Our Design
1. Acetate production module
Previous researches have revealed a relation between short chain fatty acids (SCFAs) intake and white fat tissue consumption. Mice taking SCFAs on a high-fat diet (HFA) shows a notably slower growth rate of body weight than those on an HFD without intake of SCFAs.[1] It is presumed that by raising the level of their receptor GPR43 in white fat tissue, SCFAs promotes direct consumption of the latter in the body.[2] Among all the SCFAs, acetic acid is most efficiently absorbed and utilized by human body. Therefore, it would be hopeful to help people lose weight in a healthy way by colonizing acetate-producing microbe inside human body.
To enhance acetate production of our engineered microbe, we overexpress two enzymes, phosphotransacetylase (PTA) and acetokinase (ACK), both genes of which are derived from E. coli K-12 genome. Using the very common intermediate acetyl-CoA as substrate, these two enzymes catalyze production of acetic acid through a two-step enzymatic reaction with high efficiency.[3]
Fig 1 Acetate production pathway
2. β-oxidation module
Apart from consuming fat that in human body tissue, it is equally important to prevent assimilation of excess energy. Through market investigation we found out that many currently used drugs aim to achieve that by inhibiting fat digestion. However, accumulation of undigested fat inside digestive track can have multiple negative influences of human health, including disorder of intestinal microbiome. Here we propose a different approach to declining fat absorption – by having our engineered bacteria consume the extra in-taken fat.
The uptaken fat is hydrolyzed into glycerol and higher fatty acids and subsequently absorbed. As a major product of this hydrolyzation, higher fatty acids contain most energy stored in fat. It is therefore practical to promote degradation of higher fatty acids which would otherwise be assimilated by human body. To do this, we overexpress two key enzymes involved in β-oxidation of higher fatty acids, fatty acyl-CoA synthetase (FadD) and fatty acyl-CoA dehydrogenase (FadE).[4] Again, both enzymes are derived from genome of E. coli K-12. Not only does enhancement of β-oxidation consume an extra amount of higher fatty acids, it also gives a rise on the yield of acetate by producing substrate of acetate production – acetyl-CoA.
Fig 2 Fatty Acid β-oxidation pathway
3. Inhibitory sensors
As mentioned before, an inhibitory promoter is to be placed upstream of acetate overproduction pathway, so that the pathway can be automatically shut down during digestion. Several candidates are taken into consideration, including glucose inhibitory promoter rpoH P5 and fatty acyl-CoA inhibitory promoter fabB. In our experiments we tested the effect of rpoH P5 promoter to demonstrate our idea.
4. Bilateral switch
With the two pathways introduced into our engineered bacteria, it is now conferred the ability to consume fat assimilated and unassimilated. However, these two pathways have different optimal time of effect. For acetate production pathway, it is better not to be expressed during digestion process, for its product, acetic acid, would otherwise change the pH inside digestive track and therefore interfere with digestive enzymes. On the other hand, the best time to enhance β-oxidation is right when digestion occurs, for the hydrolysis of fat contained in a meal will yield a considerable amount of higher fatty acids during the process.
To coordinate this difference, we include a bilateral switch into our circuit so that our bacteria can express either pathway in a given time. It is constructed as follows:
- First, a recombinase system is introduced into our microbe, it includes a bacteriophage integrase and its recombination directionality factor (RDF). The integrase alone recognizes its two recognition sites attB and attP on DNA double helix, inverts the DNA sequence in between and converts attB and attP sites into two new recognition sites attL and attR.[5] When both integrase and RDF are present in the system, however, RDF changes the activity of the former, making it prone to reverse the process and generate attB and attP sites once again.[6]
Fig 3 The principal of recombinase system
- Then, a constitutive promoter flanked by the two recognition sites attB and attP is put in between acetate production pathway and β-oxidation pathway. So when only integrase is expressed, the promoter initiates acetate production; when both integrase and RDF are expressed, the promoter turns to activate β-oxidation.
Fig 4 Bilateral switch design
5. Inductive sensors
As described above, using our bilateral switch, we manage to alter between the two pathways by controlling the mere presence and absence of RDF. Consequently, a gene encoding RDF of which the expression is controlled by a promoter sensitive to digestion signal, together with a constantly expressed gene encoding integrase, will validate an automatic switch between acetate production and β-oxidation.
Several inductive digestion signals are taken into consideration – glucose, major digestive product of starch; bile acids, which are released into small intestine upon the beginning of digestion; and higher fatty acids, which if chosen, would serve as a signal to the initiation of its own degradation. Several candidates are potential to accomplish this task – deoxycholate activated promoter micF, glycocholate activated promoter osmY, and fatty acyl-CoA activated promoter fadD. Each is tested to decide the fittest for our circuit.
6. Colonization
To successfully colonize engineered bacteria inside human intestine, it is essential that neither do native microbes compete exclude engineered bacteria nor do engineered bacteria interfere too much with intestinal microbiome.
First of all, we choose E. coli Nissle 1917 as the chassis for our final product. E. coli Nissle 1917 is a probiotic strain derived from human intestine and therefore, it does not easily trigger immune response. Furthermore, to prevent our engineered bacteria from being excluded, we enhance their ability to proliferate, rendering them stronger power to compete. It has been proven by BNU-China 2018 that by overexpressing glucose dehydrogenase (GDH), a key enzyme in pentose phosphate pathway, a strain can gain a notable proliferative advantage (Fig.5). The same enzyme can be used to promote colonization of our microbe.
Fig 5 GDH overexpression promotes cell growth (adapted, with permission from BNU-China 2018[8])
On the other hand, we do not want our microbe to disorder native microbiome either. So, we include a quorum sensing system orthogonal to that of native microbes in our design. With the limitation of quorum sensing mechanism, the engineered bacteria slower their proliferation rate once they reach a certain population threshold, ensuring its not interfering with intestinal microbiome.
7. Biosafety
As a project that aims to colonize engineered bacteria inside human body, biosafety has always been a most important issue. Two safety modules are designed and tested for two different usages. First, once the microbe is colonized inside human intestine, we would like it to remain in our control, that is, we want to enable our users to terminate using our engineered bacteria at any time as they like. Second, we do not want our bacteria to contaminate the environment outside human body.
To enable our users to specifically eliminate the engineered bacteria, we put a gene encoding a translational inhibitor (MazF) downstream of an L-arabinose-activated promoter into our circuit. L-arabinose is an inducer harmless to human body and can be used by direct in-taking. Besides, there is relatively low a level of arabinose in human diet, so the kill switch will not be accidentally triggered. Meanwhile, as a translational inhibitor, MazF does not cause lysis of bacterial cells. Therefore, activation of this kill switch does no harm to the native microbes.
Fig 6 Induced suicide
As regards to contamination prevention, we make use of a toxin-antitoxin system – genes encoding a protein synthesis inhibitor (RelE) and its antitoxin (RelB) are associated in a constantly expresses polycistron. Translation of RelE gene is promoted by a regular RBS, whereas that of RelB, a temperature-sensitive RBS. Only when the temperature reaches human body temperature is the expression of RelB complete. When the engineered bacteria get out of human body, temperature drops, inhibiting expression of the antitoxin RelB, and the bacteria die of the excessive toxin RelE, preventing contamination.
Fig 7 Kill switch
