Lukeman Lab
Presented by Team Toehold Conga Nanny
USING DNA ORIGAMI FOR
A NANOSCALE ELECTROCHEMICAL POSITIONER MODEL
Presented by Team Toehold Conga Nanny
1. GEL
1.1. Positioners Form Bands on Agarose Gel
1.2. The Best Concentration at which the Formation of Dimer is Reduced
1.3. Dye is Attached to Positioners
1.4. Synthesized New Floppy AQ Positioners to Make AQ Peak Visible
1.5. Created Inverted Positioners for MB Analysis
2. IMAGING
2.1. AFM Images of Positioners
2.2. TEM Images of Positioners
2.3. Measured Angles of Positioners in TEM Images
2.4 Compare the Measured Angles to Angles Found by a Referenced Research Group
3. ELECTROCHEMISTRY
4. NHS ESTER COUPLED STRANDS
5. FERROCENE NHS ESTER COUPLED STRANDS
Positioners Form Bands on Agarose Gel
Once we started the project, our goal was to make sure that positioners properly formed. Gel electrophoresis is one of the main techniques that allow us to see if origami assemblies were created. If the tight band formed and were above the plasmid band, we know that something heavier was formed. Unfortunately, we cannot tell from the gel if the shape formed properly. Another important factor is the pre-stain images of positioners. For the concentrated samples, we were able to see tight bands from the dye in the pre-stained image and that tells us that methylene blue strands attached to the positioner. This analysis is very important because it initially tells us if the sample is good enough to continue further analysis, especially for electrochemical analysis.
Figure 1. Gel 1-Unstained Gel, MB Fluorescence
Figure 2. SYBR Gold Stained Gel
The best concentration at which the formation of dimer is reduced
As we continued to anneal our initial 40 nM positioners with 20mM MgCl2 HEPES buffer, we have realized that our dimer to monomer ratio was high. Therefore, we created an experiment that allowed us to find better conditions for positioner annealing without intense dimerization. The gels below show the results of that experiment for two different positioners with different adjusters. As a result, for further experiments we have decided to try annealing positioners at a concentration of 10 nM with 15 mM MgCl2 HEPES buffer. And later on, we decided to make the positioner concentration even lower, 1 nM (represented by the 3rd gel).
Figure 3. SYBR Gold Stained Gel. Positioner 19 annealed using different concentrations of plasmid (40 nM, 20 nM, and 10 nM) and using different concentrations of MgCl2 in 1xHEPES buffer (20 mM, 15mM, 10 mM, and 5 mM).
Figure 4. SYBR Gold Stained Gel. Positioner 30 annealed using different concentrations of plasmid (40 nM, 20 nM, and 10 nM) and using different concentrations of MgCl2 in 1xHEPES buffer (20 mM, 15mM, 10 mM, and 5 mM).
Figure 5. SYBR Gold Stained Gel. Positioners 16 and 35 annealed using different concentrations of plasmid (10 nM, 3 nM, and 1 nM) in 15 mM MgCl2 in 1xHEPES buffer. (Note that the 5th and the 7th lanes represent positioners 35 and not 16)
Dye is attached to positioners
As was mentioned prior, it is crucial to make sure that the dyes/redox reporters attach to the positioner because otherwise there would be no chance of collecting electron transfer data. The pre-stained gel is the key to identify if the dye got attached to the positioner. The gels below represent different conditions of experiments with dyes for two different positioners.
Figure 6. Unstained Gel, MB fluorescence
Figure 7. SYBR Gold Stained Gel
Figure 8. Unstained Gel, MB fluorescence
Figure 9. SYBR Gold Stained Gel
Synthesized new floppy AQ positioners to make AQ peak visible
The electrochemical analysis revealed that the AQ peak is hiding behind the shoulder. That led us to create new strands with the sequence that allowed AQ dye to be more flexible with respect to the base arm. The electrochemical analysis showed defined peaks for positioner 25. By looking at the gel, destapled positioner 25 showed up as a faint band compare to positioner 19 that did not.
Figure 10. SYBR Gold Stained Gel
Figure 11. SYBR Gold Stained Gel
Created inverted positioners for MB analysis
We have created and analyzed inverted positioners to see if MB peak would be in the visible range of SWV while being on the base of the positioner instead of the arm. From electrochemical data, we were able to get some peaks but we did not have enough time to make it a reproducible peak.
