Revision as of 08:20, 26 October 2012 by Nanhai (Talk | contribs)

Project Description

The above figure is the overview of our design

Utilizing human microbiota to tackle diseases has long been the keen desire of scientists. This year, we WHU-China team engineered a kind of probiotics "E. coslim" from Escherichia coli, hoping to provide a new approach for treating obesity. Specifically, three genetic devices were designed. The first two devices were assembled to sense and response to fatty acids and glucose. To achieve these goals, promoters repressed by FadR and CRP were devised and synthesized respectively. When functional genes are placed downstream of these promoters appropriately, the first two devices are supposed to degrade fatty acids and convert glucose into cellulose respectively, thus preventing excessive calorie intake as well as producing prebiotics. Meanwhile, the third device was designed to control the density of "E. coslim". Also, xylose inducing death device was planned to terminate the effects of "E.coslim" at will to prevent horizontal gene transfer in future applications. As a whole, by modeling, we are developing "E. coslim" to regulate the microbiome composition in intestine to reduce risks of obesity.

Obesity: a severe global problem

Obesity refers to a health condition that body fat is accumulated to some extent. According to WHO, body mass index (BMI) is an index of weight-for-height that is commonly used to classify obesity in adults. It is a risk factor for various diseases, such as cardiovascular diseases (mainly heart disease and stroke), type 2 diabetes, musculoskeletal disorders (especially osteoarthritis), some cancers (endometrial, breast, and colon).

As it is shown in figure 1 and 2, a large amount of people from all over the world are overweight in both developed and developing countries and it is and will become more and more severe.

Figure 1(from reference [4]): Past and projected prevalence of overweight (BMI ≥25 kg/m2)

Figure 2: Prevalence of obesity in different countries.
(Picture from The Wellington Grey blog)

The Cause of Obesity

Obesity is most commonly caused by a combination of excessive food energy intake, lack of physical activity, and genetic susceptibility, although a few cases are caused primarily by genes, endocrine disorders, medications or psychiatric illness.

However, the problem of obesity emerged globally only several decades ago. Since the change of genome of a species requires a long time, the outbreak of obesity is unlikely to be caused by changes in human genome. For most individuals, controlling food intake and doing physical activity in a proper way are effective strategies to lose weight. But for some people whose health condition or current life pace keep them away from systemic and regular exercise and dieting, modulating the composition of microorganisms in intestine might act as an alternative.

Reports by Gordon have shown that, apart from human genome, the collective genome of microorganisms (microbiome) in human intestine is associated with our obesity [1]. Furthermore, microbiome is able to be changed through control of food intake [1].

Two groups of beneficial bacteria are dominant in the human gut, the Bacteroidetes and the Firmicutes. The relative proportion of Firmicutes increases in obese people by comparison with lean people [2].

Figure 3: How excess of energy contributes to obesity

Pertinent study by Gordon attested their initial hypothesis that changes in microbial component have a causal relationship with obesity, thus might have potential therapeutic implications [2] [3]. Colonization of germ-free mice with an ‘obese microbiota’ results in a significantly greater increase in total body fat than colonization with a ‘lean microbiota’ [3].
Figure from reference [3]

Present strategies to lose weight

Dieting, Exercise, Drugs and Surgery are major ways to lose weight. However, they all have many drawbacks. Dieting may cause nutritional imbalance and can be a heavy mental burden since the people may not be able to enjoy the food they want. Exercise requires regular time and is ineffective in many cases. Drugs and surgery may have many side effects and are many times more costly.

Our idea

Previous situations and insights construct our theoretical fundament. We try to utilize synthetic biology to provide a cheap, convenient, effective and safe approach for treating obesity. Instead of passive alternation of microbiota, we are trying to construct an engineered E.coli----- E.coslim to positively change microbiota in intestine. As Figure 3 shown, we place E.coslim in the role of sensing and consuming excessive energy, which thus leads to the double effects: lowering the proportion of Firmicutes and increasing that of Bacteroidetes, and decreasing the energy available in one’s intestine.

To achieve these two goals, we designed four devices, fatty acids consumption, cellulose synthesis, colonization and death device of E.coslim.


[1] Ruth E. Ley1, Peter J. Turnbaugh1, Samuel Klein1 & Jeffrey I. Gordon1 Microbial ecology: Human gut microbes associated with obesity. Nature 444, 1022-1023 (21 December 2006)
[2] Peter J. Turnbaugh1, Ruth E. Ley1, Michael A. Mahowald1, Vincent Magrini2, Elaine R. Mardis1,2 & Jeffrey I. Gordon1 An obesity-associated gut microbiome with increased capacity for energy harvest. Vol 444|21/28 December 2006| doi: 10.1038/nature05414
[3] Ley RE. Obesity and the human microbiome. Curr Opin Gastroenterol. 2010 Jan; 26(1):5-11.
[4] Y Claire Wang Health and economic burden of the projected obesity trends in the USA and the UK. Lancet. 2011


Fatty acids and sugar should be the primary targets for genetically engineered probiotics that can help people lose weight. For probiotics to degrade fatty acids and to convert glucose into cellulose, they must be able to sense and be regulated by emergence of those substrates. Otherwise, the system may not only be inefficient but also may cause serious problems. However, there is no promoter available in nature that can solely regulated by glucose or fatty acids. Therefore, to achieve our goals, we designed an indirect pathway and a direct synthetic promoter to sense and be regulated by glucose concentration. Also another synthetic promoter was designed to sense and be regulated by fatty acids.

