Team:LINKS China/Engineering


Click to add water




Type IV Plus (T4P) is a type of pili presents in various bacteria and archaea membranes. Type IV pili are secreted by these microbes for decoration of their surface and enabling adhesion to various substrates. Electric conductive pili (e-pili) is a T4P that is capable of conducting electricity and having long-range electron exchange between itself and the external humid environment for generation electric potential energy[1].

To produce e-pili on large scale as both the source and material of the clean energy, LINKS_China aimed to transform Escherichia Coli (E.coli) into a stable chassis of e-pili production and build a stable and high power density humidity power generation cell. While the production of functional conductive e-pili with K-12 E. coli has been substantiated, the pili’s performance in a humidity power generation device is still unclear. Therefore, we initially engineered a strain of E. coli BL21 to express e-pili by inserting a plasmid of pili generating system containing EHEC T4P assembly machinery called pili generator and Gs pilA, the major pilin monomer gene of e-pili from Geobacter. Sulfurreducens. The pili generating system includes 13 components: pilA, the Gs pilin, and pili generator gene cluster which constitutes the rest 12 components: hofBC, hofMNOPQ, ppdAB-ygdB-ppdC, and gspO(fig. 1a, b).

Our project constitutes four main stages: 1) Production of Gs e-pili in Escherichia Coli; 2) Expression of different types of e-pilis and their functional analysis; 3) Pili Yield Analysis, and 4) Enlargement of pili yield. In these stages, we constructed and expressed three types of e-pili, designed a new measurement for quantitative analysis of our pili production, and carried out several optimizations on genes of pilA and pili generator to enlarge the yield of pili production. The e-pili character of electricity generation has been confirmed with detectable voltage and current in our battery units. Additionally, here we will give a proposal of the cultivation conditions for having the highest e-pili production after comparison of several controlled variables for cultivation, and prove the function of pili generator over-expression in pili production in E.coli BL21.


1. Production of e-pili in Escherichia Coli

1.1 Major Pili Plasmid Construction

To acquire a complete pili generating system, several steps were included in the construction. We first ordered oligo DNAs from Gs pilA (BBa_K3552000) gene and assembled them by Splicing Overlap Extension PCR to obtain its complete gene sequence(fig. 2a). We amplified each gene cluster of pili generator (hofBC, hofMNOPQ, ppdAB-ygdB-ppdC, and gspO) from E.coli DH5α and assembled all genes we obtained separately on pSB1C3 vector by Gibson Assembly to form three sub-plasmids: P2: Gs pilA and hofBC (BBa_K3552000 and BBa_K3552003); P3: hofMNOPQ (BBa_K3552004); and P4: ppdAB-ygdB-ppdC and gspO (BBa_K3552005) (fig. 2b). The electrophoresis photography shows all genes in the correct molecular mass, confirming the successful synthesis of segments for the major plasmid(fig. 2c). Additionally, we constructed another sub-plasmid and another sub-plasmid P1: plasmid containing the pET28a vector for the backbone of our major plasmid. Finally, four sub-plasmids were assembled through Golden Gate Assembly to form the final major plasmid of Gs pilA with pili generator from E.coli DH5α (BBa_K3552009) for pili production(fig. 2e, d).


Since using the pili generator from the strain Enterohaemorrhagic Escherichia Coli (EHEC) will give a better production of T4P[3], we conducted point mutation separately on individual genes in the three sub-plasmids we constructed previously (P2, P3, and P4) for further genetic construction, deriving the original gene sequence of pili generator from E.coli DH5α into that of E.coli EDL933 and obtaining three new sub-plasmids: P5(BBa_K3552006), P6(BBa_K3552007), and P7(BBa_K3552008) (fig. 3a, b). We then performed the same assembling method as before to assemble three new sub-plasmids with P1, obtaining a new major plasmid possesses Gs pilA with EHEC pili generator (BBa_K3552010) (fig. 3c, d). The result of sequencing analysis confirmed the successful mutation conducted and the new plasmid construction(fig. 3e).


1.2. Pili expression and extraction

We chose E.coli BL21 for our pili expression because it is the most commonly used engineering strain and it was proved that it can be used for T4P production. On the cultivation part, we cultured the bacteria in solid M9 mediums for its relatively clean content background and infertility, and the bacteria can be able to attach on the surface of the solid medium, which can stimulate the pili growth on bacteria membrane. Additionally, we provide the bacteria with glycerol as the carbon source because it was confirmed to help improve the conductivity of the pili produced.

