Proof of Concept

Expand upon your Silver medal work for Proposed Implementation and develop a proof of concept for your project.

To make a proof of concept for our project, we have successfully demonstrated the mode of phage – intestinal commensal E. coli – antigen production by the following experiments:

  1. Testing the T7 phage susceptibility of fecal E. coli isolates.
  2. Measuring cell growth and GFP protein production curves for the E. coli isolates infected by T7 phage::GFP.
  3. Quantifying ovalbumin protein concentration in the E. coli isolates infected by T7 phage::OVA.
  4. Purifying the His-OVA antigen from our T7 phage::OVA induction system.

For future work and implementation, we are planning to verify OVA-specific antibody production in mice treated with T7 phage::OVA. For now, we can just provide a somewhat indirect data of Prof. Jan’s lab to support our hypothesis. In the meantime, we’re submitting an Animal Use Form under the investigation by Institutional Animal Care and Use Committee (IACUC) of University of Chung Shan Medical University. And now we’re waiting for their approval.

For safety issue, the intestinal commensal E. coli isolated from the mouse feces were all conducted in the Prof. Ming-Shiou Jan’s laboratory. The phage susceptibility test was measured with the commensal E. coli previously isolated and determined by Prof. Jan’s lab several years ago. For the E. coli isolated with the help of Prof. Jan’s research assistant, we just learned the process and the principles, as well as observed the results. We didn’t use them for further studies.

Below is our experimental flow chart and highlights of our results, future work and the imaginary implementation in the real world.


PHAGE SUSCEPTIBILITY OF INTESTINAL E. coli

Isolation of mouse intestinal E. coli

To isolate the intestinal commensal E. coli from BALB/c mice, the fresh mouse feces was collected and plated onto Eosin Methylene Blue (EMB) agar (Sigma-Aldrich) after homogenized in a sterile PBS buffer1. EMB is a differential medium, which inhibits Gram-positive bacterial growth and distinguishes E. coli from other non-fermenting Gram-negative bacteria with colors. The metallic green color of colonies was shown on EMB agar after growing at 37°C overnight, which were referred to as mouse intestinal commensal E. coli (Fig. 1).

Figure 1 | BALB/c mouse intestinal E. coli isolation. The diluent of homogenized feces in PBS was spread onto an EMB agar plate. The bacteria were grown at 37°C overnight.

Susceptibility to T7 phages

The mouse intestinal commensal E. coli previously isolated and determined by Prof. Jan’s lab were tested for their susceptibility to the wild-type T7 phages or the engineered T7 phages with GFP or OVA gene. The T7 phage susceptibility of the isolated E. coli were determined by the lysis of bacteria and OD600 values of overnight culture broth. We got an average susceptibility ratio of 11.5%±3% (Fig. 2). The standard errors may come from the test with unknown phage titer against the isolated E. coli individually. The optimized MOI is at 0.1 to 10 ratio of phage to bacteria and determined by individual interaction. Nevertheless, the data are consistent with the observation by Kasman LM as shown in Table 1 from the paper published on Virol J. in 20052, in which the wild-type T7 phages can infect commensal E. coli from mouse feces or human clinical specimens.

Figure 2 | The T7 phage susceptibility of BALB/c mouse intestinal E. coli. The 48 E. coli isolates were grown to an OD600 around 0.6 and incubated with 1l of the indicated T7 phages (i.e., wild-type, GFP and OVA) overnight. The complete lysis of bacteria was an indicator of the phage susceptibility measured at OD600 or observed by naked eyes.

Table 1 | Susceptibility of commensal E. coli strains to lysis by laboratory coliphage (percent susceptible strains) published by Kasman LM on Virol J. in 2005.

Titration

After determining the susceptibility of the mouse intestinal E. coli, we’d like to know the amplification effectiveness of the engineered T7 phage. Titration assay3 was performed by spotting serial dilutions of the T7 phage::OVA on a lawn of one of the isolated E. coli. As shown in Fig. 3, the concentration was expressed as plaque forming capability at a level of 108 PFU/ml. Taken together, the data showed that the phage susceptibility and titer are comparable to those of the wild-type T7 phage, indicating the engineered phages are as competent as the wild-type T7 phage against E. coli strains from the laboratory or a mouse intestine (Fig. 2 and 3).