Indirect Pathway Design

In a cell, the total amount of ATP, ADP and AMP molecules remains constant. Low glucose concentration results in high activity of adenylate cyclase converting ATP into cAMP, who binds and converts cAMP receptor protein (abbreviated as CRP) to DNA-binding configuration. Conversely, when glucose concentration gets high, more ATP and less cAMP will be produced, resulting in low DNA-binding activity of CRP.

We embeded gene cI of lambda phage(BBa_P0451) downstream promoter PcstA (BBa_K118011), which can be activated by the binding of CRP, and genes of red fluorescence protein(RFP, BBa_I13507) respectively downstream the promoter BBa_R0051 repressed by protein cI. In this way we construct an indirect regulation pathway with sensus glucose, transcription activator CRP and transcription repressor cI. If the device works as design, output of RFP will be increased following the elevation of glucose concentration, and vice versa.

The indirect regulatory pathway


Construction of plasmid for indirect regulation pathway

In this experiment, RFP reported the function of the indirect regulation pathway.

BBa_K861173: BBa_I13507, an mRFP generator with RBS and terminator, was embedded after CRP activated promoter BBa_K118011.

BBa_K861172: BBa_P0451, a cI generator with RBS and terminator, was embedded after promoter BBa_K118011 activated by CRP.

BBa_K861169: BBa_K861172 and BBa_I763007, a cI repressed RFP generator, were assembled .

BBa_K861174: BBa_K137115, constitutively expressing cI generator with promoter, RBS and terminator, was assembled to BBa_I763007

All new composite parts mentioned above were transformed to competent cells of Escherichia coli str. DH5a. All positive clones are validated using PCR, restriction enzyme digestion and DNA sequencing.

Cell culture fluorescence measurement

Minimal medium with different concentration of glucose(1mM, 4mM, 10 mM , 20 mM , 50 mM ,100 mM) was transferred into a 96-well plate, 200 μL for each well. Then each well was inoculated with 2 μL of seed liquor which was activated overnight in M9 minimal medium with 50mM glucose at 37℃. The wells without inoculation were regarded as blank controls to revise the results. Under each condition, three parallel samples were set. The plate was incubated at 37℃, 150rpm. Cell culture fluorescence was recorded on a SpectraMax M2 plate reader (Molecular Devices). Excitation at 584 nm and emission at 607 nm were used. All fluorescence was normalized with cell density by measuring the absorbance at 600 nm.

Capturing fluorescent image

Cell morphology was observed through fluorescence microscope, and the images of bacteria with different glucose concentration were captured.

Fluorescent analysis of cyto-imaging

A program named FANCY was designed to recognize single cell and calculate the fluorescence strength according to the images. For more information, please click Here.


Purified plasmids constructed before were digested with XbaI and PstI for confirmation. The agarose gel electrophoresis showed that the lengths were correct. At last, the plasmids were sent for sequencing. Results showed no mutation.

The result for fluorescence measurement of indirect device

In the cell culture fluorescence measurement experiment, fluorescence of BBa_K861173 decreased coordinating with glucose concentration, while BBa_K861169 was reversed.The fluorescence of BBa_K861174 was too low to record, so we do not show it here. All of the results coincided with expected results indicating that we have successfully constructed the promoter which was activated by high concentration of glucose.

The fluorescent image of indirect device in different concentration of glucose

Fluorescent images indicate that all cells were growing normally, because the size and morphology are both the same as that of the cells in LB medium. The fluorescence of the cells in the images shows the same discipline as results from the fluorescence measurement experiments.
The results of FANCY are showed as bellow,fluorescent intensity of PcstA+cI+RFP increased with the glucose concentration,while that of the PcstA+RFP decreased with glucose concentration.It conforms well the results that showed above.

The fluorescent intensity of indirect device from program FANCY


All results of the three experiments indicate the device works as expected. Next, RFP will be replaced with genes of cellulose synthesis. So the excess glucose can be transformed into cellulose.
Although the indirect regulation pathway was tested effective,it works through an intermediate product, protein cI. This determines that the device will be less sensitive to glucose than a direct regulation pathway without intermediate.


Direct Regulatory Promoter Design

1. Glucose sensor

To turn CRP into a repressor, we firstly consulted papers about CRP binding site and find out that 22-bp palindromic consensus site of the sequence AAATGTGATCT*AGATCACATTT. Unfortunately, this consensus sequence contain XbaI restriction site. We then designed the binding site by changing the most frequent bases in the binding site into second most frequent bases. Specifically, two binding sites were designed: AAATGTGATTTAAATCACATTT, AAATGTGATTATAATCACATTT.

