Team:ECUST China/Experiments

Cellulose degradation

For this part, we want our chassis organism degrade lignocellulose into cellobiose firstly, so we designed a plasmid which can express C. fimi endoglucanase CenA and C. fimi β-1,4-exoglucanase Cex and can secrete these two enzymes out ofE. coliusing hemolysin secretion system, which is the best-characterized and most widely used TISS is the ahemolysin (HlyA) secretion pathway from uropathogenic E. coli. The secretory machinery of this pathway consists of three components: HlyB, an ATP-binding cassette (ABC transporter); HlyD, a membrane fusion protein; and TolC, an outer membrane protein( Su L. et al, 2012).

Figure 1.1. Gene circuit of cellulase secretion system
cex: Cellulase exoglucanase gene, cenA: Cellulase endoglucanase gene,hlyA, hlyB, hlyD: α-hemolysin system gene

In order to use E. coli DH5α to express recombinant protein into the growth medium through the a-hemolysin secretion pathway, the α-hemolysin specific signal peptide HlyAs, which corresponds to residues 965-1024 of α-hemolysin, was fused to the C-terminus of CenA and Cex, which genes was modified by overlapping PCR. In addition, the genes encoding the components of translocator, HlyB and HlyD, were amplified together by PCR using part:BBa_K1166002as template.

Enzyme activity of Cex-HlyA/CenA-HlyA

To make sure our final construction would work, we first analyzed the enzyme activities of the fusion protein CenA-HlyA and Cex-HlyA. hlyA was amplified from part:BBa_K1166002 and cenA was amplified from part:BBa_K523015, while cex was amplified from the gene synthesized by Genewiz, and the sequence was obtained from NCBI. The PCR products were cloned into a modified pET28b (+) plasmid pIN2 and enzyme activity within cells harboring cloned genes were tested qualitatively using carboxymethyl cellulose (CMC) substrates and FPA (Filtered Paper Assay) substrates( Duedu KO, French CE, 2016 ) .

Figure 1.2. Plasmids for Cex-HlyA and CenA-HlyA expression
cex: Cellulase exoglucanase gene, cenA: Cellulase endoglucanase gene,hlyA, hlyB, hlyD: α-hemolysin system gene

Figure 1.3. PCR amplification results by 1% Agarose Gel
Lane 1: linearized pET-cenA backbone; Lane 3: pET-cex Lane 4: hlyA-cenA; Lane 6: hlyA-cex; Lane 8: CenA; Lane 9: hlyB&hlyD

Figure 1.4. Plates of DH5α transformants (cenA-hlyA/pIN2 and cex-hlyA/pIN2)

Due to mRFP had been inserted into our backbone pIN2 for the construction of the invertor system, the negative colonies would show red for expressing mRFP, and white colonies were mostly positive. We conducted colony PCR and Sanger sequencing for further confirmation.

Figure 1.5. Colony PCR of cenA-hlyA/pIN2 plate of cenA gene

Figure 1.6. Sanger sequencing results for cenA-hlyA/pIN2 and cex-hlyA/pIN2 plasmids

After successfully constructed both clones, we followed Crude enzyme extraction protocol and cellulase enzyme activity assay protocol to obtained the enzyme activity of Cex-HlyA and CenA-HlyA, and the results are shown below. We also measured the enzyme activity curve of CenA alone for the characterization of Part: BBa_K523015.

Figure 1.7. Enzyme activity of different temperature and pH, measured by CMCNa assay

Figure 1.8. Enzyme activity of measured by CMCNa assay and FPA assay (pH=7, temp.=37℃)

To fully verify our cellulase hydrolysis products were mostly cellobiose for the activation of cellobiose promoter and as a result turning off cellulase expression and turning on bacterial cellulose synthase, we performed TLC on silica gel plates using a ethyl acetate/ methanol /acetic acid/distilled water (12:3:3:2, V/V/V/V) mixture as the mobile phase for our FPA assay products and CMCNa assay products and results are as follows:

Figure 1.9. TLC results for FPA assay products and CMCNa assay products
Our products show spots at the same horizontal position as cellobiose standard solutions, revealing that the reaction products have the component of cellobiose

Secretion efficiency of Cex-HlyA/CenA-HlyA

After validating the secretion ability of a-hemolysin system by HlyA fused with RFP, we also want to make sure that Cex and CenA can be secreted and to analyze the secretion efficiency separately, we planned to construct pCDFDuet-MCS1-cenA-HlyA-MCS2-HlyBD and pCDFDuet-MCS1-cex-HlyA-MCS2-HlyBD.

