Overview

To develop MOVE system, we divided our experiments into three sections. Below, we outline each section along with the respective achievements.

• Membrane Vesicle

  We obtained key results suggesting that the production of membrane vesicles is regulated in response to glucose concentration.

• Surface Display

  Strong evidence was obtained demonstrating that the scaffold protein-SpyCatcher-SpyTag-functional protein complex is synthesized within Escherichia coli (E. coli) and displayed on the membrane.

• shRNA

  We successfully engineered an E. coli strain harboring a plasmid for shRNA production.

The results presented on this page demonstrate glucose concentration-dependent MV production, the surface display of proteins on the membrane, and the successful production of shRNA. Therefore, by integrating these techniques, the concept of MOVE, in which shRNA is encapsulated within MV that have surface display proteins, appears highly feasible and promising.

Membrane Vesicle

The production of membrane vesicles (MVs) via the Polymer Intracellular Accumulation-triggered system for MV Production (PIA-MVP) is significantly influenced by glucose concentration. As glucose levels increase, glycolysis is activated, leading to enhanced production of poly(3-hydroxybutyrate) (PHB) (Figure 1). This, in turn, raises the intracellular pressure, creating conditions more conducive to MV release. To determine the optimal glucose concentration for MV release, we prepared LB media with varying glucose concentrations and assessed the resulting differences in PHB production. This approach provided indirect evidence that MV release can be regulated by glucose concentration. For further details, please refer to Design .

The following specific experiments were conducted to validate the findings:

• Transformation with pGEM-PhaC_Re AB
• Cultivation of E. coli under varying glucose concentrations
• Measurement of optical density
• Fluorescence microscopy analysis
• FM4-64 staining of MVs

Transformation with pGEM-PhaC_Re AB

We transformed E. coli with the plasmid pGEM-PhaC_Re AB, which carries the genes encoding the three essential enzymes for poly(3-hydroxybutyrate) (PHB) synthesis: β-ketothiolase (PhaA), NADPH-dependent acetoacetyl-CoA reductase (PhaB), and PHB synthase (PhaC), necessary for PHB production via the PIA-MVP system. The transformation was successful, as indicated by normal bacterial growth.

Cultivation of E. coli under varying glucose concentrations

As previously mentioned, glucose concentration plays a significant role in MV production via the PIA-MVP system. To identify the optimal glucose concentration for MV production, we prepared LB media with varying glucose concentrations and cultured E. coli harboring pGEM-PhaC_Re AB for 48 hours. During the cultivation, bubbles were observed in the media (Figure 2), indicating successful culture progression, as reported in previous literature [1] .

Measurement of optical density

To confirm the normal growth of E. coli, we measured OD600 every 24 hours (Figure 3). A comparison between E. coli with and without pGEM-PhaC_Re AB revealed no significant differences in growth rates. This suggests that the cultures were proceeding under normal conditions.

Fluorescence microscopy analysis

We stained the E. coli cultures with Nile Red and observed them under a fluorescence microscope. Nile Red is a well-known reagent for staining neutral lipids, and in this experiment, it primarily stained poly(3-hydroxybutyrate) (PHB). The observations (Figure 4) revealed that E. coli containing the plasmid exhibited significantly higher fluorescence compared to E. coli without the plasmid. This indicates that the plasmid we constructed is functioning correctly, resulting in the production of PHB.

In addition, Figure 4 were further analyzed. Using ImageJ[9] , background noise was removed through thresholding, and regions emitting fluorescence were isolated. Subsequently, the “Analyze Particles” function was employed to quantify the fluorescence intensity of individual cells that exceeded the threshold. Fluorescence intensity for each sample was then corrected based on data obtained from the control (without plasmid) and normalized to the measured area. These procedures allowed for the quantification of fluorescence intensity from each image, revealing a bell-shaped curve with the highest fluorescence intensity observed at 2% glucose concentration (Figure. 5). This finding indicates that glucose concentration influences PHB accumulation.

FM4-64 staining of MVs

The culture medium of E. coli was centrifuged and filtered to remove the bacterial cells, followed by ultracentrifugation to obtain MVs as a pellet. The MVs were then washed with PBS and stained using FM4-64, a dye that labels cellular membranes. The stained MVs were excited at a wavelength of 558 nm, and the fluorescence emission was measured at 734 nm. The results is shown Figure 6.