Firstly, to construct promoter that can be directly repressed by CRP, we put modified CRP binding site between -35 and -10 region of promoter BBa_ J23110 (TTTACGGCTAGCTCAGTCCTAGGTACAATGCTAGC). However, it did not work out as we suppose to. In our reporter assay, this promoter failed to express higher level of RFP when bacteria grow in M9 medium with higher glucose concentration.

Then, we tried a different strategy. we designed another promoter overlapped with CRP binding site to satisfy the sequence of -10 region of the promoter (TTGACAGCTAGCTCAAATGTGATTTAAATCACATTT). The promoter also failed to show the desired function, the expression of RFP showed that the promoter is unstable.

Finally, we designed promoter with modified CRP binding site overlapped with the -10 region of the promoter. To satisfy better CRP binding, we firstly changed the sequence in -10 region to meet the CRP consensus sequence (TTGACAGCTAGCTCAAATGTGATTATAATCACATTT). We named it Pcar, which was short for Promoter of CRP As a Repressor . This time, Pcar showed exciting property as we expected.

2. Fatty acid sensor

In fact, fatty acid sensor had already been tried by iGEM2006_Tokyo_Alliance (BBa_J54220).Yet their design failed to give desirable function.
To design promoter that is under the sole regulation of fatty acid concentration, double FadR binding sites in promoter of FadL are placed and overlapped downstream of constitutive promoter J23110 (TTTACGGCTAGCTCAGTCCTAGGTACAATGCTAGCTGGTCCGACCTATACTCTCGCC ACTGGTCTGATTTCTAAGA). Also, FadR was overexpressed to prevent leaky expression of the promoter.


Although the indirect regulation pathway was tested effective, we went on attempting a more compact and widely useful direct regulation design. Hence we altered a constitutive promoter (BBa_J23119) to CRP repressible ones. We have established a new technical standard for our strategy of repressible promoter design (for more information, click on Standard), but we shall focus on the design itself now.

We designed promoter Pcar based on promoter BBa_J23119, inserting CRP-binding site to overlap on six base pairs with promoter -10 region. Since steric hindrance of CRP dimer blocks the function of -10 region, gene downstream will be repressed when glucose concentration is low. That is, most CRP appears in DNA-binding configuration. The repressive effect is undermined when glucose concentration increases. Accordingly, we changed CRP from an activator to a repressor, simplifying the device with potential advantages of higher sensibility and efficiency. As experimental results show, promoter Pcar works as we expect.


The direct regulatory pathway


Design of the promoter Pcar which is activated by glucose

Promoter Pcar, glucose biosensor plasmid, is derived from constitutive promoter (BBa_J23119) by adding a CRP binding site upstream the promoter which has several base pairs overlapping with polymerase binding site. The sequence was synthesized with restriction enzyme cutting site for EcoRI and XbaI at the 5' terminal and SpeI at 3' terminal. The sequence of promoter Pcar has cohesive terminus at both ends, so it is very convenient for us to construct the plasmid for functional detection.The sequence of Pcar is as followed:

The design concept of promoter Pcar

Construction of plasmid for direct regulation pathway

In this experiment, RFP reported the function of the indirect regulation pathway.

BBa_K861179: BBa_I13507, an mRFP generator with RBS and terminator was embedded downstream the constitutive promoter BBa_J23119

BBa_K861176: BBa_I13507 was embedded downstream the artificial promoter Pcar.

BBa_K861178: a constitutive expressed CRP(BBa_J23116+BBa_K861161) was assembled with K861176.

All new composite parts mentioned above were transformed to competent cells of Escherichia coli str. DH5a. All positive clones are validated using PCR, restriction enzyme digestion and DNA sequencing.

Functional detection

The same methods with that of the indirect regulation pathway were used to confirm that the promoter worked as expected. For details, please click Here.


Construction of the plasmid for functional detection

The sizes of the Promoters Pcar and J23119 are less than 100 bp and are proved to be correct by agarose gel electrophoresis. Restriction Digestion of the plasmid BBa_I13507 only have one lad on the agarose gel, it told that the plasmid was digested well. After transformation, competent cells were cultured on agar plate with 50 μg/L of ampicillin. Both red and white bacterial colonies emerged on one plate. The red ones were the correct clones revealing promoter embedded successfully, while the white ones were negative. The red clones were picked out and cultured in LB medium for plasmid extraction. Purified plasmids were digested with XbaI and PstI for confirmation. The bands of 2000bp and 1000bp showed that the promoter had been embedded successfully. At last, the plasmids we acquired were sent for sequencing, and the result shows no mutation exist.

The agarose gel for digestion comfirmation

Cell culture fluorescence measurement.