Figure 12. SYBR Gold Stained Gel
Figure 13. SYBR Gold Stained Gel
AFM Images of Positioners
Figure 14 . AFM image of postioner with adjuster 13
Figure 15. AFM image of postioner with adjuster 30
We realized that we needed another method in imaging our positioners. AFM involved 3D surface typography. It uses a probe that measures the interactions between the tip and our positioner. This information allows up to see the scale and height of our positioner, which is useful but was not enough for us. We decided to switch to TEM as our primary means of imaging. TEM beams high energy electrons that pass through and interacts with a thin layer of a sample. This allows us to see a high-resolution 2D image that can tell us more about the structural details of our positioner.
TEM Images of Positioners
Figure 16. TEM image of postioner with no adjuster
Figure 17. TEM image of postioner with adjuster that has 10-base long
Figure 18. TEM image of postioner with adjuster that has 15-base long
Figure 19. TEM image of postioner with adjuster that has 19-base long
Figure 20. TEM image of postioner with adjuster that has 22-base long
Figure 21. TEM image of postioner with adjuster that has 25-base long
Figure 22. TEM image of postioner with adjuster that has 30-base long
TEM allowed us to see clearer images of the arm, base, and the adjuster helices in between. It also allowed us to see a flaw in our design. The images showed that in some of the positioners, the adjuster helices separated from one of the arms, leaving the positioner “open” and not attached to the center. This leads to varying angles between a set of positioners with the same adjuster length. We are currently working on a new design for the positioner that will have a stronger bond between the arm and base to the connecting adjuster helices.
Measured Angles of Positioners in TEM Images
Below are graphs showing the angles of positioners for each adjuster length. There is not clear enough evidence that the angle of the positioner increases with increasing adjuster length. We are still working on the design of our positioner that will lead to our desired results.
Figure 23. Distribution of angles for Adjuster blunt
Figure 24. Distribution of angles for Adjuster 10
Figure 25. Distribution of angles for Adjuster 16
Figure 26. Distribution of angles for Adjuster 19
Figure 27. Distribution of angles for Adjuster 22
Figure 28. Distribution of angles for Adjuster 25
Figure 29. Distribution of angles for Adjuster 30
Compare the measured angles to angles found by a referenced research group
Electrochemistry
With one of our main goals in mind, acquiring an SWV with clear methylene blue and anthraquinone peaks on a square wave voltammogram, a variety of variables were experimented with. These variables include:
-
Placement of dyes on origami
-
Rigidity of dyes on origami
-
Deposition time of origami
-
Type of monolayer
-
Deposition time of monolayer
-
Reducing origami before or after destapling
The original protocol included the following:
-
Rigid MB on arm of positioner. Rigid AQ on base of positioner.
-
3-hour origami deposition
-
MCH monolayer
-
2-hour monolayer deposition
-
Origami reduced with TCEP after destapling
The following protocol did not yield any MB or AQ peaks on the SWV.
Figure 30. 6-27-19 X16 3hr origami deposition, 2 hr MCH deposition - High frequencies
This summer we adjusted the protocol until we got to a set of variables that gave SWV’s with reproducible peaks. These variables are:
-
MB rigid on arm of positioner
-
AQ rigid on base of positioner
-
Overnight deposition of origami
-
MCH monolayer
-
3-hour monolayer deposition
-
Origami reduced with TCEP after destapling
​
The next step taken was to increase the monolayer deposition to an overnight deposition and compare it with a chip prepared using the original protocol. This step was taken to try to make the baseline flatter in hopes of visualizing the peaks. This was unsuccessful.
Figure 31. 6-28-19 X16B 3hr origami deposition, 3hr vs overnight MCH deposition
The next set of variables we changed were decreasing/increasing the origami deposition time to 2 hours and 6 hours, and running SWV after re-incubating the chips in MCH for 3 hours (total of 6 hours) and overnight. Changing the origami deposition time was meant to test whether a shorter or longer origami deposition time would result in a higher yield of origami sticking to the surface. Changing the MCH deposition times were meant to see if the shoulder would flatten out and to get an idea of whether or not a long MCH deposition would displace the origami. These attempts were also unsuccessful.
