrophic or essential genes.
As we are developing a living therapeutics solution that involves introducing engineered bacteria into the human urinary tract, auxotrophic systems are not feasible for our project. We have opted for a toxin/antitoxin system kill switch that can stabilize plasmids and promote chassis survivability in the urinary tract without posing any fitness defects. The toxin/antitoxin MazEF system reduces the action of the toxin protein MazF by counter-acting an antitoxin protein MazE. Importantly, MazF is a species-specific endoribonuclease and should cause no damage to human cells. We have used this system to enable growth under controlled conditions. In our design, our double status kill switch design is OFF and the lethal phenotype of our chassis should be inactive in urinary environments to allow our chassis to survive when bound to our hydrogel matrix.
Since we needed our kill switch to be inactivated when conditions are favourable, we needed to consider input sources that were specific to the urinary tract. Both urea and creatinine compounds are present in the urinary tract under specific threshold ranges that we wanted to harness to create an effective kill switch. In healthy adults, urea concentration range from 100-10,000 mmol/L while creatinine is present between 10-100 μM in healthy adults [6,8, 23, 24].
Table 1.- Normal concentrations of urea and creatinine in urine of Adult (>18 years old). Data from the Human Metabolome Database [6-8]
|Urea in Urine||Adult (>18 years old)||12285 μmol/mmol creatinine to 55008.542 +/- 103035.182 μmol/mmol creatinine||Normal|
|Creatinine in Urine||Adult (>18 years old)||2480-22900 μM||Normal|
We have obtained inspiration from the NYMY-Taipei 2008 Project where the team created a urea sensor based mechanism to remove urea from the urine. The biosensor comprises the PureD promoter regulated by the UreR transcriptional regulator of the urease operon of Proteus mirabilis as shown in Figure 1. While looking for parts we could use for our biosensor we noticed that the sequence of PureD (BBa_K116201) created by the team had no sequence specified. Consequently, we have re-included the PureD sequence in the registry as a new basic part BBa_K3710003 and created the codon-optimised UreR transcriptional regulator (BBa_K3710002) to create a urea biosensor as a new composite part (BBa_K3710005). The mechanism of the biosensor is displayed in Figure 1, where the urea-specific transcriptional regulator UreR senses the presence of urea and binds to the PureD promoter region to enhance expression. The intergenic region PureD is a bidirectional promoter located between the UreR and UreD urease subunits which is regulated by the UreR transcriptional regulator. For our biosensor design, we have incorporated a Red fluorescent protein (mRFP1) that would signal the presence of urea in a system as shown in Figure 1
Figure 1.- Urea sensor mechanism. Urea is displayed in Yellow.
Originally, we wanted to create a creatinine biosensor as the metabolite is widely present in the urine and the ratio between urea and creatinine thresholds could be used to create a urine-specific kill switch. After we found an example of a creatinine-inducible system in the production of creatinine degrading enzymes by Pseudomonas putida C-83. However, the genome sequence of Pseudomonas putida C-83 had not been deposited to the NCBI database as of the time of writing, making it difficult to look into the genomic context of creatinine degradation [22, 25]. Nevertheless, the P. putida strains listed on the KEGG pathway that contain creatininase (EC 220.127.116.11) all share a very similar genomic organisation. Creatininase is clustered with a putative GABA permease and an AraC-type transcriptional regulator that might be involved in the activation of gene expression in response to creatinine. However, the regulator translational start site (TSS) seems to vary between the individual species, making it unclear where the actual TSS is [21, 22].