The correct clones were cultured in 96-well plate at 37℃ for 24 hours, then the fluorescence and absorbance at 600 nm were recorded on a SpectraMax M2 plate reader. All fluorescence was normalized with absorbance at 600 nm. The results represented the fluorescence of every well.
The fluorescence of K861179 was about 10000 Relative Light Units. It did not vary with concentrations of glucose. However we found a positive correlation between the fluorescence of K861176 and concentration of glucose. At a glucose concentration lower than 4mM, the fluorescence was very low, but at high concentration of glucose like 100mM, the fluorescence was much less than that of K861179.

The result of promoter Pcar from fluorescence measurement

Capturing of the fluorescent image

Fluorescent images indicated that all the cells were growing normally, because the size and morphology were both the same with cells in LB medium. The fluorescence of the cells in the images show the same discipline with results from the fluorescence measurement experiments. Fluorescence of K861179 was very strong but it didn't change with the glucose concentration. On the contrary, fluorescence of K861176 was relatively weak but increased with concentration of glucose.

Fluorescent images of Pcar and J23119 in different glucose concentration

Fluorescent analysis of cyto-imaging

The result of FANCY is showed as bellow, single cell was recognized from fluorescence images and fluorescence intensity was calculated.In the table, datas show that RFP expression was activated in high glucose concentration, which conforms well with results above.For more information about FANCY,click Here.

Fluorescent intensity of Pcar from program FANCY



The promoter Pcar is a promoter designed for the Escherichia coli and it is derived from a constitutive promoter BBa_J23119. Pcar includes the CRP-binding site and the RNA polymerase-binding site which overlap each other several base pairs. Therefore, because of the steric hindrance between CRP and RNA polymerase, gene downstream of the promoter will be repressed at high concentration of CRP. In the cells, low glucose concentration results in increasing activity by adenylate cyclase. cAMP binds to the cAMP receptor protein, which, in its bound form, is able to bind tightly to the specific DNA site in the promoter and to repress the gene downstream. On the contrary, high glucose concentration will result in the expression of the promoter.




The design of Pcar had sparkled the design of PfadR. In nature, beta oxidation needs these five genes, and is regulated by many factors besides fatty acids concentration. Therefore, there is no promoter that can meet our demand to be solely regulated by fatty acids. Therefore, we tried to develop a synthetic promoter that will be regulated by fatty acids. We found that the last 3 base pairs of constitutive promoter BBa_J23110 are the same to the first three base pairs of FadR binding site of FadL. Therefore, we thought that maybe by preventing the initiation of transcription, we can achieve our goal.

Design of the Promoter PfadR Repressed by Fatty Acids

Promoter PfadR, is derived from BBa_J23110. Specifically, FadR binding site of FadL gene was placed overlapping with the last 3 base pairs of BBa_J23110. The sequence was synthesized with restriction sites for EcoRI and XbaI at the 5' terminal and SpeI at 3' terminal. We use overlapping PCR to get the double strand DNA. The sequence design of PfadR is as followed:




For M9 medium using oleic acid as sole carbon source, oleic acid was first emulsified 10% Triton X-100. M9 medium was then slowly added with constant vortex. M9 medium with high concentration of oleic acid was diluted by M9 medium with triton to form various concentrations.

For M9 medium using glucose as sole carbon source, M9 medium with high concentration of glucose was diluted by M9 medium to form various concentrations.

After 24h of incubation in 24 well plates in 37°C, bacteria culture was centrifuged at 3000rmp for 5min, washed and resuspended in PBS. We detected the OD600 and fluorescence of using SpectraMax M2 plate reader (Molecular Devices) .Excitation at 584 nm and emission at 607 nm were used. All fluorescence was normalized with cell density by measuring the absorbance at 600 nm.


Normalized using Fluorescence/0D600

Blue: Constitutive promoter J23110

Red: PfadR

Glucose Concentration gradient: 0.5, 1, 5, 10 mM

Fatty acid Concentration gradient: 0.025, 0.05, 0.1, 0.25, 0.5, 1, 1.5 mM

PfadR and BBa_J23110 promoter strength at different glucose and fatty acids concentration

As shown from the results, the promoter shows about 3 times induction from glucose to 1.5umol/L fatty acids medium and the fluorescence of PfadR is about one sixth of BBa_J23110. This may be because the tandem FadR binding site has made it more difficult for Polymerase to start gene transcription. Also, in medium that use glucose as sole carbon source, PfadR seems to be leaky. However, since our bacteria is wild type E.coli, Fab genes was not mutated, meaning that bacteria can synthesis fatty acids. Therefore, there may be a basal level fatty acids concentration inside the cell, making the transcription not being totally repressed.

It should also be noticed that, from fatty acids concentration 0.025umol/L to 1.5umol/L, the induction is not very obvious. F0.25, 0.5 and 1 seemed to have similar fluorescence strength. The promoter is not sensitive enough. To further improve the function of PfadR, we are planning to modify the sequence of FadR binding sites to make FadR overlap more will promoter region. Also, we will try to overexpressed FadR protein to see its effects on PfadR.