Figure 1.10. Plasmids for cex-hlyA-hlyBhlyD and cenA-hlyA-hlyBhlyD expression

Figure 1.11. PCR amplification results by 1% Agarose Gel
Lane 1: linearized pCDFDuet-MCS2 backbone; Lane 2: HlyBHlyD Lane 7: CenA; Lane 8: HlyA-CenA; Lane 9: HlyA-Cex

Figure 1.12. Colony PCR pCDFDuet-MCS1-MCS2-HlyBD plate (HlyBHlyD gene and nick 1/2)

Figure 1.13. Schematic presentation of overlapping PCR drawn by Iris Hu

Firstly, the hlyBD was cloned into MCS2 by overlapping PCR. After obtaining pCDFDuet-MCS1-MCS2-hlyBD, we tried multiple methods to insert cenA-hlyA into MCS1, such as seamless cloning, double enzymes digestion and ligation, and overlapping PCR, but all these attempts failed for a mysterious missing of a part of hlyBD in the final construction while the sequence of pCDFDuet-MCS1-MCS2-HlyBD was proved to be correct by Sanger sequencing of the whole sequence.

Figure 1.14.cenA-hlyA fragments and pCDFDuet backbone fragments after Double Enzyme Digestion & cex fragments prepared for overlapping PCR

Figure 1.15. Colony PCR of pCDFDuet-MCS1-cenA-HlyA-MCS2-HlyBD for cenA gene and pCDFDuet-MCS1-cex-HlyA-MCS2-HlyBD for nick 1/2/3.

Figure 1.16. Sanger sequencing results of pCDFDuet-MCS1-MCS2-HlyBD, pCDFDuet-MCS1-cenA-HlyA-MCS2-HlyBD, pCDFDuet-MCS1-cenA-HlyA-MCS2-HlyBD
The results shows that the sequence of the first plasmid is correct while the sequence of later two plasmids are incorrect.

At last, due to the intense deadline, we couldn’t finish this construction on time, nevertheless the results of Enzyme activity within cells and secretion efficiency demonstrated by mRFP can shed light on the feasibility of our ultimate goal.

Supernatant enzyme activity of Cex-HlyA/CenA-HlyA

Interestingly, we inserted cenA-hlyA, cex-hlyA, hlyBhlyD into the plasmid pIN2 of our invertor system just before deadline. After consulting former iGEM characterizations, we learned that the secretion efficiency is relatively low, so the supernatant was concentrated by 20 times for better visualization.

Also, we used the above two plasmids: cex-hlyA/pIN2 and cenA-hlyA/pIN2 to measure to secreted enzyme activity for Cex and CenA, and the results are as below.

Figure 1.17. Supernatant enzyme activity of CenA measured by CMC-Na assay (pH=7.0 temp.=37℃)

Collaboration with Hubei University

Due to the fact that all of these proteins were fused with Flag-tag, we were able to detect their presence by Western Blot using a mouse-derived anti-Flag primary antibody and detection antibody probed with fluorescence.

Figure 1.18. Western blot of ZM4 and E. coli T1 different secretion system

Collaboration with Hubei University

From the Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and Western blot, we can conclude that the Secretion Efficiency of PhoD-CelA outshines other two types, while ZM4 showed higher expression efficiency of GAP promoter than E. coli T1.

Figure 1.19. SDS-PAGE of ZM4 and E. coli T1 different secretion system

Bacterial cellulose synthesis

This part consists of two parts:

  • Cellobiose to glucose
  • Glucose to bacterial cellulose
Since the substrate of BC synthesis is UDP-Glc, cellobiose should be degraded to glucose or glucose-phosphate firstly by Cep94A and then be transformed to bacterial cellulose via AcsABCD.

