Truncated Octahedron Building: Proof of Concept

Step 0. Polygen

Using the Polygen Software, we did a prediction of the different sequences needed to start the assembly of the truncated octahedron DNA nanocages. The obtained output is listed below.

Polynucleotide 1:


Polynucleotide 2:


Polynucleotide 3:


Polynucleotide 4:


Polynucleotide 5:


Polynucleotide 6:


Polynucleotide 7:


Polynucleotide 8:


Step 1. Assembly

1.1 Materials
  • 40 pmol/µL of each DNA Polynucleotide (Exxtend)
  • 10 units/µLT4 Polynucleotide Kinase (Fermentas)
  • 10X T4 Polynucleotide Kinase buffer
  • 3 weiss units/µL T4 DNA Ligase (Promega)
  • 2X T4 DNA Ligase buffer
  • 100mM ATP (Sigma Aldrich)
  • Nuclease-free water

1.2 Protocol

The polynucleotides used for assembly of truncated octahedron DNA nanocages were ordered in the company Exxtend Biotecnologia Ltda. First we resuspended the pellet in ultrapure water forming a 50 pmol/µL solution, following the company's recommendations. Then, to use it for our assembly experiment, we diluted again to make a 40 pmol/µL solution.

The assembly procedure was carried in two steps. For each one of them, a different mix was prepared and a program in the thermocycler was used.

First mix:

  • 20 µL of each polynucleotide (40 pmol/µL) = 160 µL
  • 30 µL 10X T4 Polynucleotide Kinase buffer
  • 3 µL ATP (100mM)
  • 6 µL T4 Polynucleotide Kinase (10 units/µL)
  • 101 µL nuclease-free water
________________________________________ Total: 300 µL

In 6 PCR tubes, were aliquoted 50 µL of the first mix and then, runned the first program on a thermocycler:

First program (kinase reaction - 5’ phosphorylation - and self-assembly of the cage):

  • 30 min 37°C
  • 4 min 65°C
  • 4 min 64°C
  • 4 min 63°C
  • ...4 min in each Celsius degree)
  • 4 min 31°C
  • 4 min 30°C

When the program pauses at 30°C, add each PCR tube 50 µL of the second mix:

Second mix:

  • 150 µL 2X T4 DNA Ligase buffer
  • 6 µL ATP (100 mM)
  • 35.82 µL T4 DNA Ligase (3 weiss units/µL)
  • 108.18 µL nuclease-free water
________________________________________ Total: 300 µL

Resume in the thermocycler and run the second program (ligase reaction):

  • 4 min 30°C
  • 4 min 29°C
  • 4 min 28°C
  • ... (4 min at each degree Celsius)
  • 4 min 17°C
  • 4 min 16°C

We leaved the samples at 16° for 65 hours (approximate). Until use, the samples were stocked at -4°C.


Step 2. Gel electrophoresis

2.1 Materials

  • Polyacrylamide:bis-acrylamide 19:1 powder
  • 10X TBE Buffer
  • 1X TBE Buffer
  • TEMED (Invitrogen)
  • 10% APS
  • 1kb DNA ladder<
  • Ethidium bromide
  • Ultrapure agarose (Invitrogen)
  • 6X loading buffer (Sinapse Inc)
  • TAE buffer with 12.5 mM of MgCl2

2.2 Protocol

In order to visualize the results of the cage assembly and to purify our DNA nanocages, we performed gel electrophoretic fractionation with the samples resulting from step 1.

Gel 1: Native DNA 5% polyacrylamide gel with 19:1 proportion of acrylamide:bis-acrylamide. Polyacrylamide:bis-acrylamide 19:1 powder was dissolved in water to make a 40% (v/v) polyacrylamide mix.

Glass plates sizes:

(1) 19.5 cm x 18.5 cm

(2) 19.5 cm x 15.5 cm

Gel size:

- 17 cm x 14.5 cm x 0.1 cm

For 30 mL of 5% gel mix:

  • 3.75 mL of 40% (v/v) polyacrylamide mix
  • 3 mL 10X TBE buffer
  • 23.25 mL of water

The electrophoresis chamber was settled with the glass plates and the bottom and sides were sealed with 2% agarose in 1X TBE. Immediately before pouring the gel mix between the plates we added the following regents to catalyze the polymerization reaction:

  • 24 µL TEMED
  • 480 µL 10% APS (fresh)

After pouring the gel mix we placed the comb and leaved polymerizing overnight.

20 µL of one PCR tube (samples containing the cages) plus 4 µL of 6X loading buffer were loaded in one well of the gel. In another well, 10 µL of the 1kb DNA ladder were loaded. The gel ran in 1X TBE at 60 volts for 1.5 hour and then, at 120 volts for about 2 hours (or until the blue color almost ran out).

The gels were stained in ethidium bromide (10 µL in 80 mL of water) for 20 minutes and visualised under UV light.

Figure 1. Result, under UV light, of electrophoresis of the DNA cage sample in a 5% polyacrylamide gel.


