A summary of how we developed a supramolecular nanostructure for triggered simultaneous release of multiple drugs.
To demonstrate our system, we will create a DNA origami structure to anchor onto the surface of the liposome membrane. This acts as both a rigid separation membrane between the two liposomes to avoid them fully fusing together while also serving as a platform for the system to be extended to its designed purposes. Utilizing the overall negative charge of DNA, we electrostatically adhere positively charge gold nanoparticles to the DNA Origami in a stoichiometric balance that ensures that two DNA origami structures, each anchoring into a liposome, come together, and thus joining a pair of liposomes. Next, by introducing a short burst of near-infrared light, we excite the gold nanoparticles, which consequently become a controlled proximal heat source to destabilize the lipid membranes of the adjacent liposomes. The symmetrical design assures that the pair of liposomes, each potentially carrying a different drug, will release in equal amounts.
Discussion of Results
We document and evaluate our results. Refer to the Experimental Process and Lab Book for detailed information regarding our procedures!
1. Forming DNA origami
Designing DNA Origami Structure
We utilize a single layer DNA Origami structure with a hollow square design that has several design considerations.
First of all, we have decided against making the structure overly rigid, as to not exert stresses upon the lipid bilayer membrane of the liposome once cholesterol ends are anchored into the liposome. We also apply twist correction methods (x’s circled in green, Figure 1b) to flatten the structure for the same reasons. Similarly, we place the cholesterol anchor extension along the inside hollow edge, where the fluctuations are fewer. We alter the size parameters in an effort to determine how the size of the DNA origami structures and windows affects the binding ratio to liposomes. We mainly focus on creating 60x60nm and 80x80nm structures. Referring to Figure 3a below, we realized that we would experience stacking interactions. Figure 1a&b show our initial designs that had not yet accommodated for this issue, while Figure 1c shows an alternative 60x60 design to accommodate for stacking interactions as well as to increase the vertical stability of the structure.
We use the Cando web modelling software to investigate the fluctuations and stiffness of the structure. (Figure 2)
We use a Nanodrop 1000 UV-Vis Spectrophotometer and gel electrophoresis to confirm the concentration and purity of DNA before and after PEG purification (Figure 3b, 4). These numbers are applied to our stoichiometric calculations and help us maximize yields of subsequent structures. We then use a Nanosurf Easyscan 2 AFM to view the DNA origami and further evaluate its formation with gel electrophoresis. In Figure 3a we show 60x60 design #1, corresponding to gel lane #8 in Figure 3c. It is apparent that there some some stacking interactions in our AFM image (and they are likely skewed in a single direction due to issues in the washing step). We attempt to resolve stacking by initially excluding the edge staple strands, circled in blue in Figure 1a, from the thermocycling mixture for design #1 but implement this change permanently later on in design #2 (Figure 1c).
We also attempt to achieve higher resolution images later on with an Aslyum Research MFP-3D AFM, but later samples appear to have some contamination (Figure 3b).
Control of our sample concentrations was a vital task in order to maintain stoichiometric conditions for following reactions, such as liposome anchoring and gold nanoparticle adherence. DNA was purified by PEG purification, though sometimes with leftover TE buffer. We show that we can precipitate DNA after thermocycling to alter the concentration for versatile usage as a reagent in subsequent experiments. Data is shown in Figure 4 of DNA samples prepared prior to anchoring DNA to create a Liposome-DNA structure for TEM imaging. Taken with Nanodrop 1000 UV-Vis Spectrophotometer.
While we were unfortunately able to perfectly image our DNA origami, we believe we have accumulated enough data to conclude that our formation was successful. We believe some of the noise in AFM images and streaking in the gels may be due to contamination or deterioration. Our yields may be in the low end, and we keep this in mind as we continue.
2. Testing electrostatic adhesion of positively charge gold nanoparticles to DNA origami
DNA Origami Size and Charge Confirmation
We used a Malvern Instruments Zetasizer to test both the size and the charge of the DNA origami. When sized, two peaks were observed (Figure 5b). This matches our expectations as it indicates a non-unity aspect ratio in our structure. The particle sizer measures the reflection of incident light, which is reflected at different angles when incident on different planes (Figure 5a).
While these tests are advantageous to confirming the formation of DNA origami, our main focus was to evaluate the zeta potential of the DNA origami to confirm its ability to make strong electrostatic bonds with our gold nanoparticles. The measured zeta potential for our structure is approximately between -24.5mV to -27.5mV, indicating a very strongly negative zeta potential (Figure 6). AuNPs are provided from Nanopartz, and are made to a +25mV zeta potential by an industrial standard manufacturing method.