Excluding the sample with 0% glucose, all samples exhibited higher MV production compared to the control. Furthermore, MV production increased in proportion to the glucose concentration, consistent with our initial hypothesis. However, the sample with 0% glucose appeared to show a higher MV yield than any other sample. This anomaly is likely due to contamination by membrane components from the bacterial cells during the purification process. This interpretation is supported by the observation that the pellet obtained after ultracentrifugation in the 0% glucose sample differed in appearance from those of the other samples.

The results of the MV production experiments demonstrated that our engineered Escherichia coli strain is capable of producing MVs, and that the production levels can be regulated by adjusting the glucose concentration.

Surface Display

We aimed to target fungi by presenting functional proteins on the surface of our engineered system. Specifically, we designed four functional proteins: Mgfp-5, Chitinase C, GH19 Chitinase, and Endo-1,3(4)-β-glucanase. Mgfp-5, an adhesive protein derived from the freshwater mussel Mytilus galloprovincialis [2] , adheres to a variety of surfaces, allowing MOVE to attach to plant leaves and other substrates, preventing dispersal by wind or rain. Chitinase C, an enzyme that hydrolyzes glycosidic bonds in chitin, a major component of fungal cell walls [3] , creates pores in the target cell wall, facilitating MV uptake. Similarly, GH19 Chitinase, an endo-type enzyme that also hydrolyzes chitin’s glycosidic bonds [4] , is expected to enhance the same effect as Chitinase C. Additionally, Endo-1,3(4)-β-glucanase degrades β-glucan, a key component of fungal cell walls, when displayed on the surface [5] [6] .

These functional proteins are expected to assist in the penetration of MVs into the target cells. For surface display, we utilized the SpyCatcher-SpyTag system, which allows easy surface presentation by inserting specific sequences into the proteins [7] [8] . Our designed sequences were validated by modeling to ensure proper surface display (see Model for details).

In the Wet Lab, we confirmed the expression of these surface-displayed proteins by incorporating the designed sequences into plasmids and transforming them into E. coli. The experiments conducted are listed below:

• Addition of GGA sites
• Synthesis of plasmids for surface-displayed protein production
• Synthesis of plasmids for scaffold protein production
• Construction of E. coli strains with surface-displayed proteins using the SpyCatcher-SpyTag system
• Verification of surface display through SDS-
• Western blotting using an anti-HisTag antibody

Addition of GGA Sites

To construct the plasmids, we added Golden Gate Assembly (GGA) sites (BBa_K5269032) containing BsmB I recognition and cleavage sites to both ends of the plasmid vector pBluescript II SK(-) and each DNA fragment. For the DNA fragments listed in Table 1, PCR was performed using specific primers. Similarly, PCR was performed on the DNA fragments listed in Table 2 (No. 1–9) using primers designed to match the desired ends. After gel electrophoresis (Figure 5) confirmed the presence of the expected bands, gel purification was conducted to obtain the desired fragments. The GGA site was then introduced into pBluescript II SK(-) through In-Fusion Assembly, enabling the subsequent assembly of the surface display protein genes.

Synthesis of plasmids for surface-displayed protein production

First, we synthesized plasmids carrying the genes for functional proteins. Using Golden Gate Assembly (GGA), we ligated each DNA fragment with the added GGA sites and the pBluescript II SK(-) vector, obtaining the target plasmids. GGA is an assembly method that facilitates seamless ligation of vectors and inserts using type IIs restriction enzymes. These plasmids were transformed into competent DH5α cells, followed by colony PCR to verify successful transformation. Gel electrophoresis was performed (Figure 8), and when the target bands were observed, large-scale culturing and plasmid extraction were conducted to mass-produce the surface-displayed protein plasmids. We attempted to synthesize 11 different plasmids with varying numbers and types of proteins, of which 8 were successfully synthesized. Table 4 lists the combinations of surface-displayed proteins and indicates whether the synthesis was successful.

