Team:Hong Kong HKU/Results
Results
We tested the various properties of our nanostructure with different in-vitro analytical methods:
| Perspectives | Examination |
|---|---|
| 1) Tetrahedron Assembly | Gel Electrophoresis (PAGE and Agarose) |
| 2) Strand Displacement | Gel Electrophoresis (PAGE and Agarose) PCR clean up Mutant inputs |
| 3) G-quadruplex formation | Gel Electrophoresis (PAGE and Agarose) |
To recall our design, the input will induce a strand displacement, eventually change the oligo 1 in conformation to a G-quardruplex structure.
The G-quardruplex (Gq) will exhibit peroxidase activity in with the incorporation of hemin.
Wile hydrogen peroxide is used as substrate, ABTS is utilized to give out the resulting signal.
Simplified Beacon
It is not difficult to observe a simplified molecular beacon can be assembled Oligo 1 and Oligo 5, in which they possess a complementary region as counterparts.
This region is a functional unit of a YES-Gate, which are the essential components of the tetrahedron structure (see later for more).
We started off with this simplified version of the nanostructure to do several prelimary testings.
One of the testings was to check whether they are able to anneal as a response to the addition of input.
Below shows the results of 12% PAGE for the assembly of the beacon and the addition of the input.
The beacon was made by adding equal quantity of Oligos 1 and 5 in TM Buffer.
It is noted that the formation of complex nanostructures is allowed by a standard procedure of heating it up followed by step-wise cooling.
The detailed protocol regarding Nanostructure Assembly could be found here.
12% PAGE stained with GelRed and photographed under UV light to check for beacon assembly and input addition.
Tetrahedron nanostructure
The tetrahedron nanostructure is much more complex.
It consists of a total of 5 oligos and has a wider range of combinations subjected to investigation.
We have performed 12% PAGE to investigate into the migration of all 31 combinations of the 5 oligos (oligos alone included).
Several combinations of oligos are expected to anneal to form dimers, trimers and quadruplexes.
The results from PAGE were consistent with the list below.
It is quite easy to notice that the assembled tetrahedron stationed at the very top of the gel, meaning that perhaps it might be too large in size that is hard to move along.
A lower arcylamide percentage of the gel might give a signifcantly different result.
With the same tube of tetrahedron sample, we therefore later on performed a 8% PAGE to observe any differences on the migration of the tetrahedron.
8% PAGE gel stained with GelRed and visualized under UV light to check for differences in migration compared to gels with higher acrylamide concentration (12%)
From the above PAGE gels, still, it was well observed that the tetrahedron remained at the very top position of the gel.
It reflects the diameter of the tetrahedron to be larger than that of the pore sizes of 8% polyacrylamide gels.
We proceed to replicate the experiment on an 1% agarose gel.
Now, the oligos alone are expected to be dots as they are too small to be resolved well at 1% Agarose.
1% agarose gel stained with GelRed and visualized under UV light to show assembly of the tetrahedral structure from the five oligos O1 to O5.
It is clear that there are still some trapped at the well of the Agarose gel. From such observations, and from literature,
the gel pore size of 1% Agarose gel ranges from 100nm to 500nm.
The diameter of our nanostructure is expected to be greater than 500nm.
ref: Jean-Louis Viovy (2000). "Electrophoresis of DNA and other polyelectrolytes: Physical mechanisms".Reviews of Modern Physics. 72: 813–872. Bibcode:2000RvMP...72..813V.doi:10.1103/RevModPhys.72.813.
^ Jump up to:a b Philip Serwer (1983). "Agarose gels: Properties and use for electrophoresis". Electrophoresis. 4 (6): 375–382. doi:10.1002/elps.1150040602.
PAGE and Agarose Gel Electrophoresis
The figure recalls the strand displacement reaction.
Another replication of the identical PAGE experiment with the abive is conducted at 15% arcylamide percentage.
Apart from providing another strong evidence of the successful assembly of the tetrahedral nanostructure (Lane 7 and Lane 10),
it is even more clearer to observe a signifcant difference in the band in Lane 6 (Oligo 5), Lane 8 (Input Oligo) and Lane 9 (Oligo 5 + Input Oligo).
The formation of a larger complex proves a strand displacement to occur when the two oligo meets.
15% PAGE gel stained with GelRed and visualized under UV light to demonstrate successful assembly of the tetrahedron and displacement by input strand.
Similar to the assembly protocol, the experiment was replicated on a 1% Agarose gel:
Once again, the bands in Lanes 8 and Lane 9 demostrated a strand displacement.