Structure Forming Confirmation with Spectrophotometry
Spectrophotometry can be used to detect the electrostatic interaction between AuNPs to DNA origami structures. When binded together, we expect another peak to form on the graph, confirming the presence and yield of a bound structure1.
Unfortunately not enough time and material was available to explore this confirmation method.
Provided this information, we are motivated to believe that the presence of a fixed DNA structure on the liposome surface should attract gold nanoparticles into its proximity.
We were unable to complete a full investigation into this aspect.
3. Anchoring DNA origami to liposomes
As an intermediate step to creating our “sandwich” structure, we utilize the biologically inspired mechanism of an amphipathic molecule to embed our origami structure’s cholesterol anchors into the lipid bilayer membrane.
The first design had cholesterol-TEG-modified anchors included in the DNA origami folding reaction. The problem with this is that cholesterol-modified structures can aggregate with each other because of hydrophobic interactions2. The second design resolves this by attaching the cholesterol-TEG-modified oligonucleotide to the liposomes before binding the DNA Origami as shown in Figure 7.
We initially measure the success of this attachment by running the sample through an anionic exchange column, and also running gel electrophoresis. If the structures were well-formed, we expected to see very little segregation between well-formed structures and remaining bands, which would be populated by unbound liposomes or DNA origamis.
Several methods are explored to evaluate the most successful anchoring method.
Incubation in solution in different order. Liposome, Chol, then DNA. Liposome plus DNA+Chol together
We tested two assembly orders: 1. assembling DNA structures to cholesterol-anchor strands then to liposomes 2. assembling cholesterol-anchor strands to liposomes then to DNA structures. The structures were assembled, and we attempted to take Cryo-TEM images of the samples. All assembled samples aggregated with reach other, resulting in inability to image the samples (images not shown). Samples where structures were assembled in order of adding cholesterol-anchor strands to liposomes and then to DNA structures was observed to have aggregated more than others.
Anionic exchange resin. Embed DNA into the resin and begin building
We tested a third assembly order, but this time the structures were assembled in the presence of anion exchange resin. We added the DNA origami structure to DEAE anion exchange resin, introduced cholesterol-anchor strands to resin-DNA structure, and finally added the liposomes.
Our results from an anionic exchange column indicate that binding did occur but at very low yields. As shown in Figure 8, samples were eluted within the first three washes. This is likely due to the low DNA content in presence of the resin, causing DNA to be eluted in presence of low NaCl concentration (100mM).
We attempted several methods for anchoring. However, we are lead to believe, due to our subsequent results, that there is more to investigate in this area to repeatably achieve consistent results for the binding geometries of our final nanostructure.
4. Formation of complete “sandwich” structure by electrostatic interactions
Cryo-TEM Evaluation of Structure Formation
TEM images were attempted on an FEI Tecnai G2 200kV TEM. Unfortunately, TEM images were not rendered and implied several possibilities. As a recap, Cryo-TEM is performed on a thin sheet of ice where the sample is frozen. The TEM shoots an electron beam to test the diffraction through the sample. However, the electron beam may not pass through areas, in our equipment, where the thickness is larger than 500nm.
We tested 3 samples. Ratio of cholesterol anchors were added in 1.6 times excess.
- Cholesterol strands were attached to DNA first, then liposomes were incubated. The sample is without gold. Approx. 2.45e12 DNA structures were added with 2.46e12 liposomes.
- Cholesterol strands were attached to anchored to DNA first, then liposomes were incubated. This sample is with 2uL of 73.7nM gold. Approx. 1.96e12 DNA structures were added with 2.46e12 Liposome structures.
- Cholesterol strands are anchored to liposomes first, then DNA origami were incubated. This sample is without gold. Approx. 1.47e12 DNA structures were added with 1.48e12 Liposome structures.
Each condition indicated that there were likely too few areas with thickness less than 500nm. This could indicate several possibilities that we have yet to explore in consideration of time.
First, liposomes are provided from Avanti Polar Lipids at nominal diameters between 60-70nm with a minor population at 200nm. However at this size (<100nm), liposomes are prone to fusion in lifespans >3 days, according to our advisors. Samples 1, 2, and 3 are performed on calcein-loaded liposomes which had dated 6 days already at time of TEM imaging.
Another implication may be the uncontrolled linking of DNA origami structures between liposomes. Considering size and proximity of cholesterol strand linkers, DNA origami structures may have linked an excess number of liposomes to each of their anchor sites, thereby creating long chains or large agglomerations of liposomes with DNA origami as its linker.