Consequently, we decided to look into alternatives and finally settled on a sarcosine biosensor. Sarcosine (N-methylglycine) is an intermediate metabolite in the intermediate in the metabolism of choline, carnitine, creatine, and glyphosate and is encoded by the sarcosine oxidation operon sox in bacteria . In the urinary tract, creatinine is converted into creatine by creatininase and creatine is subsequently converted into sarcosine and urea by creatinase [21, 22]. Sarcosine oxidase, the enzyme that subsequently breaks down sarcosine is produced in response to sarcosine. For our kill switch design we have created a sarcosine sensor (Figure 2) using the SouR sarcosine transcriptional regulator (BBa_K3710000) and the promoter region pglyA1, which we have named pSouR (BBa_K3710001) to create a sarcosine sensor composite part. Similarly to out urea sensor, we have incorporated a Red fluorescent protein that would signal the presence of sarcosine in a system as shown in figure 2.
Figure 2.- Sarcosine sensor mechanism. Sarcosine is displayed in Green.
As sarcosine is an intermediate in the creatinine metabolism, we found that the threshold range value should be in the micro molar range as shown in table 2. Consequently, with the help from our advisors, we have tested the detection range of our sarcosine sensor design as part of our wet lab work. Our wet lab indicated that the detection range of sarcosine by our biosensor is much higher (in the millimolar range, above 12.5 mM) compared to the range present in the urinary tract (micromolar) (Table 2).
Table 2.- Normal concentrations of sarcosine in urine of Adult (>18 years old). Data from the Human Metabolome Database [6-8]
|Sarcosine in Urine||Adult (>18 years old)||0.00100 - 2.9 (0.5-5.4) μmol/mmol creatinine||Normal|
Our final kill switch design is organised as an AND gate where both urea and sarcosine have to be present in the system in order for our chassis to survive in urinary environments. Our biocontainment strategy utilises the threshold range values normally present in urine to prevent our chassis from surviving outside optimal environemnts. The kill switch functions by actively secreting the MazF toxin along with UreR and SouR via the Pcons constitutive promoter to produce a toxic environment and kill the chassis when either urea, sarcosine, or both are absent (Figure 3). When both sarcosine and urea are present, the MazF toxin is still produced, but activation of the biosensors induce the expression of the antitoxin MazE to neutralise the toxin and prevent the induction of the kill switch. When the UreR and SouR transcriptional regulators both sense urea and sarcosine, these factors then bind to the promoters pureD and PsouR, which transcribe the T7 N- and C-Terminal proteins which form an AND gate. These proteins form an intact T7 polymerase to allow for expression under the PT7 promoter which transcribes for the antitoxin MazE downstream, countering the effects of the MazF toxin as shown in Figure 3.
Figure 3.- Final Kill Switch Design mechanism in an AND gate organisation controlled by Urea and Sarcosine.
Previous iGEM teams including the Thessaly 2020 team working with toxin/antitoxin systems have reported issues with basal levels of expression of the toxin. To overcome this limitation, the team described placing the antitoxin gene under a constitutive promoter and overexpressing the toxin gene when conditions are unfavorable. We considered this design for our own kill switch where the antitoxin MazE would be expressed constitutively along with the toxin gene MazF (figure 4). A repressor would then repress the activity of the toxin when conditions are favourable or when both urea and sarcosine are expressed. However for this design to be functional we would require a strong repressor for the toxin gene and we were concerned that the practical balance between the promoter strength of the mazE and the mazF genes may be difficult to control. For instance, if the constitutive promoter is too strong, there would be a lot of MazE being produced, which will negate small amounts of MazF that is later produced upon signal induction. In that scenario, the system would need a lot of MazF to be produced to counter MazE, which may render our kill switch as ineffective or not working as fast as initially expected. Additionally, it might be the case that MazE is very stable in our system, which might lead it to accumulate massively and also result in an ineffective kill switch. In the future we would like to test this in the lab to further characterise how the sarcosine and urea sensors interact with the MazEF toxin/antitoxin system in the urinary tract.