1% agarose gel stained with GelRed and visualized under UV light to demonstrate successful assembly of the tetrahedron and displacement by input strand.
O1+O5 beacon and others
The following is the result of a 15% PAGE experiements that includes the analysis of the O1+O5 beacon, O1-O4 beacon and the tetrahedron.
This gel image provides another promising evidence of the assembly of O1+O5.
With reference to Lane 7 and Lane 8, Lane 4 and Lane 10, and Lane 9,
it was showed that O5 is necessary for the strand displacement as the input strand binds to and only to Oligo 5 as the speciflied target.
15% PAGE gel stained with GelRed and visualized under UV light for analysis of O1+O5 beacon, O1-O4 beacon and the tetrahedron.
PCR Method
Due to the small size of the output product, or, the dimer between the input strand and Oligo 5 (30 nucleotides long),
we manipulated this as the primers to amplify the template strand.
The template strand was made from 2 complementary single-stranded sequences in the thermocycler,
and was designed in such that both the sense and antisense strands of the output dsDNA can anneal and hence act as primers.
We first run an Agarose Gel electrophoresis against the tetrahedron displacement,
which provides another solid evidence of the strand displacement.
We then extracted the output product from the gel.
Below shows the difference between before and after cutting the 1% Agarose gel for extraction.
The yellow indicate the site where the expected displaced products should be.
1% agarose gel stained with GelRed and visualized under UV light for demonstration of strand displacement and extraction of output product.
We then set the PCR program with the aid of NEB Tm calculator for the anneal temperature to start the PCR run and performed the PAGE with appropriate controls.
We used nanodrop to measure the concentrations of all the tubes that had been in the thermocycler. Each PCR tube contains 2nM of the template strand, as contained components according to the protocol available on the NEB website.
| From Gel Extraction | After PCR | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| OUT | TD1 | TD2 | TD3 | OUT | TD1 | TD2 | TD3 | Control | |
| Nanodrop result (ng/μL) | 7.5 | 8.6 | 4.6 | 5.0 | 396.1 | 398.4 | 391.6 | 403.9 | 1.4 |
Image J analyais of the gel was performed so as to analyze the relative increase to the marker (OUT), which acted as a positive control.
| A | B | C | D | E | |
|---|---|---|---|---|---|
| Output (O5 & input) | + | + | + | + | |
| Template | + | + | + | + | + |
| Peak intensities (arbitrary units) | 1571 | 9835 | 7627 | 7440 | 8024 |
It is clear that the concentration increased and this shows the correct output product via strand displacement.
Mutants
We tested in total 6 mutants, where 3 experienced single-base mutation, each of two experienced double- and triple-base mutation, and one random mutant strand of the same length as out target DNA.
We also aspired to try the RNA inputs to test our probe, which proves the application with miRNAs as targets, which are promising biomarkers.
| Sequence | Length | |
|---|---|---|
| DNA Input | CAATCAGGGTCTAACTCCACTGGGTGCCAT | 30 |
| DNA Mutant | CAGGCAGTATCATGCGACATTGGGTGCAGC | 30 |
| DNA Input | CAAUCAGGGUCUAACUCCACUGGGUGCCAU | 30 |
| DNA Mutant | CAGGCAGUAUCAUGCGACAUUGGGUGCAGC | 30 |
We performed PAGE and compared the relative intensities of the output product using ImageJ®.
The relative intensities are expected to reflect the rate of strand displacement.
DNA
Gel image showing correct vs random input (same length)
12% PAGE gel stained with GelRed and visualized under UV light to analyze output products.
ABTS Assay is one of the methods to test the G-quadruplex (Gq) formation.
While the the upper strand of our nanostructure, Oligo 5, is displaced (shown in the schematic diagram) as the input hybridizes with the bottom strand, Oligo 1,
which contains split Gq sequences quickly folds into a 4-strand structure, in the presence of potassium ions.
When Gq forms a complex with hemin, it exhibits peroxidase activity and functions as a DNAzyme.
It reacts with ABTS or TMB, together with H2O2 and gives out colour change from colorless to green, or to blue for TMB.
We perform ABTS assay on a 96-well plate where triplicates are made possible.
The experimental procedures of the ABTS assay can be found in here.
Fold Change
In the early stage of our experiment, we tested both the tetrahedral DNA nanostructure and a simplified version of the active component of the tetrahedron
(consisting of the O1's G-quadruplex side and O5 of the tetrahedron) for DNA detection.
