Results

Wet lab result

Strain construction

There are two main strains that we constructed in our project. One of them is the pilin producing strain V.PC and the other one is the report strain V.CR-mod(fig.1 and fig.2). During the strain construction process, we tried three different transformation protocols and finally found out the method that works to our V.natriegens strain.

在这次的项目中，我们构建了两个主要的菌株。其中一个是生产导电菌毛的V.PC菌株，另一个是报告菌株V.CR-mod。在菌株构建的过程中，我们一共尝试了三种不同的转化方法，最终找到了有效的电转细胞制备方式与电转参数。

Strain V.PC has plasmid pPilin and pdCas9-sg.pPilin(fig.1a)encodes e-pili monomer gene from Geobactor metallireducens, designated as G.PilA. pdCas9-sg(fig.1b) encodes dCas9 from Streptococcus pyogenes and sgRNA that targets at the promoter of type Ⅳ pilus monomer in V.natriegens(V.PilA promoter). V.CR-mod is an accessory strain which also has two plasmids, pGFPmod2 and pdCas9-sg. pGFPmod2(fig.2a) is a report plasmid to test the efficiency of CRISPRi system. sfGFP is under the control of V.PilA promoter, which is the target of sgRNA in pdCas9-sg.

菌株V.PC转入了pPilin和pdCas9-sg两个质粒。pPilin编码金属还原地杆菌导电菌毛的组装单体，被命名为G.PilA。pdCas9-sg编码来自化脓链球菌的dCas9基因以及靶向需钠弧菌四型菌毛单体启动子（V.PilA promoter）的sgRNA。V.CR-mod是一个辅助菌株，它也就有两个质粒，分别是pGFPmod2和pdCas9-sg。pGFPmod2是一个用于测试CRISPRi系统工作效率的报告质粒。在这个报告质粒中，sfGFP的表达由V.PilA promoter控制，而V.PilA promoter正是CRISRPi质粒pdCas9-sg中sgRNA的靶向位点。

Vibrio natriegens pili suppression

CRISPRi system construction

Because Vibrio natriegens itself can produce type Ⅳ pilus, we need to suppress the endogenous production of type Ⅳ pilus before producing the e-pili. Therefore, we designed a CRISPRi system to suppress the expression of V.PilA.The reason why we didn't choose CRISPR to knock out V.PilA was that it has been reported in the literature that DNA double-stranded breaks caused by the CRISPR/Cas9 system could be extremely toxic to Vibrio spp because non-homologous end-joining (NHEJ) activity is almost undetectable in the microorganism.

因为需钠弧菌本身可以产生Ⅳ型鞭毛，所以在产生e-鞭毛之前，我们需要抑制内源性Ⅳ型鞭毛的产生。因此，我们设计了一个CRISPRi系统来抑制V.PilA的表达。我们之所以没有选择CRISPR来敲除V.PilA，是因为文献中报道了CRISPR/Cas9系统引起的DNA双链断裂对嗜盐弧菌等微生物的毒性极高，因为非同源末端连接（NHEJ）活性在该微生物中几乎无法检测到。

We initially designed four different sgRNAs targeting at the sequence near transcription start site(TSS) of V.PilA. We ordered the first version of pdCas9-sg which had sgRNA 1 and gene segments containing the other three sgRNAs. Then, using high fidelity reversed PCR and Gibson assembly, we acquired 4 different versions of pdCas9-sg plasmids(fig.3b).

我们最初设计了四种不同的 sgRNA，靶向 V.PilA 的转录起始位点（TSS）附近的序列。我们订购了第一版 pdCas9-sg，其中包含 sgRNA 1 和含有其他三个 sgRNA 的基因片段。然后，通过高保真反向 PCR 和 Gibson 组装，我们获得了 4 个不同版本的 pdCas9-sg 质粒（图 3b）。

System efficiency measurement

pGFPmod2 is the final version of our report plasmid for CRISPRi efficiency measurement.(for designing process of report plasmid, visit engineering success ) It has the gene of sfGFP under the control of V.PilA promoter and the N terminus sequence of V.PilA links to the N terminus of sfGFP. The methionines in the V.PilA sequence was mutated so the folding of sfGFP would not be influenced by an attached fragment(fig.4).