To address the problem of sensing substrates in our metabolic devices, we have developed a few promoters following the strategy of modifying unconserved regions to protein binding sites. As steric hindrance can stop RNA polymerase as well as other transcription factors from correctly binding and functioning, the binding protein become the repressor of our synthetic promoters.

To different promoters, the details of modification comes in different ways. For instance, the PfadR (BBa_J861060) is modified downstream -10 region, while Pcar (BBa_J861171) is modified between -35 and -10 region. Indeed we have tried some other modifications (for more information, click on design of promoters) but without desirable functions and we still do not know why. However, all modifications aimed at the same goal, to introduce overlapping sequences between binding site of the designed repressor and the promoter.

To test the function of these promoters, we used fluorescence measurements. First, both the modified promoter and the unmodified one (as control) were assembled with mRFP genes as reporter. Hence, after culturing and testing fluorescence intensities using different methods (see indirect regulation and FANCY), we can learn if the synthetic promoter is regulated as expected.

After a few successes, we are endeavoring setting up a new technical standard on design, construction and functional test of promoters.


Fatty Acid Degradation Device

The above is the video introduction of fatty acid degradation device. For Chinese mainland visitor, please visit here for the video


To help people lose weight without the need of food restriction, we designed a genetically modified E.coli that can sense and degrade excessive fatty acid intake by the host. We hope that, together with other two devices we designed, we can introduce our E.coslim as resident in intestine to consume the excessive calories intake by the host and regulate intestinal microbiota.


Genes that are responsible for degradation and transportation of fatty acids (FAs) from E.coli K12 and from Salmonella enterica LT2 were cloned. Also, a promoter named PfadR that can be regulated solely by fatty acids was also designed. By placing those fatty acid degradation genes downstream of the artificially designed promoter PfadR (BBa_K861060), we hope to create a device that degrades FAs only when the concentration of FAs is high.

Long chain fatty acids are firstly imported by the transmembrane protein FadL. After FAs get into cells, a CoA will be added by inner membrane-associated FadD (acyl-CoA synthase). β-oxidation is initiated by FadE(acyl-CoA dehydrogenase), which will convert acyl-CoA into enoyl-CoA. The following cycles of hydration, oxidation, and thiolytic cleavage are carried out by tetrameric complex consisting of two FadA and two FadB proteins or two FadI and two FadJ in anaerobic condition. FadR is a transcriptional regulator that, when not binds to acyl-CoA, can either serve as an activator for fatty acid synthesis gene like FabA, FabB and etc., or a repressor for fatty acid degradation gene like FadA, FadB, FadD FadE, FadL, FadI, FadJ and etc. After long chain fatty acids are converted to fatty acyl- CoA by FadD, it can bind to FadR. The binding will alter the conformation of FadR, making FadR unable to bind to the DNA sequence it recognizes to fulfill its function. Therefore, FadR can no longer activate or repress the transcription of genes downstream FadR binding sites. However, to our knowledge, there is no promoter available in nature that can respond solely to FadR since those promoters are often regulated by glucose concentration or oxidative stress and many other factors.
In our design, FadL, FadD, FadE, FadA, FadB, FadI and FadJ from Escherichia coli K12, and FadA, FadB and FadE from Salmonella enterica LT2 are placed downstream a synthetic promoter PfadR to make them solely regulated by fatty acid concentration.


Cloning of the gene

First, the genome of Escherichia coli K12 str. DH5ɑ and Salmonella enterica LT2 (symbolized as S-) were extracted and amplified by PCR using primers for each gene. The sequences of the primers used are as bellow (5’---3’).

Design of the Promoter PfadR Repressed by Fatty Acids

Promoter PfadR, is derived from BBa_J23110. Specifically, FadR binding site of FadL gene is placed overlapping with the last 3 base pairs of BBa_J23110. The sequence was synthesized with restriction sites for EcoRI and XbaI at the 5' terminal and SpeI at 3' terminal. We used overlapping PCR to get the double strand DNA. The primer design of PfadR is as followed:

Construction of Biobricks

Fatty acid degradation project is divided into two parts: promoter, and gene function.
To discover the optimal combination of those fatty acid genes, we:
(1) use PCR to clone those genes in E.coli K12 and Salmonella enterica LT2
, (2) restriction digest and ligate those gene into pSB1C3,
(3) restriction digest and ligate those gene with RBS(B0030),
(4) RBS-FadA, RBS-FadI, and RBS-S-FadA is ligated with both BBa_R0011 promoter and our PfadR
RBS-FadR, RBS-FadB, RBS-FadJ, RBS-FadE, RBS-FadD, RBS-FadL, RBS-S-FadB, and RBS-S-FadE are ligated with B0034,
(5) Promoter-RBS-FadA is ligated with RBS-FadB-Terminator, PROMOTER-RBS-FadI is ligated with RBS-FadJ-Terminator and Promoter-RBS-S-FadA is ligated with RBS-S-FadB-Terminator. RBS-FadE-Terminator, RBS-FadD-Terminator, RBS-FadL-Terminator, and RBS-S-FadE-Terminator, are ligated with BBa_R0011 promoter, PfadR and various constitutive promoters. For Promoter PfadR
(1) promoter PfadR was synthesized using overlapping PCR
(2) RFP reporter was ligated downstream the promoter and ligted into pSB6A1
(3) J23116+ RBS+ FadR+ Terminator was ligated to PfadR+ RFP in pSB6A1