Figure 2.1 Gene circuit of bacterial cellulose synthesis

Cellobiose to glucose

Short cellulose from pulp fibers were initially degraded to cellobiose which could accumulate in the cytosol, then the cellobiose needed the further degradation by cellobiose phosphorylase (Cep94A).

Cellobiose phosphorylase is an enzyme that catalyzes the chemical reaction cellobiose + phosphate → alpha-D-glucose 1-phosphate + D-glucose. We obtained the sequence of cellobiose phosphorylase from Part:BBa_K2449003. Unfortunately, this part was not contained in iGEM distribution kit, so we synthesized this gene (cep94A) from Genewiz.

Figure 2.2 Gene circuit of cep94A

In order to characterize the function of Cep94A specifically, we cloned the cep94A into p15a. Firstly, we successfully got the target gene and backbone through PCR. Because of the genes were all flanked with homologous end sequence, we used the Gibson assembly to construct the final plasmid.

Figure 2.3 The PCR results of cep94A and p15a backbone

Ligated plasmids were transferred into E.coli DH5α competent cell, then applied onto the Kan-containing LB plate. After incubating the cells at 37℃ for 12 hr, we selected 8 colonies of each plate and used the test primer to verify whether the fragments linked properly.

Figure 2.4 Colony PCR results shown on 1% agarose gel

We chose the positive colonies (number1,4,7) which were also sequenced successfully to perform the further characterization experiment.

Figure 2.5 SDS-PAGE results of Cep94A (about 91kDa)

To make out whether the Cep94A protein was expressed, SDS-PAGE was performed. The result showed that Cep94A was actually expressed (about 91kDa) after induction.

After verifying the expression of Cep94A, we decided to measure the Cep94A enzyme activity to perform the quantitative experimental characterization. After induction for 16 hours, the cells were collected and resuspended in 10 ml of PBS buffer, afterwards sonicated to obtain 10 ml supernatant. The enzyme activity of Cep94A was measured by the DNS method: 2 μL cellobiose +1 μL enzyme solution + 17 μL citrate buffer (pH =4.8) was reacted at 50 °C for 30 min. Then 30 μL DNS reagent was added, boiling for 10 min. The reacton mixture was diluted 5 times and transferred to 96-well microplate to measure the OD540 with a microplate reader.

Figure 2.6 The DNS assay of Cep94A

Enzyme activity = moles of substrate converted per unit time. So we could measure the activity of Cep94A by monitoring the rate at which a product (glucose) formed. Firstly, we drew the glucose standard curve: y = 0.6746x - 0.0491,R² = 0.9775 (y represents OD540, x represents glucose concentration), then the enzyme activity was calculated based on the glucose standard curve. Pluging OD540=0.093 into y, we worked out that x (glucose concentration) = 0.211mg/ml. According to the formula: Cep94A enzyme activity (U/mL) =enzyme dilution factor×moles of glucose /reaction time, we calculated that the enzyme activity of Cep94A was 0.781U/ml.

Figure 2.7 Glucose standard curve

Glucose to bacterial cellulose

After the degradation of cellulose into glucose by above-mentioned steps, our chasis was supposed to utilize the glucose immediately and produce bacterial cellulose. We chose the cellulose synthase from the Gluconacetobacter. xylinusacsA (the catalytic domain), acsB( the cyclicdi-GMP-binding domain), acsC and acsD (two accessory protein). These four genes, invented by Imperial London 2014 iGEM team, were in the distribution kit the iGEM send to us.

Since the plasmid containing acsAB, acsC,acsD was too big to construct (almost 12kb), we decided to characterize acsAB and acsC, acsD separately. By using the mini-prepared plasmid as template, we got the desired fragments as follows:

Figure 2.8 The PCR results of target genes of BC synthesis

Since the fragments were all flanked with homologous end sequence, we used the Gibson assembly to obtain our final plasmid.Ligated plasmids were transferred intoE. coliDH5α competent cell, then applied onto the Chl-containing LB plate and incubated at 37℃ for 12hr. After colony PCR and sequencing, we obtained the positive clones.