Figure 32. 7-1-19 X10 A 2 hr origami, 3 hr MCH deposition
Figure 33. 7-1-19 X10 A 2 hr origami, 3 hr re-incubation (6hr total) deposition
Figure 34. 7-1-19 X10 A 2 hr origami, Overnight re-incubation deposition
Figure 35. 7-1-19 X10 B 6 hr origami deposition, 3 hr MCH deposition
Figure 36. 7-1-19 X10 B 6 hr origami deposition, re-incubate overnight MCH deposition
After still not being able to visualize any peaks, we assumed there may not be enough origami adhered to the gold surface of the chip; therefore, we increased the origami deposition time to overnight while slighting increasing the MCH deposition from the baseline to 4.25 hours. At this point we began to see very slight peaks and knew we were headed in the right direction. We assumed these results were due to changing the origami deposition time to overnight.
Figure 37. 7-8-19 X25 A Overnight origami deposition, 4.25 hr MCH deposition
The next hypothesis was mixing the origami and MCH and depositing them together on the chip overnight. This was unsuccessful in producing any peaks on the SWV.
Figure 38. 7-8-19 X25 C Overnight origami+MCH deposition
Following this, we reverted to the protocol that gave us the best results, modifying the deposition times slightly. The origami was deposited overnight followed by a 3 hour MCH deposition. After hypothesizing further ways to optimize the peak that was beginning to show on the SWV, we decided to change the origami reducing protocol. Instead of reducing the origami after destapling, we reduced the origami before destapling. We assumed the TCEP in the origami solution for a long time may hinder the origami; this protocol ensures the TCEP would be filtered out. This set of variables gave the best MB peak thus far.
Figure 39. 7-9-19 X19 A overnight origami deposition, 3 hr MCH deposition, reduced with TCEP before destapling
Now that our MB peak is resolved, there is still an issue in resolving the AQ peak because it’s located in the region where the shoulder is still present. We changed the monolayer from MCH to HDT in attempt to create a better monolayer and get rid of the shoulder. The HDT didn’t result in a better baseline and the MB peak was not as reproducible as before when using MCH. In another attempt to flatten the baseline, we ran argon through the buffer before running the SWV, but that also did not give results resolving the issue. We still have not been able to achieve a protocol in which the baseline is flat enough to visualize the AQ peak on SWV.
Figure 40. 7-15-19 X22 A 3hr HDT deposition, overnight origami deposition
Although the MB peak was present, there was difficulty visualizing it, especially at higher frequencies, due to noise. In order to reduce noise, we placed the cell and electrodes in a faraday cage. This drastically reduced signal to noise and the peak was now apparent at increased frequencies.
Figure 41. 7-15-19 X22 D no faraday cage
Figure 42. 7-15-19 X22 D with faraday cage
Some other changes we made were to the design of the positioner in terms of the dyes. One approach was making the AQ dye floppy instead of rigid. We made this change in attempts to resolve the AQ peak in case its rigidity was resulting in inefficiency of electron transfer.
Figure 43. 7-24-19 X19 A floppy AQ
We also tried putting MB on both dye positions in order to see if the peak were to double, indicating electron transfer was happening at the base of the positioner.
Figure 44. 7-31-19 X25 B MB both dye positions
Finally, we tried inverting the MB and AQ, also in hopes of seeing if the dye on the arm was giving any signal. Since MB peaks show in a region that has been decipherable, we anticipated seeing a peak in that region, indicating the dyes on the arm were performing the way they should. For the most part, these attempts were not successful or reproducible.
Figure 45. 7-17-19 X19 B Inverted AQ and MB dye positions
NHS Ester Coupled Strands
Based off our lab’s results from electrochemistry on our DNA origami, we decided that we wanted to try to develop DNA origami that would stick to the gold surface of our chips more, which led us to completing a coupling reaction between dithiol strands and a 5’ or 3’ NH2 terminal strand because dithiols have a greater affinity towards gold. This means that, if we incorporate our coupled dithiol strands into our DNA origami, stronger bonding between the origami and the gold surface may give us stronger signals from the dyes attached to the origami. Additionally, the commercially available strands with the thiols that we originally used in our DNA origami were cheap, but these thioctic acid strands (disulfides) were expensive. Therefore, we decided to make our own based off our 5’ C6 amine-labeled strands.
