Design

GIFTs are designed as structures that allow selective transport of substances. This system is made of a large liposome that has smaller liposomes inside it, forming a “GUV-in-GUV” structure. Each inner compartment can take in and release molecules through DNA origami nanopores that work with a docking–undocking system.
In the following sections, we explain the system in more detail from two main viewpoints: the GUV-in-GUV structure and the DNA origami nanopores.

The membrane structure of GIFTs consists of a large liposome that contains smaller liposomes inside. The design helps create separate areas inside the large liposome. In this project, we used GUVs (Giant Unilamellar Vesicles) for both the inner and outer liposomes, since they are easier to observe by a microscope. In this project, we refer to the inner vesicle as the inner liposome and the outer one as the outer liposome. When producing the GUV-in-GUV structure, three factors are especially important: the lipid composition, the composition of the solution, and the centrifugal force used during preparation.

Lipid Composition 

When making the GUV-in-GUV structure, it is necessary to keep a distance between the inner liposome and the outer liposome. In this project, we controlled the distance between the liposomes by adjusting their lipid composition. Lipid membrane is made of DOPC, DOPG, and cholesterol in a molar ratio of 3:1:6. Because DOPG carries a negative charge, it helps maintain a proper distance between the inner and outer liposomes. As a result, this lipid combination improved the yield of GUV-in-GUV formation compared with other lipid compositions.

Solution Composition 

To improve the yield of the GUV-in-GUV structure, we needed to adjust the density of each solution to make a proper difference. In our setup, the inner solution contained 300 mM trehalose, the outer solution contained 75 mM trehalose and 225 mM glucose, and the second outer solution contained 300 mM glucose. 

This combination made it easier for liposomes to pass through the oil–water interfaces during formation. In particular, the trehalose concentration in the outer solution strongly affected whether the GUV-in-GUV structure could form successfully. If the trehalose concentration was too low, the outer liposome membrane could not form properly. On the other hand, if it was too high, many outer liposomes without inner liposomes were produced.

Centrifugal Force 

We fabricated GUV-in-GUV structures using a centrifugation method. Liposomes are formed when the emulsion passes through the interface between the oil and water layers.
In the first step, inner liposomes were formed by centrifugation at 8000 × g for 5 minutes.
In the second step, we made outer liposomes by centrifuging the emulsion solution that contained the inner liposomes. The outer liposomes were formed by stepwise centrifugation at 3000 × g for 5 minutes, 6000 × g for 5 minutes, and 8000 × g for 5 minutes each. We found that performing this three-stage centrifugation in the second step enabled the successful formation of GUV-in-GUV structures. We consider that a gradual increase in the centrifugal force helps the process. In the process of 3000 × g for 5 minutes, the emulsion containing the inner liposomes moves closer to the interface. This allows the outer liposome to encapsulate the inner one more effectively. And liposomes were formed by passing through the interface at 8000 × g for 5 min.

Detection Method 

We used fluorescent dye-modified lipids to observe each membrane of GUV-in-GUV structures. NBD-PE was added to the inner liposome, and rhodamine-PE was added to the outer liposome. When a GUV-in-GUV structure was formed, a red ring was expected to appear outside a green ring under a fluorescence microscope. In some cases, when the diameters of the inner and outer liposomes were almost the same, the red and green rings appeared to overlap. However, since our goal was to produce a double-membrane structure with a clear gap between the two membranes, we considered samples where the two fluorescent rings overlapped to be unsuccessful.

In this GIFTs structure, the DNA origami nanopore and Signal DNA have two advantages for molecular transport. The first is that the nanopore can repeatedly dock and undock through the docking–undocking process via strand-displacement reactions. The second is that the system allows selective binding of a specific DNA origami nanopore. In the following sections, we summarize these two aspects: the structure of the DNA origami nanopore and the Docking–Undocking System.

Structure of the DNA Origami Nanopore 

The DNA origami nanopore in GIFTs is composed of two parts: a Wing part and a Pore part. On the back side of the Wing part, a total of cholesterol molecules are attached via the DNA strands. These cholesterol modifications allow the DNA origami nanopore to stably anchor and insert into the lipid membrane. The Pore part has two main functions. First, it penetrates the lipid bilayer, connecting the inside and outside of the liposome. Second, it serves as a connecting site during the Docking–Undocking process.

Design and Simulation of DNA Origami Nanopore 

The DNA origami nanopore was designed using Cadnano and simulated with oxDNA. M13mp18 single-stranded DNA was used as the scaffold strand. The oxDNA simulation was successful. The resulting RMSD plot is shown below.

Using oxDNA simulations, we analyzed the structural stability of the DNA origami nanopores designed in this project. The results confirmed that the structural stability and dynamics expected during design were successfully reproduced. Analysis of the RMSD (root-mean-square deviation) and the number of hydrogen bonds during the simulation showed that the DNA structures maintained shapes close to the initial design. In particular, the double-stranded regions retained most of their base pairs, demonstrating sufficient thermodynamic stability.

Docking–Undocking system via Signal DNA 

The DNA origami nanopores in GIFTs can repeatedly dock and undock in response to signal DNA, allowing multiple cycles of binding and unbinding.
We named the DNA strands used in the Docking–Undocking system as shown below.

The Docking–Undocking System proceeds through the following steps.
1. Preparing for Docking

2. Docking and Lid Opening

3. Undocking

4. Installing a Lid and Docking Blockers

5. Removing Undocking Signals

We investigated the above strand displacement reactions using Visual DSD. The simulation results showed that the Docking–Undocking reactions we proposed proceeded as designed and exhibited the expected kinetic behavior. In Visual DSD, the simulations were performed deterministically.

We designed the DNA sequences involved in this Docking–Undocking System and analyzed their behavior using NUPACK. We minimized off-target binding and the formation of undesired secondary structures as much as possible. From the simulation results, we ultimately allowed the presence of a few complexes that were unstable. These included intra-molecular bindings with an equilibrium probability below 0.7 or complexes with very little folding, as these were considered acceptable.

Fig.6 NUPACK analysis of designed DNA strands in the Docking–Undocking system.

Simulated concentrations of possible complexes (left) and minimum free energy (MFE) structures with ensemble base-pairing probabilities (right). The analysis was performed to confirm that undesired secondary structures and off-target bindings were minimized. 

Parallelization of the Docking–Undocking System 

The Docking–Undocking system is initiated when the Docking Signal displaces the Docking Blocker from the iSNARE. By modifying the recognition sequences between the Docking Signal and the Docking Blocker (for example, Docking Signal α binds to Docking Blocker α), it is possible to remove only the corresponding Lid in response to a specific signal. Specific DNA sequences were used to distinguish between nanopores that share the same overall design. This enables selective activation of individual nanopores.

Iwabuchi S., Fukami N., Sato Y., and Nomura S.-i. M., “Construction of Artificial Cell-type Molecular Robots,” Seibutsu Butsuri (Biophysics), vol. 62, no. 3, pp. 178-180, 2022, doi: 10.2142/biophys.62.178 

Berhanu S., Ueda T., and Kuruma Y., “Artificial photosynthetic cell producing energy for protein synthesis,” Nat. Commun., vol. 10, art. no. 1325, Mar. 2019, doi:10.1038/s41467-019-09147-4.