Synthesis of plasmids for scaffold protein production

Next, we synthesized plasmids carrying the genes for scaffold proteins. The DNA fragments for INP or OmpA were used, along with the pTf16 vector. After adjusting the fragment ends, we synthesized the plasmids via In-Fusion Assembly. The genes, primers, and vector used are listed in Table 4 . For the gel electrophoresis results, refer to Figures 9 and 10 . Following the same procedure as for the surface-displayed protein plasmids, we were able to scale up production of the scaffold protein plasmids.

Construction of E. coli strains with surface-displayed proteins using the SpyCatcher-SpyTag system

We further engineered an E. coli strain harboring both plasmids, enabling the display of multiple proteins using the SpyCatcher-SpyTag system. First, we transformed the competent DH5α cells with the scaffold protein-producing plasmid. Colonies were screened by colony PCR, and those that tested positive for the presence of scaffold protein DNA were selected to generate competent cells harboring the scaffold protein plasmid. Next, these competent cells were transformed with the surface-displayed protein-producing plasmid. Colonies that were confirmed by colony PCR to contain both scaffold protein DNA and surface-displayed protein DNA were subjected to large-scale culturing, resulting in the successful production of the desired E. coli strain. For the combinations of surface-displayed proteins and the success of their synthesis, refer to Table 5.

Verification of surface display through SDS-PAGE

Using the SpyCatcher-SpyTag system, the scaffold protein expressed from pTf16 connects its N-terminal domain with the surface-displayed proteins, forming the Scaffold protein–Surface protein complex. To investigate the localization of these proteins within E. coli cells, we fractionated the cells into periplasmic, cytoplasmic, and membrane fractions. Protein presence in each fraction was analyzed by SDS-PAGE (Figure 11). The E. coli strains used for the analysis, along with the corresponding scaffold protein–surface protein combinations and their theoretical molecular weights, are summarized in Table 6.

For strains O_1 through I_2 (Lane 1-12), each strain presented a single type of protein. Corresponding bands were observed across all protein sections, confirming that the proteins were correctly expressed. Additionally, the detection of bands in the membrane fraction suggests that our surface display system is functioning properly. Furthermore, the absence of individual bands for the scaffold protein SpyCatcher and SpyTag-surface display protein implies that the SpyCatcher-SpyTag interaction is successfully occurring.

In strains O_3 through I_4 (Lane 13-24), we attempted to produce two Scaffold protein-Display protein constructs. In the cytoplasmic and membrane fractions of O_3, bands of the expected sizes for both proteins were detected as intended. However, for all other samples, only one of the two protein bands was observed. Upon closer examination of the undetected proteins, it was found that in all strains, the protein positioned downstream of the promoter was not being produced.

Based on these SDS-PAGE results, it is suggested that single protein surface display on the E. coli membrane was successfully achieved.

Western blotting using an anti-His-Tag antibody

To confirm that the bands observed in SDS-PAGE corresponded to the target proteins, we conducted Western blotting using an anti-His-Tag antibody targeting the C-terminal 6×His sequence of the display proteins (Figure 12).

Bands were observed in all fractions of O_1, the periplasmic and cytoplasmic fractions of I_1, and the membrane fraction of OmpA-Mgfp5 in O_3. Therefore, the successful surface display in strains O_1 and O_3 is further supported. The lack of corresponding bands in Western blotting for some of the bands detected in SDS-PAGE is likely due to the insufficient quantity of the produced proteins.

For the O_3 and I_3 strains, which were not detected by Western blotting, His-tag purification using magnetic beads was performed. This allowed us to compare the results with SDS-PAGE data and confirm the presence of surface-expressed proteins in the membrane fraction. Lane 1 and Lane 2 represent the purified and flow-through fractions of O_3, respectively, while Lane 3 and Lane 4 correspond to the purified and flow-through fractions of I_3 (Figure 13). These results confirm that His-Tagged proteins were successfully purified using magnetic beads. This finding further indicates that the scaffold protein–surface protein complex is localized in the membrane fraction.

Overall, the results of the surface display experiments demonstrate that the SpyCatcher-SpyTag system enables the presentation of a single target protein on the surface of E. coli cells. However, based on the results from strains O_3 through I_4, displaying more than one type of protein on a single E. coli cell presents challenges due to expression level issues, necessitating further considerations in sequence design.