Results and Discussion: We can see that, while the DNA ladder entered the gel, the cage didn’t seem to have entered. As referred in the literature1,2, we expected to see a band between 500 bases and 1 kilobase of linear DNA ladder. This probably happened due to tridimensional shape and nanomolecular properties of the cage, which are hampering the cage entry in the gel. As the running buffer we used didn’t contain Mg2+, the cages weren’t stabilized and formed clusters, which weren’t fractionated by electrophoresis. This would explain why we see unusual smears in Figure 1.

To solve this problem we would need to repeat the gel electrophoresis changing some variables such as gel composition, running time and voltage. After a set of experiments for padronization, we have concluded that, for our samples, the best electrophoresis conditions would be a 0.8% agarose gel composed by 1X TAE with magnesium to stabilize the cages. The procedures are described below.

Gel 2: To obtain a 0.8% agarose gel, we dissolved 0.64 g of ultrapure agarose in 80 ml of 1X TAE buffer with 12.5 mM of MgCl2. We added 4 µL of ethidium bromide. The gel polymerized in a support of 10 cm X 10 cm after we had placed a 10-well comb over it.

100 µL of a PCR tube (samples containing the cages) plus 20 µL of 6X loading buffer were loaded in four wells of the gel. In another well, 10 µL of the 1kb DNA ladder were loaded.

We performed the electrophoretic run in 1X TAE buffer with 12.5 mM of MgCl2 for 1.5 hour, gradually increasing the voltage from 90 to 110 volts during the run. The electrophoretic cube used, Owl™ EasyCast™ B1 Electrophoresis Systems (Thermo Scientific™), was cooled to avoid gel overheating.

We visualized the gel using MiniBIS Pro - DNR Bio-Imaging Systems (+ Gel capture software) with UV light and took a picture of it, shown in Figure 2.

Figure 2: Result, under UV light, of electrophoresis of the DNA cage sample in a 0.8% agarose gel. The stronger bands represents the fraction we believe that contains the DNA cages.


Results and Discussion: In 0.8% agarose gel, we were able to fractionate our samples and see bands corresponding to the expected molecular weight equivalent to the cages. There is still a visible smear, which may be caused by some cage aggregation. In order to visualize and characterize our samples, we needed to extract it from the gel.


Step 3. DNA gel extraction

The bands corresponding to the cages were visualized with a benchtop transilluminator and excised from the gel with a scalpel. The gel pieces were put in a 1.5 ml microtube.

Spin Column Filtration

DNA agarose gel extraction was performed using a commercial kit from Zymo Research, the Zymoclean Gel DNA Recovery Kit. The procedure is:

  • Add 3 volumes of a solution with chaotropic agent to the gel pieces;
  • Melt the gel in a laboratory water bath at 50ºC for 15 minutes;
  • Place the resulting solution in a spin column with a silica filter over a collection tube and centrifuge it at 10.000 rpm for one minute. Repeat this step twice and then discard the flow through;
  • Put a washing buffer with ethanol over the filter and centrifuge it at 10.000 rpm twice.
  • Discard the flow through;
  • Elute the DNA nanocages from the filter with 10 µL of ultrapure water.

Results and discussion: We measured the sample absorbance to estimate DNA concentration and purity in Nanodrop™ ND-1000 V330 spectrophotometer. The samples concentration ranged from 18.4 to 85.6 ng/µL and some had an absorbance peak at 230 nm, probably indicating contamination with guanidinium thiocyanate, present in the extraction kit reagents. These results showed that the method used was purifying DNA with low efficiency and letting possible contaminants remain in the sample.

We did a preview analysis of these samples with DLS and electron microscopy negative staining. The DLS (performed as described below, in Step 4 section) read showed that the solution was very polydispersive (PdI = 0.899) and the medium molecule size was 413 nm. This results indicate that the DNA cages were aggregating and forming clusters. The electron microscopy, also showed big DNA clusters and no cage particles were differentiable.

Thus, in order to obtain a better cage samples we needed to change our purification protocol to a more efficient method that preferably did not have to heat the DNA nor add possible contaminants to it. Besides, the cages should be eluted in buffer with magnesium and not pure water, for the salt presence could stabilize the cages structure and avoid cluster formation. The alternative chosen was to perform an electroelution with the gel fragments directly in buffer with magnesium.

Electroelution in a dialysis bag

The electroelution was performed according to these steps:

  • Add 3 volumes of a solution with chaotropic agent to the gel pieces;
  • Place the bands excised from the gel inside a dialysis bag in the same direction they were before the fragments were excised.
  • Add 1ml of 1X TAE/Mg2+ to the bag and place it inside the same electrophoretic cube where the gel electrophoresis occurred, fulfilled with 1X TAE/Mg2+.
  • Apply a 60 V voltage for 1 hour, for the DNA pass from the gel to the buffer.
  • Invert the electrodes and applied a 100 V voltage for 1 minute.
  • Collect the buffer inside the bag in a 1.5 ml microtube.