pGFPmod2是我们用于CRISPRi效率测定的报告质粒的最终版本（报告质粒的设计过程请访问engineering success页面），它具有V.PilA启动子控制的sfGFP基因，V.PilA的N末端序列与sfGFP的N末端相连。V.PilA 序列中的蛋氨酸被突变，因此 sfGFP 的折叠不会受到连接片段的影响（图 4）。

pGFPmod2 was cotransferred with pdCas-sg,pdCas-sg2,pdCas-sg3 and pdCas-sg4 seperatly into V.natriegens(fig .6). Overnight cultures of these four strains were spread out on plates and the cells were collected with culture medium for Green flourescence measurement. Based on the measurement results, it can be observed that all CRISPRi plasmids, except for pdCas9-sg1, have a certain inhibitory effect. Among them, pdCas9-sg4 shows the best inhibitory effect. However, due to time constraints, we only used the initial pdCas9-sg1 plasmid to construct the bacterial strain for pilus expression.

将 pGFPmod2 与 pdCas-sg、pdCas-sg2、pdCas-sg3 和 pdCas-sg4 分别共转到需钠弧菌菌株中（图 6）。这四个菌株的过夜培养物被涂布在平板上，并用培养基收集细胞进行绿色荧光测量。从测量结果可以看出，除了pdCas9-sg1之外，其余CRISPRi质粒均有一定抑制效果，其中pdCas9-sg4的抑制效果最佳。但由于时间限制，我们仅使用了最初的pdCas9sg-1质粒进行菌毛表达菌株的构建。

e-Pili production

Pili harvest

Overnight culture of e-Pili producing strain V.PC was spread out on plates for pili harvest. Meanwhile, we also spread overnight wild type V.natriegens culture on plates.It could be seen that the apprearance of lawn of these two strains were different.The lawn of V.PC, which produced e-Pili of G.metallireducens, had winkled surface(fig.7a), while the lawn of wild type V.natriegens was smooth(fig.7b).

将产生导电菌毛的菌株 V.PC 的过夜培养物铺展在平板上进行导电菌毛收获。从图 7a 可以看出，产生金属还原地杆菌导电菌毛的 V.PC 菌株的平板表面呈褶皱状，而野生型 需钠弧菌 菌株的平板表面则很光滑（图 7b）

The whole procedure of e-pili harvest included cell collection and resuspension, shearing pili from cells, removal of cell debris and precipitation of e-pili.The harvested e-pili solution was on average 2~3 μg/μL.

收获导电菌毛的整个过程包括细胞收集和重悬浮、剪切细胞中的导电菌毛、清除细胞碎片和沉淀导电菌毛。

The harvested pili solution was purified using Ni-NTA resin. To roughly test the conductivity of this pili, we used gold electrode and multimeter to measure the resistance of the unpurified e-pili solution and purified solution. Pili solution was dropped between two gold electrodes and waited about 10 min for the solution to air-dry. After air-drying, the pili solution became a protein film. Resistance was measured when the protein film was still moisted. The vedios below illustrate different resistance of purified or unpurified e-pili solution. It could be seen that the resistance of unpurified pili solution had a resistance of about 24.3kΩ while purified pili solution had a lower resistance of about 17kΩ.

We also measured the resistance of ddH2O we used to solve the pili and NaCl solution(5%) as control.The resistance of ddH2O was about 1.87MΩ and that of the NaCl solution was about 57 kΩ.We also performed the same test on BSA standard solution(adjusted to the same concentration as purified pili solution) and the result was about 220 kΩ.

为了粗略测试导电菌毛的电导率，我们使用金电极和万用表测量了未经纯化的导电菌毛溶液和纯化溶液的电阻。将导电菌毛溶液滴在两个金电极之间，等待约 10 分钟使溶液风干。风干 20 分钟后，导电菌毛溶液变成一层蛋白质膜。下面的图片展示了纯化和未纯化导电菌毛溶液的不同电阻。可以看出，未经纯化的导电菌毛溶液的电阻约为 24.3kΩ，而纯化的导电菌毛溶液的电阻较低，约为 17kΩ。

Then, we used TEM (transmission electron microscope) to view the micro-structure of the pilis. In fig.9, several nanowires could be seen after negative staining (fig.9) However, there were many impurities in our sample as shown by the TEM result. We speculated that these impurities may be caused by insufficient speed provided by our centrifuge and unskilled operation of purification using Ni-NTA resin. Apart from that, there were much fewer pili in our sample compared with samples harvested in previous research[1]. We inferred that the expressing efficiency or assembly ability of G.PilA in V.natriegens was not enough. It could be part of the reason why we failed to detect G.PilA in Western Blot that the concentration of His-tagged pili was lower than the detection threshold.

