Short Description

Peptidosomes are the new fundamental approach for generating and applying encapsulated bacteria. By the creation of spherical compartments containing a liquid environment inside, bacteria are still able to grow and fulfil a given task. The mesh-like structure of the sphere allows the selective exchange of compounds via diffusion, but holds the bacteria trapped inside. Therefore, we are able to benefit from the entrapped cells’ abilities, while still ensuring that they are not released into their surroundings. Peptidosomes can be further enhanced by incorporating magnetic or biological beads – which themselves can be functionalized with proteins – into their peptide-based fibrillary shell.

Background

Bacteria are omnipresent in biotechnology and applied projects. They can be used as hosts to produce nearly any biological compound of interest such as: drugs, vaccines, enzymes, antibiotics or even fuels and solvents. Their fast life cycle and comparable low requirements of living conditions highlight their industrial relevance. Over the last decades, the main focus to increase yields laid on extensive metabolic engineering and optimizing growth conditions.

Yet, there are more aspect which need to be considered when producing a compound of interest. First, where is the product of interest found: inside of the producing strain or will it be secreted to the surrounding media? Second, what is necessary to separate the valuable end-product from the bacteria? And maybe most important, how to assure a safe use of genetically engineered production strains? If you would like to know more on encapsulated bacteria in Peptdisomes and boosting the production of a compound of interest check out our: Signal Peptide Toolbox and Peptide Secretion sections.

To address these major biological and technical questions, the TU Dresden iGEM team presents EncaBcillus. Using Bacillus subtilis as model organism we introduce a new fundamental approach for cultivation of bacteria: the Peptidsomes. These Peptidosmes are buildup of self-assembled Fmoc dipeptide phenylalanines (Fmoc-FF) and are able to form spherical cages. The cages hold back the bacteria from the surrounding but are freely diffusible for smaller molecules. We show, that B. subtilis is able to grow, while encapsulated inside of the Pepdiosomes and demonstrate the applicability of two (fluorescence and luminescence) reports for the use with Pepdiosomes.

While evaluating this new method of bacterial immobilization, we established two applications using Peptidsomes:

• An encapsulated β-lactam Biosensor
• Communication between encapsulated bacteria and their surrounding

Design

Fmoc-FF

Nanostructures are gaining great attention, due to their endless possible applications. In general, nanostructures clusters are ordered within nanoscopic dimensions. In some cases, the building blocks of the nanostructures (which can be of organic or inorganic sources) can self-assemble. This process can be driven by non-covalent forces (e.g. hydrogen bonds, van der waals forces or aromatic interactions) that dictate the organization into supramolecular structures [1]. For our project EncaBcillus, we investigated the encapsulation of bacteria by the self-assembling building block Fmoc-FF (9-fluorenylmethoxycarbonyl diphenylalanine) (Figure 1).

While studying the mechanisms that direct the self-assembly of amyloid fibrils by short aromatic peptides, it was observed that the dipeptide diphenylalanine (FF), the core recognition motif of the Alzheimer’s β-amyloid polypeptide, self-assembled into nanotubular structures in aqueous solution [2][3]. Later, while studying the interactions in the process, the chemical group 9-fluorenylmethoxycarbonyl (Fmoc) was added at the N terminal of the dipeptide, facilitating the self-assembly into typical amyloid-like fibrils [4].

Follow-up studies discovered, that it is possible to trigger the self-assembly of the dipeptide in solution. At a pH higher than 8, the dipeptide stayed in solution, but when slowly adding hydrochloric acid to reach pH levels below 8, a clear gel was formed [5].

At alkaline pH, the molecules are negatively charged and therefore repealing each other. If the pH drops, protonation of the molecules occurs, neutralizing the negative charge and permitting the self-assembly of the dipeptide into fibers [6]. The formation of this mesh was shown to be induced by exposing a solution of Fmoc-FF with gaseous CO₂ [7].

---

Peptidosome creation

---

Bacteria encapsulated in Peptidosomes

In EncaBcillus, we wanted to demonstrate and characterise functional Peptidosome encapsulating Bacillus subtilis. This fundamental new approach of applying Fmoc-FF with bacteria, will create endless new possibilities for bacterial immobilisation.