We transformed the plasmid (BBa_K3552010) into E.coli BL21. The bacteria was first cultivated on an LB medium plate with kanamycin at 30°C, 24h, then scraped off the LB plates with liquid M9 solution and coated the collected bacteria solution on M9 medium plates with IPTG and kanamycin. Finally, the coated M9 plates were cultured at 30°C, 48h. We harvested the bacteria from M9 plates after the cultivation by scrapping them off again with liquid M9 solution and collect the bacteria solution. By conducting the extraction and purification of the samples, we harvested an amount of Gs pili solution in 150mM ethanol-amine after filtrating with a 100kDa membrane in nitrogen gas(fig. 4a, b, c, d).

For physical confirmation of e-pili produced, we ordered SEM Photography for our pili. In the photographs, we can see that the pili were distributed evenly on the surface in porous structures between each pili individuals. The pili seemed to have a structure that differs from the common structure of proteins(fig. 4e, f). This porous structure of the pili membrane is like a sponge that can assure a water potential gradient would form within the pili membrane, which is crucial and beneficial for e-pili to generate electricity.


Although we have discovered the porous structure of the pili membrane shown by the SEM result, it was still hard to affirm that the nanowire in the photographies is our target e-pili. In order to confirm that the E.coli did produce the Gs pilA nanowire, we ameliorated the sub-plasmid P5 by adding 6xHis-tags to the end of the Gs pilA gene sequence(fig. 5a, b, c). It is proved that the addition of His-tags will nearly not affect the conductivity and electricity generation of the e-pili produced. Then, we conducted Western blot experiments, using the His-tag antibody attaching the His-tags on pili and the secondary antibodies sticking to the primary antibodies for coloration. According to the positive strands on the membrane, pilA was confirmed to be produced in the whole-cell and existed in our purified material, with an expected molecular mass of 10kDa(fig. 6a).


2. Expression of different types of e-pilis and their functional analysis

After we have successfully produced the Gs pilA, we started to seek for different types of pilA presenting on bacteria. Two new pilins, Pa pilA and Gm pilA, found in other types of Geobacters, Pseudomonas Aeruginosa and Geobacter metallireducens, were included in our project, being successfully synthesized as well by using SOE PCR [2,4]. We then constructed two new plasmids of the Gm pilA and Pa pilA pili production system. Both Pa and Gm pili were also proven to be produced successfully in E.coli BL21 with an expected molecular mass of 10kDa, according to the Western Blot results(fig. 6b, c).


In order to analyze their function of electricity generation, we applied the extracted pili to make our battery unit. A square of graphene electrode was painted on the glass slide and dried on 80°C hot plate, being as the bottom electrode of the battery. Then, 25ul pili solution was dropped on the graphene electrode for seven times, each time after the former droplets dried, forming a three layers pili membrane on the electrode to construct a pili battery(fig. 7a, b).

We made batteries in three types of pili according to the method. At the same time, we designed an additional plasmid of Gs pili with DH5α generator(BBa_K3552009) and harvested the pili produced which also has the feature of e-pili, applied it into batteries as the others. By measuring all batteries in direct contact, we found that each pili battery has relatively stable voltages between 0.3V to 0.4V and that Pa pili have the highest current about 3500 nA among the four sample groups. In comparison, there are no prominent differences in voltage and current between pili generated by EHEC and DH5α pili generators(fig. 7c). What's more, as the electricity flow in the e-pili film is generated along the water gradient which could be affected by environmental factors such as humidity, therefore we carried out experiments to detect the performance of Gs pili battery in different humid environments. The data illustrates that in 60% humidity the voltage reached the highest point of 0.44V(fig. 7d). These experiments indicated the potential ability of battery application of all the electric nanowires that we produced. For further information about our battery please visit our Proof of Concept page.


3. Pili Yield Analysis?

To enlarge our pili production, we first need to have a clear definition of the productivity of pili. We first determined the e-pili yields by measuring the concentration of purified e-pili solutions with the BCA Protein Assay Kit. But due to the long periods of pili purification and extraction, normally one or two days, there would be a huge pili loss during these experiments, a great error that results in a deviation to the actual amount of pili produced. Therefore, using this method to define the e-pili yields as the final results of the concentration of pili solution is absolutely biased, unclear, imprecise, and low-efficient.

In order to better define the productivity of the e-pili, we established a new measurement for a quicker, clearer, and more accurate determination of pili production within a few hours. We use his-tags antibodies to attach to the his-tags on the pili, and then the secondary antibodies will be attached to the his-tags antibodies for coloration directly on the outer membrane of the bacteria(fig. 8a). This assay will use a microplate reader to measure the light absorbance and provide a precise quantitative analysis for pili yield by directly measuring the actual pili production on the surface of E.coli, largely eliminating the wastage during the measurement. Further detail of this measurement can be found on the Measurement page.