Experimental Procedure

Cupric acetate-pyridine reaction

We used cupric acetate-pyridine as a color developing reagent to determine fatty acid consumption of genetically modified bacteria. We had modified existing methods to extract free fatty acid in M9 medium. Also, we used IPTG induced promoter BBa_R0011 to see the expression of those proteins and extract proteins from cells. For more details, please see protocol: Cupric-Soap Reaction for more details.

We also conducted in vitro experienments in which we characterized fatty acid degradation capabilities of combination of enzymes using a cell free system. Please see protocol: In vitro Experiment for more details.


Characterization of each gene

In this experiment, we wanted to test whether the ability of degrading fatty acids of our genetically modified bacteria was enhanced as expected by transforming plasmids constitutively expressing related genes in the β-oxidation pathway. The effects of the genes we tested is listed in the following chart I. The ability was reflected by the change of the concentration of the fatty acids in the medium. It was measured by cupric-acetate soap reaction as described in Protocols section. Each time we inoculated 50mg bacteria into 30ml M9 medium using fatty acid as sole carbon source, collecting the sample at the time as shown in the picture. Then the analysis of the free fatty acids was performed.

The following figures shows the effects on degrading fatty acids by expressing different genes in β-oxidation pathway in E.coli. They are under the regulation of promoters with different kinds of strength. J23107 and J23114 are constitutive promoters provided by the committee. PfadR is the promoter designed by ourselves. It consists of the sequence of a constitutive promoter and the binding sequence of the transcription factor, FadR, which is the sensor of the fatty acids. FadE is the acyl-CoA dehydrogenase, which have been proved as performing the rate limiting reaction in the pathway. S-FadE is the counterpart of FadE in the bacteria Samonella. FadD is the acyl-CoA synthase. FadL, a transmembrane protein, is responsible for transporting fatty acids into the bacteria. The control we used is the E.coli expressing galU, a gene responsible for synthesize cellulose.

Fatty acid degradation at 6h

Fatty acid degradation at 12h

Fatty acid degradation at 18h

Fatty acid degradation at 24h

Fatty acid degradation of bacteria overexpressing each gene in 24h

Based on the measurements of the consumption at given time, we conclude that overexpressing FadL increases the metabolizing ability no matter under the regulation of J23114 (BBa_K861002) or our designed promoter, PfadR(BBa_K861003). The advantage is more obvious when the time expands (consumption at 18h and 24h). It's plausible because more fadL may transport more fatty acids into the bacteria. The increased inner fatty acids concentration is quite favorable. We also notice that the later one's consumption is lower. It may attribute to the fact that our promoter is weaker than the J23114. If the copy number of FadL is less, its metabolizing rate will be slower. And the fact that the PfadR needs to be induced may also make the time needed to synthesize protein longer, which may make it less competitive. These data opposes to our assumption that overexpressing the rate limiting enzyme FadE ((BBa_K861025 and (BBa_K861026) doesn't have obvious effect. It may be because the original level of FadE is enough, thus overexpression is not needed. The strength of the promoter doesn't affects the rate much, which partially suggests the reason above.

We found that the slope between 12h and 18h and between 18 and 24h are less than the others. It may be because the bacteria have entered static status, the amount of bacteria becomes consistent. Also, after the first death phase, they entered logarithmic phase again. Since our inoculation is relatively large and the oleate is excessive, the situation that the bacteria has experience two life cycle is possible. A growth curve in the future can test the theory.

In vitro Experiment

The in vitro experiment was designed to make up for the limitation of time to test the combination of expressing different genes together (We are short of time assembling the genes together). So we used the cell extracts to do the enzyme assay. The advantage is that we can easily mix the enzyme we want together. Since the purpose of this experiment is to test whether overexpressing multiple genes are superior to a single gene, we set the amount of every gene the same in the combination for simplicity. In the future, we can test more combinations to find the best ratio.

Fatty acids remaining after 6 hours of reaction

The result was that the cell extract of bacteria overexpressing FadE (BBa_K861024), FadD (BBa_K861013), S- FadA S- FadB(BBa_K861038) (a regulon), FadI FadJ(BBa_K861037) (a regulon) separately all degraded oleate obviously faster than the control. This may be somewhat contrasting to the result of our in vivo assay, in which overexpressing the FadD and FadE did not have obvious results. However, it can be explained. The promoter is not that strong in the in vivo assay, otherwise the growth of the bacteria would be inhibited. However, in the in vitro assay, this consideration was not necessary. Also, the concentration in the final reaction system was quite high by collecting 1L medium, which may improve the metabolizing rate.