To encapsulate the bacteria we first needed to establish a lab protocol. We created Fmoc-FF solutions in water, adjusted the pH to 10.5 and responded B. subtilis cells either directly taken from overnight cultures or day cultures.

To monitor bacterial growth in Peptidosomes, day cultures were grown till an OD₆₀₀ of 0.02. Cells were pelleted by centrifugation for 5 min at 16000g. The supernatant was discarded and the pellet resuspended in Fmoc-FF solution. Droplets of 15 μl of this solution were placed on an ultrahydrophobic membrane and exposed to CO₂.

We performed multiple assays to prove the encapsulating and growth inside the Peptidsosomes. Detailed protocols can be found below. Adjustments regarding bacterial used concentrations are stated in the specific experiments.

• Plate Reader Well Scan
• Stereo Fluorescence Microscopy

Results

Characterization of Peptidosomes

Before we could encapsulate bacteria, we had to establish a robust and reproducible protocol of Peptidosome generation under lab conditions. To tackle this task, we used pH indicator cresol red solutions to visualise generated Peptidosomes (Figure 5). In addition to ease the handling of the Peptidosomes, the pH indicator colour also correlated with the status of the Peptidosomes formation. Prior the CO₂ exposure the indicator was red (Figure 5, A), after sufficient CO₂ exposure (will be discussed later) the solution changed to yellow (Figure 5, B), indicating a self-assembled Fmoc-FF layer and thus ready-to-use Peptidosomes.

These generated Peptidosomes stayed stable after CO₂ treatment and we were able to transfer them into liquid filled petri-dishes or well plates.

After we having established the protocol for the Peptidosome creation, we evaluated if we could vary the size of the Peptidosomes. For testing this, we simply generated Peptidosomes of different volumes and could demonstrate successful created and stable Peptidosomes ranging from 1 to 20 µl (Figure 6).

For discovering the optimal exposure time to CO₂ we varied it between 30 sec and 20 min. We found out, that the ones with an exposure time of 10 min were the most stable ones, showing stability after several days, even under 37°C incubation with shaking (Figure 7).

The final aim of the project is the encapsulation and cultivation of bacteria inside the Peptidosomes. For this, it is crucial that while the Fmoc-FF network should hold back the bacteria on the other side it should allow the interchange and communication with the surrounding environment. Thus, allowing fresh nutrients for the organism to enter the inside, and/or to allow the release of secreted molecules to the outside of the Peptidosomes.

Using a straight forward experiment, applying the pH-indicator we were able to prove that a diffusion between the inside of the Peptidosome and the surrounding environment is possible. We transferred Peptidosomes to water or LB medium and monitored the time the pH indicator needed to diffuse outside of the Peptidosomes.

The yellow color of both Paptidosomes (in water and LB) faded over time (Figure 8). In water (left panel), the yellow color completely vanished after about 25-30 minutes. In LB the same effect was visible after 45-50 minutes (right panel). These observations met our expectations, as the Fmoc-FF membrane is a network of tiny pores, rather than a closed shell. These pores allow the exchange with the surrounding liquid.

Proving that diffusion between the inside of the Peptidosome and the surrounding is possible, we next tested if diffusion between two Peptidosomes which are in direct contact is possible. For this experiment only one of the two connected Peptidosomes contained the coloured pH indicator solution. Over the lapse of 1.5 hours, it was possible to observe a diffusion of the colour between both peptidosome until it reached an equilibrium (Figure 9). This proved that there is intercommunication, between the two Peptidosomes.

Additionally, to the previous described method of Peptidosome generation, we also tried a microinjection-technique. By doing so, we injected a coloured solution to the inside of the Peptidosome (Figure 10). To close the membrane hole that was introduced by the thin glass capillary, the Peptidosome was after injection again exposed to CO₂ for 5 min. Followed by this procedure, the Peptidosome was transferred to water and as shown in Figure 10 E) the peptidosome was stable.