In this analysis, we did two repeated experiments with all types of pili under the same environmental conditions for cultivation except the variables: cultivation time and oxygen level. To further prove that the use of the E.coli EDL933 pili generator could result in a better yield of pili, we set the comparing group of producing Gs pili by DH5α generator, and compare the pili yield of two types of pilis.


i. Oxygen Level

Since the e-pili are found on Geobacters that are anaerobic bacteria, we discovered whether a limited mobile oxygen presence to the E.coli BL21 will give a rise in the pili yield. The E.coli is a type of amphimicrobe and therefore we suspected that an anaerobic environment could increase the pili production. We chose Gs and Pa pilA for sealed cultivation to limit the oxygen movement outside in the plates, creating a semi-hypoxic environment, comparing to the unlimited gas exchanges in the unsealed plates. According to pili staining, the color in the sealed sample did appear darker than the unsealed and we explained the color shade through the measurement optimization to get the relative data of pili yield(fig. 8b, c). In order to further discuss the effects of oxygen levels, we included all four types of pili in the following two experiments. The results from these experiments were surprisingly similar: sealed plates had a higher and more stable pili production than unsealed plates as we expected(fig. 9a, b). This proved that culturing E.coli in a hypoxic environment could result in a high pili production.


ii. Cultivation time

In the first experiment, we set two groups of time variables: 24h vs 48h cultivation time. We took two plates for each sample group for every 24h and the results show that all four samples had a better yield of pili after 48h cultivation comparing to 24h(fig. 9a).

Furthermore, In the second experiment, we did a repeated experiment with an additional group of 72h cultivation time, with a similar result as the first experiment, showing the same trend of better yield in 48h and that 24h cultured plates essentially have no pili produced(fig. 9b). The 72h cultivation might not have the highest yield and parts of the sample might have high pili production after a 48h period. For Gs and Pa pilA, 48h cultivation will be better than 72h in pili production while the trend of Gm pilA and Gs pilA with DH5α generator goes the opposite. But these results would have a certain degree of random uncertainty.

iii. Additional variables

Different pilins would result in a different rate of production of e-pili and we ranked the pili yield of each type of pili-Gs pili, Pa pili, and Gm pili-according to the data obtained from the quantitative analysis in two experiments. As expected, Gs pilA were identified to have the best yield among the three types of pilAs. The ranking of productivity of three pilA is Gs pilA, Gm pilA, and Pa pilA. Additionally, from the data in the graph, we can see that the only group of Pa pilA samples has the largest variables: Pa pilA has the weakest, yet the lowest yield production(fig. 9a, b).

EHEC are the strains that routinely produce Type IV Pili on their outer membrane while E.coli DH5α substantially does not.[3] This difference could result in the yield of pili production that differs from each other in two sets of pili generators. In both experiments, we compared the actual difference of EHEC and DH5α pili generator on pili production and we could find similar patterns that the pili generator from EHEC does perform better than that from DH5α by having a higher and stable pili yield(fig. 9a, b).

We suspected that the endogenous pili generator from?E.coli?could also have the function to generate our target e-pili. To confirm our hypothesis, we designed an additional experiment for comparison of the presence and absence of pili generator in this process. Two new plasmids were constructed and expressed with only Gs PilA gene and Pa PilA gene on pET28a vector respectively. The results of both Gs and Pa pilA without generator illustrated that e-pili can be produced without the presence of EHEC pili generator occasionally(fig. 9c). It is proven that nanowires can be produced by endogenous pili generators, yet the pili yield without a generator is highly unstable due to the large error bars in two sample sets. Comparing to the data with generators where they will be over-expressed, the production of e-pili will be stabilized and, for Gs pilA, have a higher yield. As for Pa pilA, although the pili yield would not originally be as high as that of Gs pilA, the expression with the generator still follows the trend(fig. 9d).

In conclusion, according to each part of the analysis, we summarized these points for cultivation that could harvest the most pili production: (1) sealed plate with limited oxygen level; (2) cultivation time between 48h and 72h; and (3) pili expression with EHEC pili generator.