For acsAB, we verified its function using congo red binding assay. After inducing AcsAB-containing cells with 0.4mM IPTG in 100mL LB supplied with Chl at 37 °C, cells were pelleted at 8000rpm for 5minutes and the supernatant was kept for further analysis (both empty vector controls and un-induced controls were also assayed). Cells were resuspended in 1xPBS and sonicated for 3 minutes as on/off cycles of 15 seconds. Both the pellet and supernatant were kept for further analysis. Congo Red was then added to the samples up to a final concentration of 20μM. Samples were incubated for 2h at static conditions and at room temperature to allow for Congo Red binding. Then the samples were transferred to 96-well plate for absorbance measurements at 490nm. Only if the free congo red existed (not binding to bacterial cellulose), could the absorbance at 490nm be detected.

Figure 2.9 The congo red binding assay of acsAB

The results revealed that the cells which expressed acsAB had the obviously lower absorbance at 490nm. Illustrating that there was bacterial cellulose produced in the cytosol.

In terms of acsC and acsD, it is difficult to characterize its function separately. So we just verified the expression of acsC and acsD via SDS-PAGE.

Figure 2.10 SDS-PAGE results of AcsC and AcsD

The SDS-PAGE results illustrated that AcsC and AcsD were exactly expressed(almost 138.8kDa and 17.4kDa respectively).

Having the prospective results of acsAB and acsCD, now we are working on constructing the whole plasmid which includes acsAB and acsCD. We hope that the cells with this plasmid could express functional cellulose synthase and synthesize bacterial cellulose.


In order to achieve the purpose of switching the procedure of cellulose hydrolysis and bacterial cellulose synthesis automatically. A novel regulator was designed, containing inverter system and cellobiose response element. Inverter just liked the switch of the two functional genes while cellobiose realized the automatic regulation of this switch.

Inverter system

Inverter consists of lactose operon, CⅠprotein and PλR012. We constructed the inverter into a dual-plasmid (pIN1-pIN2) and used GFP and RFP to verify the effectiveness of inverter(Yokobayashi Y et al.,2002).

Figure 3.1 Gene circuit of inverter system

All of the biobricks of the inverter were obtained from iGEM distributon kit via PCR while the plasmid backbones were stored in our laboratory.

Figure 3.2 The PCR results of the parts of pIN1

Figure 3.3 The PCR results of the parts of pIN2

After obtaining the PCR products, we ligated the backbone and target gene using Gibson-Assembly method. Ligated plasmids were transferred into E.coli DH5α competent cell, then applied onto the Chl-containing LB plate(pIN1) and Chl-containing LB plate(pIN2) . After incubating the cells at 37℃ for 12 hr, we selected several colonies of each plate and used the test primer to verify whether the fragments linked properly.

Figure 3.4 Colony PCR results shown on 1% agarose gel

We chose the positive colonies (pIN1-2 and pIN1-9; pIN2-2 and pIN2-6) which were also sequenced successfully to perform the further characterization experiment.

Detailed circuit measurements were performed in liquid cultures. pIN1 and pIN2 were initially co-transferred to E.coli DH5α competent cell. The positive colony was then grown in 5mL M9 at 37℃ for 12 hr, then transferred to 100 mL M9 for enlarge cultivation. After the cells growing to OD600=0.5, we added different concentration of IPTG(0.1μM,1μM, 10μM, 100μM, 1000μM ).

Figure 3.5 Liquid M9 media in 15hr of dual-plasmid (pIN1 + pIN2) in different concentration of IPTG

We took samples of each IPTG concentration before and after induction (almost 16hr after induction). The samples were then transferred to a 96-well microplate in which RFP fluorescence (590 nm excitation, 645 nm emission, top 50% cutoff) was measured by using a fluorescence microplate reader. The fluorescence data were normalized against cell densities which were measured by using a microplate reader at 600 nm.

Figure 3.6 Fluorescence intensity induced by different concentrations of IPTG
Measurements of the circuits in response to varying IPTG levels were summarized in the transfer curve, where each point on the curve represented fluorescence data from three independent cultures of the same circuit under the same induction conditions.