It also showed that simultaneously increasing the concentration of FadE, FadD, S- FadA and S- FadB significantly improved the degrading ability. Increasing the doze of the FadI and FadJ on the basis of above did not make any difference, while increasing FadA and FadB may be indispensible because only expressing FadD and FadE is worse than expressing the gene alone.

Device II: Cellulose Synthesis

The above is the video introduction of cellulose synthesis degradation device. For Chinese mainland visitor, please visit here for the video.


Cellulose is an essential material for keeping intestine peristalsis without producing energy, as prebiotics, feeding vegetarian bacteria flora (including Bacteroides, whose appropriate amount has proved important to prevent obesity) of intestine as well. Thus, cellulose helps people keep slim and healthy.

The developing device aims at transforming glucose into cellulose, thus producing cellulose as well as reducing energy intake. To achieve this goal, we cloned genes of enzymes responding to cellulose synthesis from the Escherichia coli str. DH5α, constructing functional expressional elements with these genes respectively downstream of promoter activated by glucose. In this way, cellulose synthetase complex is built artificially under regulation of glucose, repressed under low concentration of glucose and activated under high concentration of glucose.

In the future, this device can be integrated to the assembled E. coslim, activated when excess glucose is sensed in intestine, converting it to cellulose.

The same as device I (fatty acid degradation), on one hand, we divide our work into two parallel sections. Function section includes a series of molecular biological manipulation on four genes of the cellulose synthetase complex and another two genes responding to produce substrates for cellulose synthesis. On the other hand, the design, construction and function tests of glucose-activated promoter belong to regulation section.


Genes to be Cloned

4 genes, bcsA, bcsB, bcsZ and bcsC, from the rdar morphotype bacterium, are involved in cellulose biosynthesis.

BcsA is considered to be the catalytic subunit.

BcsB can be activated the soon it binds to c-di-GMP.

BcsZ encodes endo-1,4-D-glucanase which belongs to glycosyl hydrolase family Ⅷ. Activation of BcsZ is required for optimal synthesis and membrane translocation of cellulose.

Although BcsC is transcribed constitutively, cellulose synthesis occurs only in the circumstances of AdrA.

AdrA ,a diguanylate cyclase (DGC), cyclizestwo GTPs into c-di-GMP. In turn, the activity of cellulose synthase can be increased when binds to c-di-GMP.

GalU catalyzes the addition of UTP to α-D-glucose 1-phosphate to yield UDP-D-glucose, which is the substrate for cellulose synthase complex

GalF is a predicted subunit of a GalU/GalF protein complex involved in colanic acid building blocks biosynthesis

Plasmid construct concept

After the indirect regulatory pathway and promoter Pcar being tested to be effective,we can embed the genes which are involved in cellulose synthesis downstream any one of them.So cellulase can only be expressed when glucose concentration is high, and also the expression will increase with glucose concentration. In this way, our E.coslim will have the ability to convert excess glucose into cellulose.


Clone of genes

As for the genes that we cloned, there is no difference between E. Coli str. K12 MG1655 and more available DH5α. we purified and amplified these genes from genome of Escherichia coli str. DH5α using PCR. The primers contain standard restriction enzyme cutting sites. The sequences of the primers used are as below.








Then the genes were digested with restriction enzymes and assembled to RBS (BBa_B0030) and terminator (BBa_B0024).

Construction of the plasmid expressing cellulose synthetase controlled by promoter we designed

All coding sequences were assembled to RBS and terminator, afterwards, they were embedded downstream the promoter Pcar, which can be activated at high glucose concentration.
The biobricks constructed were showed as bellow:

  BBa_K861102: Pcar+RBS+bcsA+terminator

  BBa_K861112: Pcar+RBS+bcsB+terminator

  BBa_K861122: Pcar+RBS+bcsZ+terminator

  BBa_K861132: Pcar+RBS+bcsC+terminator

  BBa_K861142: Pcar+RBS+galU+terminator

  BBa_K861152: Pcar+RBS+galF+terminator

  BBa_K861074: Pcar+RBS+adrA+terminator

All new composite parts mentioned above were transformed to competent cells of Escherichia coli str. DH5α. All positive clones are validated using PCR, restriction enzyme digestion and DNA sequencing.

Detection of Cellulose Synthesis

To detect the cellulose synthesis, we used cellulase to degrade cellulose in the cell culture. Then total reducing sugar in the culture was measured. So the difference of total reducing sugar between culture before and after treated with cellulase represents the total cellulose synthetised by the cell. For detailed information, please click Here.


Clone of genes

The gene bcsA is 2619bp, bcsB is 2340bp, bcsZ is 1107 bp, bcsC is 3474bp, galU is 909bp and galF is 894 bp. After PCR amplification, DNA fragments were examined by agarose gel electrophoresis. All genes proved correct. Then the genes were digested with restriction enzymes and embedded into plasmid backbone pSB1A2. To confirm the accuracy of sequences, positive clones were sent for sequencing after transformation. And the results showed that no mutation existed in genes.