然后，我们使用 TEM（透射电子显微镜）观察菌毛的微观结构。从图 6 中可以看到，经过负染色后，可以看到几条纳米线（图 6）。然而，从 TEM 的结果来看，我们的样品中有很多杂质。我们推测，这些杂质可能是由于我们的离心机速度不够快，以及使用 Ni-NTA 树脂纯化时操作不熟练造成的。此外，与之前的研究[1]相比，我们的样本中的菌毛数量要少得多。我们推断，G.PilA 在 需钠弧菌 中的表达效率或组装能力不够。这可能是我们未能在 Western blotting中检测到 G.PilA 的部分原因，因为 His 标记的菌毛浓度低于检测阈值。

Model result

Optimization Results

We can observe that adding new aromatic amino acids can significantly optimize both our objective functions one and two.

可以看到，添加新的芳香族氨基酸可以显著优化目标函数一和目标函数二。

In objective function one , we found that replacing serine with aromatic amino acids at position 2 of the original amino acid sequence of the e-pili can reduce the average distance in objective one from 4.72 Å to 4.63Å. To verify the accuracy of the results, we attempted to iterate through all possible spacing scenarios for adding 1, 2, and 3 pi-residues (Fig. 2) and found that only the modification at position 2 changes the original shortest electron transfer pathway, and adding more aromatic amino acids does not alter this result.（Fig.9）

在目标函数一中，我们发现在 e-pili 原始氨基酸序列的第 2 位用芳香族氨基酸替换丝氨酸可以将目标函数一中的平均距离从 4.72 Å 缩小到 4.63 Å。 为了验证结果的准确性，我们尝试遍历了添加 1、2 和 3 个 pi残基的所有可能的间距方案（图 2），结果发现只有第 2 位的修饰改变了原来的最短电子传递途径，而添加更多的芳香族氨基酸并不会改变这一结果。（图9）

In objective function two , because the objective function we chose is used to describe the distribution density of aromatic rings within the e-pili, this means that adding 1 or 2 pi-residues cannot yield an optimal result. Therefore, we started considering the addition of three Π-residues, as shown in Fig. 1. Adding 5 aromatic amino acids can reduce the original spacing from 7.6 Å to less than 5.5 Å.

在目标函数二中，由于我们选择的目标函数用于描述 e-pili 中芳香环的分布密度，这意味着添加 1 个或 2 个 pi残基无法获得最佳结果。因此，我们开始考虑添加三个Π-残基，如图 1 所示。添加 5 个芳香族氨基酸可以将原来的间距从 7.6 Å 缩小到 5.5 Å 以下。

Model Evaluation

Based on the optimization results from the model described above, we reconstructed the structural models for single point mutants T2F and triple point mutants A31F, Q53F, and N59F to validate our conclusions:

根据上述模型的优化结果，我们重建了单点突变体 T2F 和三点突变体 A31F、Q53F 和 N59F 的结构模型，以验证我们的结论：

In optimization objective one , we replaced the threonine at position 2 of the wild-type metal-reducing rod bacterium pilin amino acid sequence with phenylalanine. We only selected the shortest electron transfer path in the pilus structure obtained by the Dijkstra algorithm (where the original aromatic rings are marked in red, and the newly added aromatic rings in mutant T2F are marked in yellow).

在优化目标一中，我们用苯丙氨酸替换了野生型金属还原地杆菌菌毛氨基酸序列中第 2 位的苏氨酸。我们只选择了通过 Dijkstra 算法得到的柔毛菌结构中最短的电子传递路径（原芳香环用红色标出，突变体 T2F 中新添加的芳香环用黄色标出）。

We found that the newly added aromatic amino acids changed the original electron transfer chain to (F1p, Y27p+4, F24p+4) (electrons are transferred from position 1 of the first monomer to positions 27 and 24 of the fifth monomer, and finally to position 1 of the second monomer to complete the cycle), where the centroid distances between adjacent aromatic rings are (4.6 Å, 5.5 Å, 4.1 Å). In the mutant, according to our model prediction, the electron transfer chain would change to (F2p, Y27p+4, F24p+4). However, we can clearly observe that the centroid distance between the aromatic rings in the electron transfer chain of this mutant is significantly larger than that of the wild type, with distances of (7.1 Å, 5.5 Å, 9.5 Å). （Fig11）