Just as shown in the figure, as the concentration of IPTG rose, the expression of mRFP decreased, indicating that the inverter system worked effectively. The LacI protein was constitutively expressed from PlacIQ and repressed the Plac promoter.Plac transcriptional activity was controlled by modulating the concentration of an externally added inducer, IPTG. The expression of the CI repressor was controlled by Plac. Repressor CI acts on PλRO12 on pIN2 to repress the transcription of the mRFP gene, the output fluorescence indicator. So without IPTG, the CI level was low and mRFP level was high; adding IPTG increased CI levels and in turn decreased mRFP levels.

Cellobiose response element

The cellobiose response element is our new biobrick. We chose the cellobiose as the response element ranther than glucose because cellobiose could accumulate in the cytosol, while glucose is easy to metabolize. Only when the cellobiose accumulated to a certain concentration, could the cellobiose operon be active to reverse the inverter.

Figure 3.7 Gene circuit of cellobiose response element

Cellobiose response element is the mutant of chb operon in E.coli. Wild‐type strains of Escherichia coli are normally unable to metabolize cellobiose. The chb operon is the inducible genetic system involved in the catabolism of N,N′‐diacetylchitobiose(Plumbridge J et al.,2004). However, single base‐pair changes in the transcriptional regulator chbR that translate into single‐amino‐acid substitutions constitute the cellobiose operon which can response cellobiose(Kachroo AH et al.,2007). So, ECUST_China iGEMers performed several types of mutation: chbRN137K, chbRY30C and chbRN238S, aiming to find a efficient cellobiose response element.

Figure 3.8 The mutation sites of chbR

During the experiment, we used mRFP as the reporter. The sequence of Pcel and chbR was obtained from E.coli K12 MG1655 via PCR .Then we performed inverse PCR to achieve the site-directed mutation of amino acids of chbR. According to the literature(Kachroo AH et al.,2007), we chose three site-directed mutation : chbRN238S, chbRY30C and chbRN137K.

Figure 3.9 Gene circuit of cellobiose response element characterization

Figure 3.10 The PCR results of pCEL backbone and chbR

After constructing the mutation plasmids: pIN1-NK and pIN1-YC-NS. Both of the mutant clones and the wild-type strain were incubated in M9 medium containing 0.4% glycerol and 0.4% casamino acids, with or without cellobiose for about 40 hours. All of the samples were transferred to 96-well plate to measure the mRFP fluorescence.

Figure 3.11 Fluorescence intensity induced by cellobiose

The results revealed that presence of both chbR mutants resulted in a high basal level of expression. More importantly, the transformant carrying ChbR-YC-NS showed an approximately threefold induction over the basal level in the presence of cellobiose whereas no induction was seen in the presence of wild-type ChbR.

Linker optimization

Since the cellulose fiber cannot cross the cell membrane, we expected the cellulase, including endoglucanase(CenA) and exoglucanase(Cex), to be secreted through α-hemolysin system.

The hemolysin (type I) secretion system belongs to the ABC transporter family, which recognizes the C-terminal amino acids of hemolysin toxin HlyA for protein secretion without requiring an N-terminal signal peptide.

So we planed to fuse the cex and cenA gene with hlyA tag at the C-terminus. Besides, we also co-expressed the HlyB and HlyD, two inner membrane proteins necessary to recognize and transport fusion protein across the inner and outer membrane directly.

Figure 4.1 Gene circuit of cellulase secretion system

But before we constructed our final plasmid, which is load with cex-hlyA, cenA-hlyA, hlyB, and hlyD(Figure above), we firstly tried to replace the function gene, namely cex and cenA, with mRFP.

Figure 4.2 Gene circuit of mRFP secretion system

We used BBa_K1166002 as the template to acquire the hlyA-tag, hlyB ,and hlyD as a whole by PCR. And in the similar way, we cloned the mRFP-containing pIN2 backbone form one of our former constructed plasmid. In order to diminish the conformation effects of the hlyA on the target gene(in this case, mRFP), we constructed 3 types of linker— flexible, rigid(Xiaoying Chen et al., 2013), and OmpT-cutting linker(R. A. Kramer Ye et al., 2001)—form the original linker which was reported to interfere with the GFP folding when fused together.