While these tests proved to be inconclusive, we assumed consistency throughout the remainder of our experiments. We assumed that it would most heavily influence the intensity of emitted calcein when triggered, but the consistency should be able to provide a baseline measurement from which we can still take conclusive measurements.
As mentioned previously, while some data implicates that interactions between liposomes, gold nanoparticles, and DNA origami likely occurred, we have not yet been able to qualify the errors that are occurring during formation.
5. Constructing control parameters to test the effectiveness of the structure towards triggered release.
Determining baseline triggered response
The Cary Eclipse Fluorescence Spectrometer is used in testing the response of calcein-loaded liposomes to gold nanoparticle excitation. In this control experiment, the DNA origami structure is excluded. Well-mixed samples of liposomes and gold nanoparticles are excited in the machine at an 800nm wavelength and the emission of calcein is measured (Figure 9).
As seen on the graphs in Figure 9, when AuNP is present in solution, the intensity falls slightly. This is likely due to a release of calcein being absorbed in the absorbance range of the AuNP
These tests evaluate the efficacy of AuNPs are our nanostructure’s trigger. From the change in intensity demonstrated in Figure 9 above and further tests, we conclude that AuNP shows promise as the trigger component in our nanostructure.
6. Simultaneous Release of Multiple Drugs
We compare the emission signal again with a Cary Eclipse Fluorescence Spectrophotometer, but this time with our DNA origami structure anchored onto the liposome surface. We expect to see higher intensity of emission in the calcein wavelength at approximately 500nm. Successful results indicate that gold nanoparticles may be successfully concentrated about the liposome, thereby increasing the efficacy of excitation to release (Figure 10)
As we increase the AuNP concentration in solution, our overall intensity falls likely to the effect of more gold nanoparticles absorbing the emission of calcein. However, as different in our control experiments in Part 5 above, the overall intensities are much lower. These problems are unfortunately those that arise from our shortened time and material availability. The batch of liposome-DNA structures used in this batch are under a different synthesis method from those used in the control experiments, and limitations in material may have limited batch volumes, which may have greatly affected yield.
As a result, it is evidenced in this data that the amount of complete and formed structures in solution may be very low.
While our final results follow our expected downward trend as the concentration of AuNP is increased, the overall intensity is low. These samples are created with an additionally tried anionic exchange resin method, as opposed to the incubation method used previous for the control experiments. While we believed that both methods should form the correct structures, it is likely that in this batch of samples, the yield was too low.
We attempt to demonstrate a nanoengineered drug delivery system capable of carrying multiple therapeutic drugs and responding to an external trigger. While the exact relationship between the DNA origami structure to the formation of the liposomal superstructure are uncertain, we have evidence that the addition of a DNA extension may aid in the release of liposomal drugs. This effect could assist in lowering the complexity of chemotherapeutic drug dosages by allowing them to be contained within a single structure and also confined to a smaller bodily area. The exploration of using gold nanoparticles as the electrostatic adhesion and trigger shows potential in the area of using electrostatic forces to construct nanoengineered supramolecular structures. The basic properties should be applicable to a wider range of non-toxic nanoparticles, enabling a wide range of triggers and treatment areas for the nanostructure. The abilities of the structure should allow for the delivery of more potent therapeutics for more efficacious treatments of cancer, due to the reduced concern of side effects.
Development of the nanostructure can still be explored in the DNA origami’s geometry. It is still inconclusive, due to the time constraints of the project, how the DNA origami geometry may affect the geometry of the final nanostructure. Several areas still remain where the parameters should be tested flexibly, such as for different sizes of liposomes, different numerical arrangements in each structure, and different liposome compositions.
In conclusion, the triggering response of our nanostructure shows great promise in concentrating potent cancer therapeutics into select areas, reducing harmful side effects during cancer therapies. DNA origami is a very versatile technique that could aid many nanotechnologies in overcoming their intrinsic limits by how simply, as we sought to show, it can externalize functionality away from another material. Development of these techniques in coordination with other sciences will greatly enhance our ability to develop novel approaches to not just therapeutic sciences but in many scientific areas.
Sastry, M., Kumar, A., Datar, S., Dharmadhikari, C., & Ganesh, K. (2001). DNA-mediated electrostatic assembly of gold nanoparticles into linear arrays by a simple drop-coating procedure. Applied Physics Letters, 78(19), 2943-2945. doi:10.1063/1.1370993 ↩
Kocabey, S., Kempter, S., List, J., Xing, Y., Bae, W., Schiffels, D.. . Liedl, T. (2015). Membrane-assisted growth of DNA origami nanostructure arrays. ACS Nano, 9(4), 3530-3539. doi:10.1021/acsnano.5b00161 ↩