---

Encapsulation of Bacillus subtilis in Peptidosomes

Before the encapsulation of B. subtilis inside the Peptidosomes, it was necessary to perform experiments to study possible interactions between the organism and the Fmoc-FF. One aspect tested was, whether the bacteria can use Fmoc-FF as a nitrogen source. And also, if the organism can survive the process of encapsulation, since the bacterial pellet is resuspended in the (high) alkaline Fmoc-FF solution.

Can Fmoc-FF use B. subtilis as a nitrogen source?

To test, if B. subtilis can degrade Fmoc-FF, we tested bacterial growth on Jensen’s medium agar plates. This medium has all the basic nutrients that bacteria needs for growth, with exception of nitrogen.

The experimental used agar plates were set-up was following:

1. Nitrogen free Jensen’s medium
2. Jensen’s medium supplemented with nitrogen (adding Fe-[III]-ammonium citrate and potassium glutamate)
3. Liquid Jensen’s medium mixed with Fmoc-FF
4. Jensen’s medium supplemented with Fmoc-FF after the plates were dried

Type one plates serve as negative control, since B. subtilis (wild type) should not be abel to grow in the absence of nitrogen. On the second plates, we added two nitrogen sources (Fe-[III]-ammonium citrate and potassium glutamate) allowing B. subtilis to grow. With plates of type three and four we intended to test if B. subtilis is capable of using the Fmoc-FF. We could only detect growth on type two plates, demonstrating that B. subtilis can neither grow with out provided nitrogen (type 1 plates) nor use Fmoc-FF as nitrogen source (type three and four plates).

Tolerance of B. subtilis against alkaline pH

In the course of Peptidosome generation, B. subtilis will be exposed to alkine conditions. Therefore we performed a pre-test to check the tolerance of B. subtilis towards high pH. Cultures were treated with four different NaOH molarities to adjust the final pH in the wells to 7, 8, 9 or 10. NaOH was added after one hour of cultivation (Figure 11, highlighted by the red dot). The experiment was carried out in a plate reader to follow the change of the optical density (OD₆₀₀) over a time period of 14 hours.

Figure 11, shows the growth of B. subtilis under various pH conditions. No significant changes in growth can be observed for pH 7 to 9. However, under pH 10 conditions, B. subtilis stops growing and shows a slight decrease in optical density. To check if the bacteria were still viable after the high pH treatment, we transferred bacteria at the end of the assay (after 14 hours) to agar plates and could obtain colony forming units. Thus, demonstrating that B. subtilis was able to recover and continue growing normally. Overall, we could clearly show that B. subtilis can survive the process of Peptidosome generation in regards to the pH conditions.

B. subtilis in Peptidosomes

After we have proven that B. subtilis can survive the designed encapsulation process, and that the organism cannot use the building block Fmoc-FF as a nitrogen source, experiments to entrap the bacteria started.

For this we resuspended an appropriate amount of bacteria in the Fmoc-FF solution. Afterwards droplets of this were placed on an ultrahydrophobic membrane and exposed to CO₂ for 10 minutes, followed by a transfer to water or LB medium.

You can find the detailed protocols here.

Well Scan experiment using B. subtilis reporter strains

For the first experiments of encapsulation, we used the two B. subtilis reporter strains: TMB4131 W168 lacA::erm P_veg-sfGFP and TMB3090 W168 sacA::cat P_veg-luxABCDE. The first strain constitutively expresses sfGFP, which we used to detect the presence of bacteria in the Peptidosomes by following the fluorescence signal. The latter strain, expresses luciferase in a constitutive way, which makes a detection of a luminescence signal possible. Having both of these well-evaluated readouts at hand, we wanted to demonstrate their applicability, with bacteria encapsulated in Peptidosomes. We performed a plate reader assay using the well-scan-mode. In this mode, the whole well is scanned to detect the exact position of a signal, either fluorescence or luminescence. Its absence is displayed in a map with a green color, while the position of the fluorescence/luminescence source appears as red. Please check out the according protocol for details.

In Figure 12 examples of the well scans of different samples are displayed.