我们发现，新添加的芳香族氨基酸将原来的电子传递链改变为（F1p, Y27p+4, F24p+4）（电子从第一个单体的第 1 位转移到第五个单体的第 27 位和第 24 位，最后转移到第二个单体的第 1 位，完成循环），其中相邻芳香环之间的中心距为（4.6 Å, 5.5 Å, 4.1 Å）。在突变体中，根据我们的模型预测，电子传递链将变为（F2p、Y27p+4、F24p+4）。然而，我们可以清楚地观察到，该突变体电子传递链中芳香环之间的中心距明显大于野生型，分别为（7.1 Å, 5.5 Å, 9.5 Å）。（图11）

This situation arose because, when determining the constraint conditions of the decision variables in the optimization model, to simplify the model, we used the average coordinates of all carbon atoms on the side chain of the modified position to approximately predict the centroid coordinates of the newly added aromatic ring. The error caused by simplifying the model ultimately led to incorrect optimization results.

This model illustrates that objective function one can no longer be optimized by adding new aromatic amino acids to the original amino acid sequence. Subsequent teams may try to optimize by changing the type of aromatic amino acids to affect the overall hydrophilicity and hydrophobicity of the pilus.

This model illustrates that objective function one can no longer be optimized by adding new aromatic amino acids to the original amino acid sequence. Subsequent teams may try to optimize by changing the type of aromatic amino acids to affect the overall hydrophilicity and hydrophobicity of the pilus.

该模型说明，在原始氨基酸序列中添加新的芳香族氨基酸已无法优化目标函数一。后续团队可能会尝试通过改变芳香族氨基酸的类型来优化，从而影响菌毛的整体亲水性和疏水性。

In objective function two, we added 3, 4, and 5 phenylalanine residues at the positions calculated by our optimization algorithm, respectively, on top of the original structure. We can intuitively observe that the introduction of aromatic amino acids significantly improves the uniform distribution of aromatic amino acids in the original pili structure, causing the core of electron transfer to spread outward from the pili's central axis. This means that when the structure of conductive pili changes under the influence of different environmental factors, there are more aromatic rings involved in the electron transfer process, potentially leading to an increase in the overall electrical conductivity of the pili.

在目标函数二中，我们在原始结构的基础上，在优化算法计算出的位置上分别添加了 3、4 和 5 个苯丙氨酸残基。我们可以直观地观察到，芳香族氨基酸的引入明显改善了原始菌毛结构中芳香族氨基酸的均匀分布，使电子传递的核心从菌毛的中轴向外扩散。这意味着当导电菌毛的结构在不同环境因素的影响下发生变化时，有更多的芳香环参与电子传递过程，从而有可能导致菌毛整体导电率的提高。

Hardware result

In our design, since our system needs to detect samples with possible pathogenic hazards. In order to maximize the protection of the operator of this hardware, we broke down our hardware into two isolated parts, a test disposable and a multi-use measurement and data processing device.

在我们的设计中，由于我们的系统需要检测具有可能的病原危害性的样本。为了最大程度的保护这台硬件的操作者，我们将我们的硬件分解成一次性检测耗材和多次使用的测量和数据处理装置这两个之间相互隔离的部分。

For the hardware part we prototyped the quantitative rapid test device Plinker and the test disposable E-Pili Cartridge. We designed the PCB for the test module and debugged and packaged the WiFi module and screen. We achieved successful connectivity between the two parts by designing and iterating on the connection method, and the Plinker was able to measure the conductivity of the E-Pili in the Cartridge.

硬件部分我们制作了定量快速检测设备Plinker和检测耗材E-Pili Cartridge的原型。我们设计了检测模块的PCB，并进行了WiFi模块、屏幕的调试和封装。我们通过设计和迭代连接方式，实现了两部分之间成功的连通，Plinker可以测量出Cartridge中菌毛的电导率。

To demonstrate the relative accuracy of our system, we also designed and did a simulation experiment to test the results of Plinker's conductivity measurements using a saline solution and a buffer. The results did prove the accuracy of our system.