Figure 4.3 3 types of linker optimization

We modified the linker by introducing desired 5’ primer sequence.Through touch-up PCR using Phanta high fidelity polymerase from Vazyme, we got all the fragments and tested their length through electrophoresis.

We purified the PCR products and detected the concentration of DNA using NanoDrop2000.

Figure 4.4 Touch-up PCR results
O:original linker, F:flexible linker, R:rigid linker, C:OmpT-cut; line1-4: pET+mRFP plasmid backbone(~3500bp); line5-8: hlyA+secretion system(~5200bp); line9: marker III
Diagram 4.1 Touch-up PCR of hlyA+secretion system

Using Gibson-Assembly method, we ligated the backbone and target gene. Ligated plasmid were transferred intoE. coliDH5α competent cell, then applied onto the Kan-containing LB plate and incubated at 37℃ for 12hr(Marc Baaden et al., 2004).

Diagram 4.2 Concentration of purified DNA measured by NanoDrop2000
Diagram 4.3 Gibson Assembly Reaction System

We selected 8 visible(purple) colonies of each four plates and used 3 pairs of primer to test whether these 2 fragments are connected.

Figure 4.5 Plates of DH5α transformants (pIN2 + mRFP + linker + hlyA + araC + pBAD + hlyB + hlyD) of 4 types of linker on LB-kan plates and in liquid M9 media.

Three different pairs of primer has been used to test the connection link. and the colony PCR results showed that positive clones were: flexible-linker F12357, rigid-linker R12367, OmpT-cleavable-linker C12357, and original-linker O4568.

So we chose F3, R2, C3, and O8 to be amplified and further tested by sequencing(BGI company). The link sequences were all proved to be in accordance with our designed linker sequence, indicating we’ve successfully constructed 4 types of plasmid with original, flexible, rigid and cleavable linker respectively. These four clones were then use to prepare protein sample to run SDS-PAGE and Western Blot.

Figure 4.6 Colony PCR results shown on 1% agarose gel.

All the colonies proved to be positive by colony PCR and first generation sequencing could be easily spotted red on plates and liquid media, directly indicating that the all the linker+HlyA tags were not interfering the emission of fluorescence of mRFP. Compared with using GFP as the secreted target gene, mRFP is apparently more suitable for the fusion of HlyA to testify the secretion efficiency of hemolysin system.

Figure 4.7 Liquid media of DH5α transformants (pIN2 + mRFP + linker + hlyA + araC + pBAD + hlyB + hlyD) of 4 types of linker on LB-kan plates and in liquid M9 media.

Positive transformants with different linker were cultured and induced by arabinose. We collected the same weight of E.coli by restricting the value(OD multiple volume) equal to 16.32. And then separate the media and cell by centrifugation. The supernate were concentrated 100 fold and the pellet were resuspended with PBS. Supernate and pellet were both pretreated to prepare protein sample and run SDS-PAGE.

Figure 4.8 Liquid LB media containing positive colonies with(Y) or without(N) arabinose induction after 16 hours. Left three media are pIN2+mRFP, and pIN2(empty vector) as negative control.

The SDS-PAGE shows that, the fusion protein—mRFP-HlyA(in the red frame, about 32.6kD)—were all expressed inside the cell and successfully secreted to the media.And in the OmpT-cleavable linker case, the mRFP and HlyA-tag were also successfully cleaved, since the cleaved mRFP(in the rgreen frame, about 26.0kD) clearly surpassed the fusion protein on the gel. Although the mRFP-HlyA fusion protein and mRFP were both slightly smaller than expected size, we highly suspected that’s because, the pI of the mRFP-HlyA and mRFP are estimated to be 5.12 and 5.07, both were lower than pH7, which might dramatically affect the mobility ratio of these protein in the SDS-PAGE buffer and shows at a lower molecular weight on the gel.

Figure 4.9 SDS-PAGE of supernatant and pellet of types of linker-transformants.

So we’ve demonstrated the OmpT-cleavable linker to be functional, and the flexible and rigid linker were also proved to be not hindering the secretion of fusion protein while offering several alternative options for perspective users who want to express their protein in an secretory form.



Shanghai, China


+86 021-64253306