Furthermore, the results for O_3 and I_3 suggest more efficient surface presentation for OmpA compared to INP. This suggests that more consideration should be given to scaffold proteins in the design.

shRNA

In the MOVE project focusing on RNA-based pesticides, RNA plays an indispensable role. Our approach involved encapsulating shRNA within membrane vesicles (MVs), which are then taken up by target organisms to induce RNA interference (RNAi), thereby functioning as a pesticide. To achieve this, we designed a plasmid that produces shRNA targeting GFP as a model system.

The following steps were undertaken:

• Synthesis of shRNA-Producing Plasmid
• Construction of E. coli Strains Harboring the shRNA-Producing Plasmid

Synthesis of shRNA-Producing Plasmid

First, we synthesized the plasmid for shRNA production. The DNA fragments included the GFP-targeted shRNA (BBa_K5269019) and the vector pBluescript II SK(-), both amplified using specific primers via PCR to adjust the fragment ends (Table 7). After confirming the presence of the expected bands through gel electrophoresis (Figure 10), the fragments were ligated using In-Fusion Assembly, successfully yielding the target plasmid.

This plasmid enables the production of shRNA, which is a critical step in our strategy to induce RNA interference (RNAi) in target organisms.

Construction of E. coli Strains Harboring the shRNA-Producing Plasmid

We transformed the shRNA-producing plasmid into competent E. coli cells (DH5α) and performed colony PCR. Gel electrophoresis (Figure 11) confirmed the presence of the expected band derived from the pBluescript II SK(-) vector. Following this, large-scale culturing was conducted, and the plasmid was extracted from the culture, enabling mass production of the shRNA-producing plasmid. This result indicates that we successfully constructed E. coli strains capable of producing shRNA.

Reverse transcription of shRNA

To confirm the expression of shRNA in Escherichia coli strains harboring shRNA-producing plasmids, reverse transcription followed by gel electrophoresis of the resulting cDNA was performed. The results confirmed that shRNA was being correctly produced.

This indicates the successful generation of E. coli strains capable of producing shRNA.

References

[1] Koh, S., Sato, M., Yamashina, K. et al. Controllable secretion of multilayer vesicles driven by microbial polymer accumulation. Sci Rep  12 , 3393 (2022).

[2] D. S. Hwang, Y. Gim, D. G. Kang, Y. K. Kim, H. J. Cha. Recombinant mussel adhesive protein Mgfp-5 as cell adhesion biomaterial. Journal of Biotechnology. 127 , (4) 727-735 (2007).

[3] Li X, Jin X, Lu X, Chu F, Shen J, Ma Y, Liu M, Zhu J. Construction and characterization of a thermostable whole-cell chitinolytic enzyme using yeast surface display. World J Microbiol Biotechnol. 2014 Oct;30(10):2577-85. doi: 10.1007/s11274-014-1681-5.

[4] GH19Chitinase ] M. Orlando, Patrick C. F. Buchholz, M. Lotti, J. PleissI. The GH19 Engineering Database: Sequence diversity, substrate scope, and evolution in glycoside hydrolase family 19. PLoS ONE 16 (10) (2021).

[5] Endo-1,3(4)-β-glucanase ] Takashi Akiyama. イネ科植物に特有のβ-グルカナーゼファミリー. 化学と生物. 1999, Volume37, No.8, P.526-531

[6] Niu C, Li X, Xu X, Bao M, Li Y, Liu C, Zheng F, Wang J, Li Q. [Research progresses in microbial 1,3-1,4-β-glucanase: protein engineering and industrial applications]. Sheng Wu Gong Cheng Xue Bao. 2019 Jul 25;35(7):1234-1246. Chinese. doi: 10.13345/j.cjb.

[7] S. Gallus, T. Peschke, M. Paulsen, T. Burgahn, C. M. Niemeyer, K. S. Rabe, Surface Display of Complex Enzymes by in Situ SpyCatcher-SpyTag Interaction. ChemBioChem 21 , 2126–2131 (2020).

[8] S. Gallus, Esther Mittmann, K. S. Rabe, A Modular System for the Rapid Comparison of Different Membrane Anchors for Surface Display on E. coli. ChemBioChem 23 , (2022).

[9]Schneider, C., Rasband, W. & Eliceiri, K. NIH Image to ImageJ: 25 years of image analysis. Nat Methods  9 , 671–675 (2012). https://doi.org/10.1038/nmeth.2089