The well scan maps appear completely green when fluorescence/luminescence signals are absent, i.e. in water and empty Peptidosomes. Wells containing a sample of the day culture and lyophilized eGFP dissolved in water show a red colour in the whole map. However, a localized red spot over a green background is observed when the source of fluorescence/luminescence is contained: the cells are trapped inside the Peptidosomes. This also shows, that Peptidosomes containing cells hinder the bacteria from diffusing and rather keeps them contained within the structure, demonstrated by the locally limited red signal. We were able to reproduce these results over a longer time period under shaking, highlighting the robustness of Peptidosomes.

Fluorescence microscopy with B. subtilis

Next, we used Fluorescence microscopy to analyse a B. subtilis strain encapsulated in Peptdiosomes, which constitutively expresses sfGFP (Figure 13). We could demonstrate, that the fluorescence signal was restricted to the Peptidosome - standing in line with the results obtained in the well scan experiments. This was the final proof that the bacteria are encapsulated in the Peptidosomes and are fully trapped.

The Growth of B.subtilis in Peptidosomes

After we checked the functionality of reporters in Peptidosomes, we were also interested to observe growth of B.subtilis while encapsulated. Since the Peptidosomes should be liquid-filled and allow the exchange of molecules, B.subtilis should be able to grow. To demonstrate this, we loaded Peptidosome with a known amount of bacteria and plated them on agar plates after different time periods of incubation at 37°C. These plates were then incubated over night and we documented the colony numbers on the next day (Table 1). We expected to observe an increase of the colony forming units in correlation of longer incubation times. Peptidosomes plated after 0 hours represent the starting bacterial concentration, as these Peptidosomes were plated on agar plates immediately after their generation.

Time [h]	0	3.5	7
Colonies	370.67 ± 57.42	593.33 ± 74.44	lawn

As observed on the results, the number of colonies in the plates increased after 3.5 h of incubation, when compared to the starting bacterial concentration (0 hours). After 7 hours it was not possible to count the amount of colonies, cause B.subtilis formed a lawn. This clearly showed that B.subtilis was fully capable to grow inside of the Peptidosome cages. Thus, Peptidosomes are indeed suitable for the establishment of bacterial cultures.

Cryo-Scanning Electron Microscopy

To fully investigate how B.subtilis is encapsulated in the Peptidosomes, we performed Cryo-Scanning Electron Microscopy (Cryo-SEM). By doing so, we should get inside if the surface of the Peptidosome is also covered by bacterial cells.

As one can clearly see in the images above, some cells are completely or only partially integrated into the membrane of the Peptidosome (Figure 14, B). These cells, which are found on the outer surface of the membrane, if released, can result in bacterial growth outside of the Peptidosome.

To address this issue, we designed a second growth experiment, in which we also plated the supernatant of each Peptidosome to check the amounts of the "released" bacteria.

During this experiment two additional counter measurements were implemented, the first one consisted in adding “washing” step, were the Peptidosomes were transferred twice to fresh media before their incubation. We hypothesized that the bacteria on the outer membrane would be released and left behind in the course of the transfers. The second treatment that we tested consisted in pre-incubating of the Peptidosome in LB broth adjusted to an acid pH, this should guarantee an increased closing of the membrane by triggering the self-assembly of the any unreacted Fmoc-FF solution still present in the Peptidosome membrane.

Treatment	Normal	Washing	pH treatment
Colonies	203.7 ± 309.1	0.7 ± 1.2	1.7 ± 1.9

By introducing the treatments, the number of colonies was dramatically reduced from several hundreds to an average of 0.7 for the washing steps and 1.7 for the low pH treatment (Table 2). From here, we concluded that the washing steps are important to reduce the number of cells released in the supernatant, whereas the treatment with acid broth does not add a significant effect on top of this. By introducing the washing steps and also reducing the concentration of cells in the Peptidosomes, the amount of bacteria present on the outer surface of the membrane was dramatically decreased. We also confirmed this by performing a second round of cryo-SEM, showen below.

In this second round of cryo-SEM the number of detectable bacteria that lie or are integrated in the membrane is significantly lower. Only a few individual cells are visible and the surface is less wrinkled. (Figure 15 A), B))

Only on the Peptidosome with the low pH treatment, a cell that was not integrated in the membrane was observed (Figure 15, D).