4. Enlargement of pili yield

One of the possibilities that limit the pili yield could be the gene cluster of pili generators overlapping each other and sharing the same sequence. Furthermore, it also could be extremely inconvenient for us to modify the gene sequence such as RBSs (ribosomal binding sites) and conduct codon optimization for biosynthetic engineering purposes. But if the clusters were decoupled directly, the same genetic sequences will be recombined at once. Hence, we developed a new strategy to further enlarge our pili yield: decouple the pili generator gene clusters with optimization of RBS and the former region of the overlap by amplifying the former sequence for undergoing point mutation to reduce the repetition as the downstream gene, ensuring the sequence of amino acids unchanged and also remarkably reducing the endogenous recombination at the same time(fig. 11a, b).?Due to the time limit, we did not finish this stage of optimization, but we are planning to complete this genetic modification in the future.

Additionally, we will introduce the design of the ribozyme addition into each decoupled generator gene sequences. As we wished to treat ribozymes as an insulator to rise the translation rate of the genes, therefore we characterized the three ribozymes and added them onto the ptac promoter respectively to increase the sensitivity of it. For further information about the addition of ribozymes please visit our improvement website.



So far, we have successfully constructed and produced three types of nanowires (Gs, Pa, and Gm pili) production system with the EHEC generator. We discovered the full physical properties and characters of electricity generation by several evaluations several. With quantitative comparisons of different variables in pili cultivation through our new measurement, we found that cultivating E.coli BL21 with target plasmid in a sealed hypoxic environment and between 48h and 72h will give the highest and stable pili yield. What's more, we have conduct codon optimization and decoupling modification to the pili generator gene cluster to increase pili yield. Concluding of all experiments above, we are able to confirm that we have developed E.coli to be the suitable chassis of electrical Type4 pili production.


Continuing the further enlargement of the pili production plan, we will conduct RBS optimization for the upstream of each gene of the decoupled pili generator gene cluster and serve these alternations as the templates in the next stage. After consulting our tutor, we will next use fluorogenic quantitative PCR to confirm the concentration relationship of mRNA of each gene and to establish mathematic models, then analyzing which genes are the most important to the cell and adjusting their RBS after the analysis. Or we will try to introduce random RBS library to cells and filtrate them by the character and production of pili grown on their surface, and therefore to pick out the best combination. As we mentioned before, ribozymes would be added for a predicted increase of translation rate as the other part of this plan. We would finally assembled all decoupled pili generator together with experiments to test the optimized function in the future.

We have identified the native pilin gene in E.coli coded for the autogenous pilins fimA and PpdD, which has been confirmed that they are non-conductive, that they could be assembled with our targeted pilin pilA together as pili subunits to form e-pili and its effects of rising the internal resistance of the e-pili. Although whether these genes will be expressed during their normal vital movement is unknown or these Gennes might not express at all, we would still plan to conduct the genetic knock out experiment on these two genes for obtaining a relatively clean background and eliminating any possible disturbance that other pilins might affect the quality of our target pili produced. We plan to use CRISPR Cas9 technology to knock out the genes and replace them with a shortened but similar sequence. We speculate that the knockout of these genes could finally result in a decrease in the internal resistance of the final e-pili products and improve the overall electricity generation and conductive function of our battery.

LINKS_China constructed a part collection in reference to the information obtained throughout our journey of electric nanowires exploration. We expect this assemblage's contribution of helping to establish a substratum of e-pili development inside out the iGEM community and inspiration to other optimization of this intriguing field of study.


1.Liu, Xiaomeng, et al.Power Generation from Ambient Humidity Using Protein Nanowires.Nature, vol. 578, no. 7796, Feb. 2020, pp. 550?C554, 10.1038/s41586-020-2010-9.

2.Liu, Xi, et al. Biological Synthesis of High-Conductive Pili in Aerobic Bacterium Pseudomonas Aeruginosa.Applied Microbiology and Biotechnology, vol. 103, no. 3, 6 Dec. 2018, pp. 1535?C1544, 10.1007/s00253-018-9484-5. Accessed 27 Oct. 2020.

3.Luna Rico, Areli, et al. Functional Reconstitution of the Type IVa Pilus Assembly System from Enterohaemorrhagic Escherichia Coli.Molecular Microbiology, vol. 111, no. 3, 21 Jan. 2019, pp. 732?C749, 10.1111/mmi.14188. Accessed 28 Sept. 2019.

4.Tan, Yang, et al.Expressing the Geobacter Metallireducens PilA in Geobacter Sulfurreducens Yields Pili with Exceptional Conductivity.MBio, vol. 8, no. 1, 17 Jan. 2017, 10.1128/mbio.02203-16.

5.Ueki, Toshiyuki, et al.An Escherichia Coli Chassis for Production of Electrically Conductive Protein Nanowires.?ACS Synthetic Biology, vol. 9, no. 3, 3 Mar. 2020, pp. 647?C654, 10.1021/acssynbio.9b00506.