Detection of Cellulose Synthesis

After treating with cellulase, total reducing sugar in supernatant and deposits was measured by methods described in our protocol. Colors in the tubes
becoming darker meant that reducing sugar increased with time. Amount of reducing sugar was calculated according to standard curve of glucose.

The formula of standard curve is as bellow:


In our experiment, cells that expressed protein which was nothing to do with cellulose synthesis was used as a control.We transformed the seven genes involved in cellulose synthesis into E.coli and meaaured cellulose production with method mentioned above. In the step of measuring total cellulose production, exceed cellulase was appended and incubated at 50 ℃ for 1 hour, and results show that some of the seven genes help increase the ability of cellulose synthesis.

In addition,AdrA is responsible for the production of a small molecule,c-di-GMP, which is known as an activator of cellulase. So we co-overexpress adrA with bcsA. Cellulose output in both supernatant and deposit were show in the following figure. The cellulose production in adrA+bcsA is more than 0.6 mg/mL, which is almost two times higher than that of control.



In our experiment, results show that cellulose yield in adrA+bcsA is higher than that of the single gene. So it indicated that the co-expression of some genes would increase cellulose production. In the next stage, we will try more combination of these genes and find the most efficient one

Actually, BcsB can be activated by c-di-GMP. But till the deadline, we have not successfully assembled Pcar and bcsB. In the following work, we are going to assemble a plasmid including all of the seven genes. Maybe in this way, cellulose production will increase greatly.

Other than application in the project of E.coslim, the cellulose device can also be used in many other fields, such as papermaking industry, biomedical materials, audio equipment and so on.


The above is the video introduction of colonization device. For Chinese mainland visitor, please visit here for the video


One big challenge of probiotics is their survival in intestine. We respond to this challenge by expressing gene adrA responsible for manufacturing the second messager c-di-GMP, a magic molecule that leads to inhibition of motility and increase of adhesion and division of E.coli.


AdrA protein can convert GTP into c-di-GMP, a magic second massager that, besides promoting the production of cellulose(for more details, see our celluose synthesis device), can reduce the expression of flagella and acute virulence gene simultaneously. In the same time, c-di-GMP can facilitate the synthesis of various adhesins and exopolysaccharides and can promote the proteolysis of replication inhibitors. As a result, AdrA makes cells become adhesive and promote the formation of biofilm, making the bacteria gain advantages to survive in the hostile environment of intestine.

Experimental Procedure

We tested the function of AdrA gene by plate assay. Formore details, please visit our protocol page.


As can be seen from the plate. Clone with AdrA constitutively expressed using BBa_J23107 is much more smaller compared to the control that only contain the plasmids of RBS, though the amount of bacteria is similar. This result showed that the AdrA was successfully expressed, elevating c-di-GMP level, leading to the increase expression of adhesins and inhibition of motility.

Death Device

The above is the video introduction of death device. For Chinese mainland visitor, please visit here for the video


We plan to place endonuclease and YhjH gene downstream the xylR repressed promoter (BBa_K861200) to make bacteria die and lose their adhesion when exposed to xylose.


In order to terminate the overreaction of fatty acid metabolism and cellulose synthesis, we design an elaborate device. Firstly, we determined to use endonuclease YhjH (BBa_K861090) responsible for c-di-GMP degradation to counteract the impact of AdrA and BglII( (BBa_K112106)), a protein that can kill the microbe without inducing cell lysis. Since AdrA but not YhjH contains this restriction site, our E.coslim will lose their adhesion more easily. To control the expression of endonuclease and YhjH, a regulator is necessary. We selected xylose as the inducer because this monomer hardly exists in nature, so the GMOs will not be killed under natural condition. Also, it is healthy and hard to be absorbed by human beings so the concentration of xylose in the gut can keep high for a long time. We are planning to device another synthetic promoter repressed by XylR from B.subtilis. We hope that eating xylose will subsequently derepress the expression of endonuclease and YhjH, ending up with the loss of adhesion and the death of our E.coslim, therefore efficiently terminate the process of lossing weight. For more information and for a brief introduction of our two-capsule design, please visit our Future Perspective .

Prevention of HGT

In human gut, there are billions of microbes. Horizontal gene transfer of those high efficient metabolic genes between GMOs and normal intestine residents may get out of control and cause disastrous effects. Therefore, a mechanism to prevent HGT is highly necessary. To prevent horizontal gene transfer, we designed a two-plasmids system. In one plasmid, xlyR repressor encoding gene (BBa_K143036) from Bacillus subtilis will be assembled with constitutively expressed promoter. On the other plasmid, metabolic gene will be coupled with xylR repressed death system. In our GMOs, since death system is suppressed by xylR in low xylose concentration, the bacteria will not die. However, when the HGT of those high efficient metabolic genes happens, the death system will not be suppressed, and the recipient will be killed.