为了证明我们的系统相对精确，我们还设计并进行了一个模拟实验，使用盐溶液和缓冲液测试Plinker测量电导率的结果。结果的确证明了我们系统的精确性。

E-Pili Cartridge

The E-Pili Cartridge needed to provide a platform that would allow full binding of the modified E-Pili and the disease marker to be tested. After several iterations we finally settled on using lateral flow immunochromatography, the most common structure used in rapid screening kits, as the substrate for our disposables. We immobilize the purified E-Pili in the test region on the NC membrane, and the sample solution added on the other side will flow across the NC membrane under the moisture gradient provided by the absorbent pad, and eventually bind to the E-Pili in the test region. We placed a interdigital electrode in the upper layer of the E-Pili, customized a new casing to accommodate the electrode, and designed its connection to the Plinker for measurement purposes.

E-Pili Cartridge需要提供一个让修饰后的菌毛和待测疾病标志物充分结合的平台。经过了多次的迭代后我们最终确定了使用快速筛查试剂盒最常用的结构---侧流免疫层析作为我们的耗材的基底。我们将纯化后的导电菌毛固定在NC膜上的待测区域中，在另一侧加入的待测样品溶液会在吸水垫提供的水分梯度的作用下在NC膜上流动，最终在待测区域和菌毛结合。我们在菌毛上层置入叉指电极，定做了新的可以容纳电极的外壳，并且设计了其与Plinker的连接方式以达成测量的目的。

Plinker

The accuracy of quantification is directly proportional to the accuracy of the conductivity measuring instrument. In order to get better quantitative results, we designed a high accuracy conductivity measurement system based on the STC15 development board. We made it as a PCB for detecting the resistance value and analyzing it.

定量的准确程度和电导率测量仪器的精度成正比。为了获得更加好的定量结果，我们设计了一块基于STC15开发板的高精度电导率测量系统。我们将其制作为PCB，用于检测电阻值和分析。

In addition, the conductivity of the E-Pili may be affected by the ambient temperature and humidity. As a system with high requirements for quantitative accuracy, our hardware must be able to realize the quality control of the data it produces and the automatic correction within a certain error range. For this purpose, we have designed a Blank program and a quality control step in this module.

除此之外，菌毛的电导率可能会受到环境温度和湿度的影响，作为一个对定量的精度具有较高要求的系统，我们的硬件必须要能够实现对于其所产生的数据的质量控制和在一定误差范围内的自动修正。为此我们在此模块中设计还设计了Blank程序和质检步骤。

Finally, we used 3D printing and laser cutting to create a casing for the Plinker that seamlessly connects to the Cartridge, and embedded electrodes to connect the two parts.

最后我们用3D打印和激光切割技术为Plinker制作了一个可以与Cartridge无缝连接的外壳，并在其中嵌入用于连接的电极，以连通这两部分。

WiFI module

We download a TCP client software on the mobile device and input the IP address and port number of the ESP8266 module for connection testing. This is done to ensure that communication is working correctly. This process verifies the stability of data transmission and reliability of communication for the WiFi module.

Simulation Experiment

We added Tris-HCl buffer to the sample hole and again added Tris-HCl buffer with an additional 5% KCL at the sample hole when the conductivity indication stabilized. The conductivities of these two solutions were measured individually (by adding these two solutions dropwise directly to the interdigital electrode) as 3.134 mS and 3.845 mS, respectively. Finally, we can get the following results:

We can observe that after the addition of the Tris-HCl buffer, the buffer flowed laterally across the NC membrane, and eventually the number stabilized at 2.712 mS (after about 1 minute). After the number stabilized, we added Tris-HCl buffer with an additional 5 percent KCl, and the number started to change after about 15 seconds, and finally the number stabilized at 3.533 mS (about 1 minute).

Although there is a certain error (0.5mS) between the conductivity value measured after chromatography and that measured by directly adding the solution dropwise to the interdigital electrode, the difference between the conductivities of different ion concentrations measured by the two methods is relatively close (0.711mS vs. 0.821mS). This result can be good proof that our system is relatively accurate.

Proof of concept

We tested the conductance of pilin solution after adding different and concentration of anti-His antibody. As shown in this figure, when adding serial concentrations of appropriate antibody into the e-pili solution, the resistance elevated gradually between zero and 150 micrograms per mililiter, then the curve reached a plateau after that. In contrast, pili solution of wild type Vibro natriegens harvested with the same method didn't show this characteristic.