Figure 3 | The titration of T7 phage:OVA against an isolated E. coli from a BALB/c mouse.


GFP PROTEIN PRODUCTION

To determine the protein production by T7 phage::GFP, the same isolated E. coli as in Fig. 3 were grown to an OD around 0.6 in the M9 minimal medium supplemented with casamino acids (0.2%), D-glucose (0.3%), vitamin B1 (1 mg/ml), MgSO4 (0.2 mM), CaCl2 (0.1 mM)4, followed by infected with the engineered phages at a MOI of 5 for 2 hr at 37°C. The data were read for OD600 and GFP at an ex/em = 483/513 nm every 10 min. Compared to wild-type T7 phage infection, the T7 phage::GFP can infect and kill the E. coli in a similar way (Fig. 4). In addition, the engineered phages are able to transfer the GFP reporter cassette into E. coli and make GFP production at the beginning of 40 min (Fig. 5), which is consistent with the time starting to lyse the bacteria (Fig. 4).

Figure 4 | The infection of the wild-type and the engineered T7 phages on a mouse intestinal commensal E. coli. The E. coli were growing to an OD around 0.6 in the wells of a 96-well plate (Falcon® 96-well Black/Clear Flat Bottom) at 37°C with 200 μl of the supplemented M9 media, followed by the phage (as indicated) infection at a MOI of 5. The values of OD600 were read every 10 min till 2 hr by a microplate reader (Synergy H1 Hybrid Multi-Mode Reader - BioTek Instruments). The growth of E. coli without phages was set as a control.

Figure 5 | The production of GFP by the engineered T7 phage::T7P-g10.RBS-His-GFP-T7T. The E. coli were growing to an OD around 0.6 in the wells of a 96-well plate (Falcon® 96-well Black/Clear Flat Bottom) at 37°C with 200μl of the supplemented M9 media, followed by the phage (as indicated) infection at a MOI of 5. The values of fluorescence intensity were read at an ex/em = 483/513 nm every 10 min till 2 hr by a microplate reader (Synergy H1 Hybrid Multi- Mode Reader - BioTek Instruments). The background levels were measured with the controls of E. coli without phages or infected with wild-type phages.

OVA PROTEIN PRODUCTION

To test the ovalbumin (OVA) protein expression by the engineered phage, the same isolated E. coli as tested for T7 phage::GFP were grown to an OD of 0.6 and infected with T7 phage::OVA at a MOI of 5 for 2 hr at 37°C. The bacterial lysates were subjected to SDS-PAGE and Coomassie blue staining. The gel clearly showed a sharp band at the predicted size of a purified chicken ovalbumin protein (~45 kDa) compared to the lysates of E. coli infected with wild-type T7 phages (Fig. 6A).

To quantify the protein concentration produced by the T7 phage::OVA, 5 x 108 cells of the isolated E. coli (OD600 = ~0.6) were infected with 2.5 x 109 phages (MOI = 5) for 2 hr at 37°C. The 50 μl of lysates along with a serial dilution of standard of purified ovalbumin proteins were subjected to an ELISA Kit for Ovalbumin (CLOUD-CLONE CORP., Product No. CEB459Ge) and performed according to the manufacturer’s instruction. The data were measured at OD450 nm and determined by regression analysis due to the assay is based on the competitive inhibition enzyme immunoassay technique. As shown in the Fig. 6B, the concentration of OVA proteins produced in the E. coli infected with T7 phage::OVA were 154±12 ng/ml (OD450 = 0.89±0.06).

Taken together, the engineered phages are able to transfer genes (i.e., GFP and OVA) into mouse intestinal E. coli and produce proteins of interest by the infected E. coli.