The results of the growth experiment with the additional treatments were confirmed microscopically. There was no significant improvement in the Peptidosomes treated with acid LB medium compared to the sole integration of washing steps. This confirms the reduced number of colonies in the supernatant of the growth experiment when carrying out the additional treatment methods (Table 2 and Figure 15).

The one loosely attached cell might be a problem for some applications where the surrounding medium is expected to be free of bacteria, therefore, the manufacturing technique of Peptidosomes should be adapted accordingly. An adequate method for this could be introducing the cells by microinjection, since there is not a chance that bacteria can be attached to the outer part of the membrane.

---

3. Surface Decoration

The properties of self-assembled Fmoc-FF membrane allow tiny objects, such as Dynabeads, to be incorporated in the membrane of Peptidosomes. Incorporating Dynabeads enables the control of the Peptidosomes movements in a magnetic field, provides a powerful delivery tool, which can be used in various applications.

Moreover, the surface of Dynabeads can be decorated with a His-Tag. For the “proof of principle” experiments His-tagged GFP were selected allowing an easy imaging procedure. First, the Invitrogen protocol was optimized for our goals and tested for Dynabeads decoration with histidine-tagged GFP (Figure 16).

To prove the surface of the Dynabeads enveloped to the Peptidosome membrane availability for the decoration the labeling procedure was applied after the Peptidosome formation in both binding/washing buffer and LB media (Figure 17).

Here, we clearly demonstrated the possibility of Peptidosome surface decoration with histidine-tagged GFP. Hereby the GFP serves as a proof of principle that can be exchanged with any other protein of interest and thus used for various applications where enzymatic activity of the surface of Peptidosome or Peptidosome immobilization is required. Surface decoration protocol is compatible with almost any goals due to its flexibility, as the decoration procedure can be performed before and after Peptidosome generation.

---

Conclusion

Our team successfully invented the new immobilization system, the Peptidosome. We established a robust protocol for creating the Peptidosomes. With that we are able to encapsulate bacteria that can live and grow inside the Peptidosome. In Addition we managed do functionalize the cages by decorating the surface. We can proudly announce that the Peptidosomes are the new fundamental approach for generating and applying encapsulated bacteria.

References

[1]	K. Tao, A. Levin, L. Adler-Abramovich, and E. Gazit, “Fmoc-modified amino acids and short peptides: simple bio-inspired building blocks for the fabrication of functional materials.,” Chem. Soc. Rev., vol. 45, pp. 3935–3953, 2016.
[2]	M. Reches and E. Gazit, “Casting Metal Nanowires Within Discrete Self-Assembled Peptide Nanotubes,” Science (80-.), vol. 300, no. 5619, pp. 625–627, 2003.
[3]	C. H. Görbitz, “The structure of nanotubes formed by diphenylalanine, the core recognition motif of Alzheimer’s β-amyloid polypeptide,” Chem. Commun., no. 22, pp. 2332–2334, 2006.
[4]	M. Reches and E. Gazit, “Self-Assembly of Peptide Nanotubes and Amyloid-like Structures by Charged-Termini-Capped Diphenylalanine Peptide Analogues,” Isr. J. Chem., vol. 45, pp. 363–371, 2005.
[5]	V. Jayawarna et al., “Nanostructured hydrogels for three-dimensional cell culture through self-assembly of fluorenylmethoxycarbonyl-dipeptides,” Adv. Mater., vol. 18, no. 5, pp. 611–614, 2006.
[6]	C. Tang, A. M. Smith, R. F. Collins, R. V. Ulijn, and A. Saiani, “Fmoc-Diphenylalanine Self-Assembly Mechanism Induces Apparent pK a Shifts,” Langmuir, vol. 25, no. 16, pp. 9447–9453, 2009.
[7]	H. G. Braun and A. Z. Cardoso, “Self-assembly of Fmoc-diphenylalanine inside liquid marbles,” Colloids Surfaces B: Biointerfaces, vol. 97, pp. 43–50, 2012.
[8]	M. Zhang, S. Feng, L. Wang, and Y. Zheng, “Lotus effect in wetting and self-cleaning,” Biotribology, vol. 5, pp. 31–43, Mar. 2016.