Team:NCTU Formosa/Engineering



  From the research, we know that periodontal disease was proved to be related to many chronic and cardiovascular diseases. Among the pathogenic organisms of periodontal disease, Porphyromonas gingivalis (P. gingivalis) is the most important bacteria that give rise to this bad condition in dogs. To tackle the problem at its roots, we put effort into finding the efficient way to reach the inhibition of Porphyromonas gingivalis first.

   Finally, we chose antimicrobial peptides, LL-37 play the part of bacteria-inhibition. Human cathelicidin-derived LL-37 is a 37-residue cationic, amphipathic α-helical peptide. It is an active component of mammalian innate immunity. LL-37 not only can inhibit P. gingivalis [1], can also inhibit Escherichia coli itself [2], preventing DenTeeth from overgrowing after entering the canine mouth.


  The study showed that [3], producing the LL-37 peptide by E. coli, would make a change in cell shapes and can permeabilize both cell membranes of E.coli. With scanning electron microscopy, the changes in cell changes observed on E. coli cells were evident. Therefore, we imaged that LL-37 can secrete out of the E. coli without cell lysate.


   Inhibition is the key to periodontal disease. In wet lab, we first expect that LL-37 could attain the mission. To observe the expression of DenTeeth’s biobrick, the RFP would be inserted behind the LL37. Because DenTeeth would be killed by LL-37 peptide, we added a composite part in the front of the LL-37 producing sequence, preventing engineered bacteria died before entering the mouth.

   As stated before, LL-37 can inhibit E. coli. If the DenTeeth could not exist with LL-37 in a proper proportion, the DenTeeth wouldn't attain the mission of solving the periodontal disease. Therefore, we decided to use the prediction model to predict whether DenTeeth could grow and express under the effect of LL-37. To confirm the feasibility of DenTeeth.

 growth curve of E. coli and P.gingivalis
Figure 1. The biobrick of DenTeeth(A: 37°C Induced biobrick; B: LL-37 Producing Biobrick)

  Only when the temperature is up than 37°C, the LuxR would be expressed (Fig.1 A). At this moment, if the concentration of bacteria is high, the amount of AHL would rise, binding with LuxR. The complex of AHL and LuxR will attach to the Plux (Fig.1 B), activating the expression of LL-37.
To see more details click to design page.arrow_circle_right

  For the dry part, by deriving differential equations based on our assumptions and substituting parameters from published articles, we could analyze the biological parameter of DenTeeth and predict the expression of LL-37.


Colony PCR Result

  We successfully built DenTeeth with LL-37 .

Figure 2. Cloning result of DenTeeth after cloning into E. coli BL21. (4685 b.p.)


  Through the modeling built with substituting parameters from published articles, the growth of E. coli and P. gingivalis under the effect of LL-37 would be inhibited significantly. However, the figure also show that E. coli could still live with LL-37. Simply speaking, we successfully test the feasibility of DenTeeth and find that we can improve the DenTeeth with both biobrick design and efficiency optimization model.

 growth curve of E. coli and P.gingivalis
Figure 3. The growth curve of E. coli and P.gingivalis
 growth curve of E. coli and P.gingivalis
Figure 4. The growth curve of E. coli and P.gingivalis with DenTeeth


  Because P. gingivalis was in the RG2, so we could not do the LL-37 functional test by inhibit the growth of P. gingivalis. After reading some related papers, we found that the killing rate of E. coli was similar to that of P. gingivalis [4]. Based on these data, we determined to make E. coli, DH5α with pET32A, as the bacteria killed by DenTeeth in the LL-37 functional test.

Parameters Description Values Units
kk killing rate [5] 0.04 1/μM·min
N LL-37 absorbed per dead cell [5] 0.35 μM/O.D
Table 1. Parameters of the sterilization system of LL37


Model Validation

Figure 5. A is control group, and B is experiment group

  In order to validate the efficiency optimization model usable in any environment, we use E. coli with pSB1K3, hydrogen peroxide, and glucose to simulate as P. gingivalis, feeding dental bones and eating foods. As a result of the different environments of the experiment and the dog's mouse. We modified the environment reaction function to meet requirements.

Figure 6. The Validation of Efficiency Optimization Model

  The above figure shows that the prediction value is very close to the experimental data, and the prediction is very accurate and precise by observing the R-square and RMSE.

  After validating the accuracy and precision of the model, we further compare the reward of two dental bone feeding policies. Policy 1 (control group) is feeding dental bone with a fixed time interval. Policy 2 (experimental group) is feeding dental bone with RL prediction results.

Figure 7. The comparison of Reward from Policy 1 and Policy 2

   Through calculating the reward of policy 1 and policy 2 by reward function, we can claim the reward of policy 2 is statistically higher than policy 1. Simply speaking, through the validation experiment, we successfully proved the optimization ability of this model.