Figure 6 | Ovalbumin protein production by the mouse intestinal E. coli infected with T7 phages::T7-g10.RBS-His-Ova-T7T. (A) 30 ng of total phage-infected bacterial lysates were run on SDS-PAGE with a 4–15% Mini-PROTEAN® TGXTM Precast Protein Gels (Bio-Rad Laboratories, Inc.) and stained with Coomassie Brilliant Blue G-250. The PageRulerTM Prestained Protein Ladder, 10 to 180 kDa (Thermo Fisher Scientific Inc) was used as a marker. (B) 50 μl of the same lysates were also conducted in an ELISA Kit for Ovalbumin (OVA) (Cloud-Clone Corp.) along with a serial dilution of standard Ova protein provided in the kit. The values were read at OD450 and calculated by a formula made by a regression analysis.


OVA PROTEIN PURIFICATION

To achieve our goal of building up the model antigen of phage ovalbumin induction system (i.e., phage – E. coli – protein production) for the application of vaccine testing, a His tag was fused at the N terminus of ovalbumin. We’d like to know whether the His-ovalbumin can be purified for future application.

The His-ovalbumin protein from the lysates of E. coli DH5alpha infected with T7 phage::OVA was purified by Nickel column through the ÄKTA start protein purification system (Cytiva). The Fig. 7 showed the elutions contain predominantly purified His-ovalbumin proteins with a predicted molecular weight of 45kDa, although some degradation forms or impurities may leave in the elutions. Taken together, the ovalbumin protein in the E. coli infected by T7 phage:OVA can be produced and purified in our system. Therefore, the ovalbumin as a model antigen can be applied in the purified form or as phage-infected bacterial lysates.

Figure 7 | His-Ovalbumin purification from the phage-infected bacterial lysates. The lysates of E. coli DH5alpha infected with T7 phage::T7-g10.RBS-His-Ova-T7T were purified by Nickel column through the ÄKTA start protein purification system (Cytiva). The protein harvested were analyzed by SDS-PAGE and Coomassie Blue Staining using 4–15% Mini-PROTEAN® TGXTM Precast Protein Gels (Bio-Rad Laboratories, Inc.) PageRulerTM Prestained Protein Ladder was used as a marker. Lane: (1) E. coli control lysates without phage infection, (2) E. coli raw lysates with phage infection, (3) flow-through, (4) total lysates before purification, (5) wash-through, (6) Elution #8, (7) Elution #9, (8) Elution #10, (9) Elution #11.


SUPPORTING EVIDENCE FOR POTEINTIAL PHAGE VACCINE

In addition to an iGEM 2020 project of TUDelft and the preclinical trial of biotech company BiomX, an observation of the adjuvant-like effect of phage- infected E. coli lysates in Prof. Jan’s lab encouraged us to design a synthetic biology approach to create phage vaccine in our future project.

Mouse immunization with phage-infected E. coli as adjuvants

Briefly, BALB/c mice were intraperitoneally (IP) vaccinated using 0.5 µg (50 µl) of the commercially purified ovalbumin protein (Chondrex, Inc.) with 50 µl of the lab-made phage-infected intestinal commensal E. coli lysates compared to with 50 µl of the commercial well-defined ImjectTM Alum Adjuvant (ThermoFisher) in a formulation of aluminum hydroxide and magnesium hydroxide. The data in Fig. 8 showed that the combination of ovalbumin and phage-infected lysates can induce OVA-specific antibody production (12.3 ng/ml) at a similar level as induced by ovalbumin with Alum Adjuvant (14.6 ng/ml), which is consistent with results published by Conrad ML, et al5. The phenomena implied that the lysates of phage-infected bacteria have an immunostimulatory property of adjuvant, which has been confirmed in the observations by Zhu J, et. al. for phage particles6 and by Lim J, et. al. for bacterial envelope7.

Based on these above researches, we’re wondering whether phage carrying antigen gene (e.g., ovalbumin) can stimulate the specific immune responses in the animal body through infecting phage’s host (e.g., commensal E. coli) to produce antigen protein in the intestine.

Figure 8 | Ova-specific IgG antibody production induced by the purified ovalbumin protein with an Alu-based adjuvant or with a WT T7 phage-infected E. coli lysate. The IgG antibodies in the serum in the i.p. immunized BALB/c mice were detected using Mouse OVA sIgG (Ovalbumin Specific IgG) ELISA Kit (FineTest®). The vaccination procedure was indicated. The data were provided by Prof. Jan and organized and explained by us.