Figure 46. NHS Ester modification Reaction with a 5'-Amino-Modifier C6 Oligonucleotide
After synthesizing the compound, we thawed, vortexed, spun the samples down, and ran an 18.5% acrylamide denaturing gel on our strands compared to their corresponding starting material strands with 10 pmol per lane for each sample at 350 Volts. After 794 Vhrs, we removed the gels from their cassets and placed them in a SYBR Gold Gel Stain for 15 minutes. Afterwards, we scanned the gels on our Typhoon FLA 9500 and discovered that there was little to no mobility change between all our starting material strand lanes and our supposedly coupled strand lanes.
Figure 47. 18.5% Denaturing Acrylamide Gel on NHS Ester Coupled Strands
** Note: Label corresponds to the strand in the lane above the label!
Left Imaged Gel Labels:
S1 = Starting Material Arm 5M
P1 = Product Arm 5M
S2 = Starting Material Base 3A
P2 = Product Base 3A
L = 10 Base Pair Ladder
S3 = Starting Material Arm 3M
P3 = Product Arm 3M
S4 = Starting Material Base 5A
P4 = Product Base 5A
Right Imaged Gel Labels:
S5 = Starting Material Anchor 1672
P5 = Product Anchor 1672
S6 = Starting Material Anchor 7181
P6 = Product Anchor 7181
S7 = Starting Material 1872
P7 = Product Anchor 1872
S8 = Starting Material Anchor 6366
P8 = Product Anchor 6366
L = 10 Base Pair Ladder
S9 = Starting Material Anchor 7171
P9 = Product Anchor 7171
S10 = Starting Material Anchor 13171
P10 = Product Anchor 13171
S11 = Starting Material Anchor 2072
P11 = Product Anchor 2072
S12 = Starting Material Anchor 9181
P12 = Product Anchor 9181
Figure 47 shows that there is little to no mobility change between the lanes for the starting material strands and the developed product strands; however, it did not confirm completely that our coupling reaction had failed. As a result, we turned to use HPLC to test these samples and determine whether coupling had occurred.
​
Although little to no difference was shown, this did not confirm completely that our coupled reaction did not occur. Instead, it indicated that the gels wouldn’t give us the answer we were looking for, so we decided to turn towards using HPLC to purify our samples based on hydrophobicity and determine whether our coupling reaction was successful.
​
We proceeded to run HPLC on all our Anchor strands and their respective starting material strands.
Figure 48. HPLC Chromatogram of Starting Material Anchor 1872
Figure 49. HPLC Chromatogram of Product Anchor 1872
Figure 48 and 49 show that a shift occurs between the starting material’s peak at approximately 9.6 minutes and the product’s peak at approximately 14.595 minutes. Additionally, the peaks are absorbed at 257.8 nm, which is close to DNA’s absorbance wavelength, and this shift supports the idea that strands were successfully coupled.
​
When we compared the chromatograms, we were able to identify that the starting material’s main peak was approximately at 10 minutes based off the respective starting material’s chromatogram. In our product’s chromatogram, we saw that this peak at 10 minutes disappeared, and a new peak appeared at 14 minutes, which indicated that our reaction seemed to be successful in producing product. To ensure that these peaks were new and not from any substances remaining on the column after washing, we ran a “blank” run of only 0.25 M NaHCO3 / Na2CO3 buffer, which was the buffer we used in our coupling reaction to produce our strands, and no major peaks appeared like those in our samples’ runs. Therefore, this shift that occurred in 7 out of 8 of our Anchor strands indicated that our coupling reaction was successful. For our Anchor 13171 strand, no peak appeared in the product chromatogram, which suggests that product was not successfully produced in that reaction tube, and we will further inspect why this strand appeared to be different than the other anchor strands.
Ferrocene NHS Ester Coupled Strands
Since the functionalization of our gold chips through electrochemistry had some difficulty, especially with getting the proper current signals from strands with Anthraquinone dyes attached to their ends, we decided to repeat this coupling reaction with Ferrocene NHS ester to see if we could replace using Anthraquinone with Ferrocene in our positioners.
Figure 50. Ferrocene NHS Ester modification Reaction with a 5'-Amino-Modifier C6 Oligonucleotide
Figure 51. HPLC Chromatogram of Starting Material Base-3AQ-NH2 with Fc-NHS Ester