Functional Test

   After finishing the design of DenTeeth, we wanted to know whether its inhibition ability can have a function, so we designed the following experiment. Our team decided to use the inhibition zone experiment to confirm that DenTeeth can inhibit other bacteria. Furthermore, we can also compare the difference of inhibition zone diameter to know the strength of inhibition intensity.

   First, we spread E. coli (DH5α) with pET32a in LB plates with AMP resistance to be our inhibition target. Second, sticking up the filter paper, we dropped the following materials on the filter paper. Finally, we placed LB plates at 37 Celsius degrees for 12 hours.

   In figure 8, there are ten different materials added to LB plates. Plate A was added with hydrogen peroxide as the positive control group. We could see that Plate A only grew bacteria near the edge. Plates B and C are E. coli (BL21) with pSB1K3 and DenTeeth, respectively. In these two plates, both of which contained K resistance. With other factors unchanged, Plate B is negative control. As we could see, our DenTeeth had an inhibitory effect. In plate D and plate E, are cell lysate of E. coli (BL21) with pSB1K3 and DenTeeth, respectively. In these two plates, both of which contain K resistance. With other factors unchanged, Plate D is negative control. As we could see, the cell lysate of DenTeeth had an inhibitory effect. According to the results of plate A to plate E, we could know whether it is a live DenTeeth or a cell lysate, both of which had the function of inhibiting bacteria.

   In plates F, G and H, are added with PBS, E. coli (BL21) with pSB1K3 + PBS and DenTeeth + PBS, respectively. To ensure PBS would not influence the inhibition results significantly, we decided to make three plates containing PBS because it would be used on cell lysis. In these three plates, we could observe that only plate H (DenTeeth + PBS) had an inhibitory effect. This result allowed us to know that PBS would not have an obvious influence on our experiment, and can know that DenTeeth still had an inhibitory effect.

   In plates I and J, are the LB broth after centrifugation from E. coli (BL21) with pSB1K3 and centrifugation from DenTeeth, respectively. With other factors fixed, we could observe that plate J had an inhibitory effect. This result allowed us to know that antimicrobial peptide LL-37 can be secreted from DenTeeth to the environment.
To see more details click to result page.arrow_circle_right

Figure 8. DenTeeth inhibition zone result in LB plates. ( The materials added in plate were in the following condition, and each total volume is 20μL. (A) 30% hydrogen peroxide, (B) E. coli (BL21) with pSB1K3, (C) DenTeeth, (D) E. coli (BL21) with pSB1K3 (Cell Lysate), (E) DenTeeth (Cell Lysate), (F) PBS, (G) E. coli (BL21) with pSB1K3 + PBS, (H) DenTeeth + PBS, (I) LB broth (Centrifugation E. coli (BL21) with pSB1K3 after liquid culturing ), (J) LB broth (Centrifugation E. coli (BL21) with pSB1K3 after liquid culturing ) )

   We repeated the DenTeeth inhibition functional test three times each. In figure 9, this bar graph indicates the three times average of inhibition zone diameters in different materials added to LB plates. This result shows that DenTeeth can secrete LL-37 to inhibit other bacteria.

Figure 9. The Bar Chart of DenTeeth inhibition Functional Test.


  After the engineering of LL-37, we could guarantee LL-37 can secrete out from DenTeeth without cell lysis and inhibit the growth of P. gingivalis. We ensured the process of DenTeeth could work smoothly and the efficiency optimization model could optimize the benefit of DenTeeth.


  1. Xue Yang, Li Niu, et al.(2020)”LL-37-Induced Autophagy Contributed to the Elimination of Live Porphyromonas gingivalis Internalized in Keratinocytes”Frontiers in Cellular and Infection Microbiology 10:561761
  2. Ja-YoungMoon, Katherine A, et al. (2006)”Expression and purification of a recombinant LL-37 from Escherichia coli”Biochimica et Biophysica Acta (BBA) - Biomembranes 1758(9):1351-8
  3. Ján Krahulec, Marcela Hyrsová et al. (2010)”High level expression and purification of antimicrobial human cathelicidin LL-37 in Escherichia coli”Appl Microbiol Biotechnol 88(1):167-75
  4. Kriebel K, Biedermann A, Kreikemeyer B, Lang H , et al(2013). “Anaerobic Co-Culture of Mesenchymal Stem Cells and Anaerobic Pathogens - A New In Vitro Model System.” PLoS ONE 8(11): e78226. doi:10.1371/journal.pone.0078226
  5. Snoussi, M., Talledo, J. P., Del Rosario, N. A., Mohammadi, S., Ha, B. Y., Košmrlj, A., & Taheri-Araghi, S., et al (2018). Heterogeneous absorption of antimicrobial peptide LL37 in Escherichia coli cells enhances population survivability. eLife, 7, e38174.
dark_mode Dark Mode
footer_bg footer_nctu footer_fox
fb ig