Animal study using phage vaccine for future work

For our iGEM project this year, it’s pity that we currently can’t use animals to verify one of our hypotheses. The form of Animal Use has been submitted to the reviewers in Institutional Animal Care and Use Committee (IACUC) of University of Chung Shan Medical University. We’re waiting the approval.


CONCLUSION

In our project, we successfully made a proof of concept by demonstrating the T7 phage can be a suitable vector to deliver genes (i.e., GFP and ovalbumin) into an intestinal commensal E. coli to produce qualified (fluorescence activities and SDS-PAGE) and quantified (ELISA assay) proteins in vitro. The phage susceptibility of the isolated E. coli test was clearly verified (plaque assay and titration). Furthermore, the His-tagged ovalbumin protein can be purified (AKTA system and SDS-PAGE) in our phage induction system for future studies. We also gave supporting evidence (Prof. Jan’s data) to strengthen our idea of phage vaccine in the mode of phage – commensal bacteria – antigen production. We are waiting Animal Use approval and can hardly wait to examine the possibility and effectiveness in the animal.


REFERENCE

  1. Kittana H, Gomes-Neto JC, Heck K, Geis AL, Segura Muñoz RR, Cody LA, Schmaltz RJ, Bindels LB, Sinha R, Hostetter JM, Benson AK, Ramer-Tait AE. Commensal Escherichia coli Strains Can Promote Intestinal Inflammation via Differential Interleukin-6 Production. Front Immunol. 2018 Oct 9;9:2318. doi: 10.3389/fimmu.2018.02318. PMID: 30356663; PMCID: PMC6189283.
  2. Kasman LM. Barriers to coliphage infection of commensal intestinal flora of laboratory mice. Virol J. 2005 Apr 15;2:34. doi: 10.1186/1743-422X-2-34. PMID: 15833115; PMCID: PMC1097760.
  3. Ács N, Gambino M, Brøndsted L. Bacteriophage Enumeration and Detection Methods. Front Microbiol. 2020 Oct 23;11:594868. doi: 10.3389/fmicb.2020.594868. PMID: 33193274; PMCID: PMC7644846.
  4. Vinay M, Franche N, Grégori G, Fantino JR, Pouillot F, Ansaldi M. Phage- Based Fluorescent Biosensor Prototypes to Specifically Detect Enteric Bacteria Such as E. coli and Salmonella enterica Typhimurium. PLoS One. 2015 Jul 17;10(7):e0131466. doi: 10.1371/journal.pone.0131466. PMID: 26186207; PMCID: PMC4506075.
  5. Conrad ML, Yildirim AO, Sonar SS, Kiliç A, Sudowe S, Lunow M, Teich R, Renz H, Garn H. Comparison of adjuvant and adjuvant-free murine experimental asthma models. Clin Exp Allergy. 2009 Aug;39(8):1246-54. doi: 10.1111/j.1365-2222.2009.03260.x. Epub 2009 May 3. PMID: 19438585; PMCID: PMC2728898.
  6. Zhu J, Jain S, Sha J, Batra H, Ananthaswamy N, Kilgore PB, Hendrix EK, Hosakote YM, Wu X, Olano JP, Kayode A, Galindo CL, Banga S, Drelich A, Tat V, Tseng CK, Chopra AK, Rao VB. A Bacteriophage-Based, Highly Efficacious, Needle- and Adjuvant-Free, Mucosal COVID-19 Vaccine. mBio. 2022 Aug 30;13(4):e0182222. doi: 10.1128/mbio.01822-22. Epub 2022 Jul 28. PMID: 35900097; PMCID: PMC9426593.
  7. Lim J, Koh VHQ, Cho SSL, Periaswamy B, Choi DPS, Vacca M, De Sessions PF, Kudela P, Lubitz W, Pastorin G, Alonso S. Harnessing the Immunomodulatory Properties of Bacterial Ghosts to Boost the Anti- mycobacterial Protective Immunity. Front Immunol. 2019 Nov 22;10:2737. doi: 10.3389/fimmu.2019.02737. PMID: 31824511; PMCID: PMC6883722.