Results and discussion: After the electroelution, the DNA concentration was measured in the Nanodrop™ spectrophotometer. The DNA concentration was 17.3 ng/µL and the absorbance curve indicated high purity. Although the DNA concentration in this 1 ml solution was small, it could be concentrated centrifuging it in an ultrafilter with the appropriate size.

A preview DLS analysis (performed as described below, in Step 4 section) indicated that clusters of DNA cage were still being formed. The PdI measured was 0.734, lower than the sample eluted in water. The molecules sized measured had two major peaks at 99.77 and 1272, higher than the expected cage size within 15 and 20 nm. Besides it was possible to see microsized cluster with bare eyes.

To precipitate these macromolecular clusters we centrifuged the solution with DNA cages for 1 min at 10.000 g. The PdI, of 0.734, was still high, however the major particle size measured was a peak of 82.36 nm.

We then centrifuged the DNA cage solution at 10.000 g for more 5 min. The results are shown in the next session.


Step 4. Visualization and characterization

The DNA solution obtained from electroelution and centrifuged for 5 minutes at 10.000 g was the sample used in for DLS and AFM. It had an estimated 13.5 ng/ml DNA concentration and the absorbance curve suggested no contaminants.

Dynamic light scattering (DLS)

The dynamic light scattering measures were done in a Zetasizer Nano series. We have made 3 sessions of 16 measures for each sample analysed and calculated the medium value. 1 ml of the solution was added to an appropriate clean cuvette and placed inside the Zetasizer. The samples temperature was stabilized for 2 minutes before the measurements.

Results and discussion: The results are illustrated by Figure 3

Figure 3. curve showing particle sizes measured by DLS.


The two peaks visible in the graph from Figure 3 represent 18.27 nm for 78.3% of the samples and 10.19 nm for 21.7% of the accounted particles. The PdI was 0.518.

These results show interesting data. The most represented particle had 18.27 nm, within the range expected for the cage size. The 10.19 nm particles are probably parts of cage or disorganized polynucleotides. The PdI shows that this solution is still polydispersive, what indicates some cluster formation. However, the Polydispersive Index from this sample was lower than before centrifugation and then the sample eluted in ultrapure water. This solution is, therefore, suitable for microscopy visualization methods.

Atomic Force Microscopy (AFM)

The AFM was performed in a PicoScan 2100 with controller MAC Mode (both from Molecular Imaging) operating in intermittent contact mode. A AFM MAC Lever tip type II from Agilent was used (resonance frequency was ~70 kHz and spring constant was ~2.8N/m).

2 to 5 µL of DNA cage solution was dropped in a mica surface (from Ted Pella Inc.) previously cleaved and let air dry. After water evaporation, measures were done with AFM.

Results and discussion: The results are illustrated in Figure 4, generated from AFM.

Figure 4: AFM generated images. The arrows indicate structures that could be our nanocages. The scale bars indicates 50 nanometers. The color code represents the particle's height.


These two images from Figure 4 indicate the presence of nano structures that are probably our cages. The structures have dimensions ranging from 30 to 40 nm and 7.5 to 10 nm of height. These results could be due to cage flattening over mica surface and deformation while the sample was air dried. Besides, AFM low resolution and interference makes it difficult to represent the structure with details, so we can’t tell if their format is exactly the predicted with AFM.

Electron Microscopy with negative staining

Electron Microscopy (EM) was performed after ample preparation and negative staining for the sample eluted in water as described in Step 3. The images obtained showed big DNA clusters and no differentiable nanocage structures (results not shown).

The same visualization method was used for the sample electroeluted in buffer and centrifuged for 5 minutes that contained, according to DLS measures, particles of the size predicted for nanocages (Figure 5). We can see electrodense regions indicated by arrows with approximate size of 20 nm. We can suppose these structures are synthetized nanocages. However, further confirmation is needed to assure that.

Figure 5: EM with negative stained generated images. The arrows indicate electrodense regions with approximate size of 20 nm that supposedly represent our nanocages.



[1] Oliveira, C. L., Juul, S., Jørgensen, H. L., et al. (2010) Structure of Nanoscale Truncated Octahedral DNA Cages: Variation of Single-Stranded Linker Regions and Influence on Assembly Yields. ACS Nano, 4 (3), 1367–1376

[2] Iacovelli, F., Alves, C., Falconi, M., et al. (2014) Influence of the Single-Strand Linker Composition on the Structural/Dynamical Properties of a Truncated Octahedral DNA Nano-Cage Family. Biopolymers 101: 992–999

Other references

Zhang, Y., Seeman N. C.. (1993) Construction of a DNA-Truncated Octahedron. J. Am. Chem. Soc., 116 (5), pp 1661–1669

He, Y., Ye, T., Su, M., et al. (2008) Hierarchical self-assembly of DNA into symmetric supramolecular polyhedra. Nature 452, 198-201