Figure 4.- Alternate Kill Switch Design mechanism in an AND gate organisation controlled by Urea and Sarcosine. Both antitoxin and toxin are constitutively expressed
Part of our project involved seeking out new urease inhibitors to tackle the ureolytic bacteria which frequently colonise the urinary tract to cause urinary tract infections. Due to the toxicity concerns regarding FDA-approved AHA acetohydroxamic acid as a urease inhibitor, we focused on plant-derived compounds as potential candidates as urease inhibitors. We decided to only perform molecular docking on compounds we have evaluated for biosafety and that showed promising potency (IC50) values. Acetohydroxamic acid has an IC50 value of around 27μM towards the Heliobacter pylori urease . Based on this value we selected a number of flavonoids, terpenoids, and alkaloids and evaluated their IC50 values from literature against ureases. Based on reported safety considerations, we further selected the safest compounds for therapeutic use. This selection was based on toxicity and safety studies we found in literature.
Table 1.- Potency and safety evaluation of potential compounds for therapeutic use.
|Bergenin||Phenol acid||21.7μm(8)||B. pasteurii||Safety strongly depends on dosage.Tested on animals such as mice for malaria inhibition and rats for its neuroprotective effects. ( Liang et al., 2013; Barai et al., 2019 )
Hepatoprotective effects studied in rat and mice models with liver damage. ( Lim et al., 2001 )
|Caffeic acid phenyl ester||Phenol acid||20.96μm||JBU||Experimentation in vitro and in vivo mainly on tumour suppression. ( Espindola et al., 2019 )|
|Epiberberine||Alkaloid||3.0μm / 2.3μm (9)||H. pylori / JBU||Tested in rats and mice, identified as a safe compound for therapeutic use ( Chen et al., 2017; Luo, Yan and Yang, 2014 ). At doses below 200 mg/kg/day, the rats studied showed no signs of mortality or morbidity ( Yi et al., 2013 ).|
|Epigallocatechin||Catechins||2.2μm (10)||H. pylori||Hepatoxic (damage to liver cells) in animal models, however there are two main factors to this: dosage and sensitivity to the extract ( Church et al., 2014; Mezera et al., 2015 ). Oral administration tested in rats. No adverse effects at 500 mg/kg/day in rats and dogs ( Isbrucker et al, 2006 ).|
|Genistein||Isoflavone||1.6μm(11)||H. pylori||In vivo mouse and rat models suggest genistein does not act as a clastogen with doses of 20 mg/kg and 2000 mg/kg per day ( McClain, Wolz, Davidowich and Bausch, 2006 ). A rat model has shown that doses of 5, 100 or 500 ppm imposed no adverse or carcinogenic effects on male Sprague-Dawley rats, but some carcinogenic effects were exhibited in female rats such as the development of mammary gland adenoma ( National Toxicology Programme, 2008).|
|Quercetin||Flavonoid||11.2μm(12) / 80μm (13)||H. pylori / JBU||“Genotoxic in salmonella, but it’s safety upon human application is approved" (Okamoto, 2008 ). A study on cytotoxicity towards specific human cells showed that quercetin demonstrated higher cytotoxicity towards the human umbilical vein endothelial ( Matsuo, Sasaki, Saga and Kaneko, 2005 ).|
|7,8,4’-Trihydroxy-2-isoflavene / 8-hydrozydaidzin||Isoflavone||0.85μm (14)||H. pylori||H301 category - toxic if swallowed, inhaled or in contact with the skin ( National Center for Biotechnology Information, u.d ).|
Literature suggested that quercetin was the safest option to consider as a potential urease inhibitor due to the wider availability of data on this compound. Previous studies have suggested that the mobile helix−turn−helix motif called the flap motif of ureases that contains key residues αLys220, αCys322, and αHis323 may interact with hydroxyl functional groups of compounds that trigger a closure of the active site of the urease enzyme . Binding to these residues covalently modifies these conserved residues of the flap motif inducing a closed conformation. We wanted to test whether Quercetin and derivatives would show promising binding affinities with the urease enzyme and we decided to perform molecular docking simulations using AutoDock Vina. Further information on this work can be found on our molecular docking page.