DNA input was added in a 1:1 molar ratio to DNA nanostructure (100nM final).
After measuring the absorbance at 420nm, we determined the fold change using the following formula.
Fold change = (absorbance of sample) ÷ (absorbance of -ve control)
Negative control refers to the reaction mixture without DNA nanostructure and DNA input.
Fold changes after the addition of DNA input was shown in the following graph, with the controls of not adding any DNA input to the DNA nanostructure.
Fig. A: Fold change in absorbance at 420nm after the addition of DNA input to the DNA nanostructures (tetrahedron and a simplified version from tetrahedron, which is termed as "beacon" in the above graph).
The absorbance was taken 9 minutes after the addition of ABTS and H2O2. Error bars show standard deviation from triplicates.
From the above bar chart, it shows that there is an increase in the absorbance with the addition of DNA input.
Hence, it is likely that G-quadruplex was formed with the addition of input.
It is also noted that the activity of the simplified version of tetrahedron (termed as "beacon" in Fig. A) is similar to that of the tetrahedron.
Mutant testing
After showing that there is indeed an increase in absorbance due to the addition of input and hence the formation of G-quadruplex,
we went further to test if such change is specific to the sequence of our input by testing random DNA sequence.
The sequences of DNA inputs tested are shown in the table below.
| Input (DNA) | Sequence | Length (nucleotide) |
|---|---|---|
| Specificd Input | CAATCAGGGTCTAACTCCACTGGGTGCCAT | 30 |
| Random mutant | CAGGCAGTATCATGCGACATTGGGTGCAGC |
Again, we carried out ABTS assay to detect the presence of G-quadruplex as a result of strand displacement by the input. The probe used in this section is a simplified version consisting of only the Gq side O1 and O5 of the tetrahedron, which is the active component of our tetrahedral DNA nanostructure. Below is the bar chart showing the absorbance of the detection beacon with the addition of a correct DNA input and random DNA, together with appropriate controls.
Fig. B: Absorbance at 420nm after the addition of different DNAs to the simplified DNA nanostructure (formed from O1's G-quadruplex side and O5 of the tetrahedron) which is termed as "beacon" in the above graph.
The absorbance was taken 10 minutes after the addition of ABTS and H2O2. Error bars show standard deviation from triplicates.
It can be seen that adding a random DNA resulted in an absorbance similar to adding no input to our detection beacon, while adding the correct input results in a much higher absorbance.
This reflected that our detection beacon can differentiate the correct DNA input from a random DNA sequence.
Limit of Detection
After knowing that our design can distinguish the correct DNA input from random sequence,
we were interested to know what was the lowest concentration of input that can be detected..
As a result, we determined the limit of detection (LOD) of our detection beacon (formed from O1's G-quadruplex side and O5 of the tetrahedron, the active component of the tetrahedral nanostructure) by ABTS assay.
Different concentrations of input (DNA) were added and their respective absorbance at 420nm was measured.
A regression line obtained is shown in the following graph.
Fig. C: Absorbance at 420nm against the concentration of DNA input to the simplified DNA nanostructure (formed from O1's G-quadruplex side and O5 of the tetrahedron).
The absorbance was taken 15 minutes after the addition of ABTS and H2O2. Error bars show standard deviation from triplicates.
The regression line obtained is y=0.0009x+0.2089 (R2=0.9637).
The LOD is calculated as follows.
CLOD = 3(sy/x)÷b, where
CLOD is the concentration LOD,
sy/x is the standard error of regression, and
b is the slope of regression line.
First, the standard error of regression is determined.
| X | Y | Y' | Y-Y' | (Y-Y')2 |
|---|---|---|---|---|
| 0 | 0.201 | 0.2089 | -0.00789999999999999 | 0.0000624099999999998 |
| 20 | 0.230333333333333 | 0.2269 | 0.00343333333333301 | 0.0000117877777777756 |
| 40 | 0.249 | 0.2449 | 0.00409999999999999 | 0.0000168099999999999 |
| 60 | 0.273666666666667 | 0.2629 | 0.010766666666667 | 0.000115921111111118 |
| 80 | 0.279 | 0.2809 | -0.00189999999999996 | 0.00000360999999999984 |
| 100 | 0.295 | 0.2989 | -0.00390000000000001 | 0.0000152100000000001 |
| SSE | 0.000225748888888893 | |||
(Y' is the predicted value from the regression line y=0.0009x+0.2089)
Standard error of regression = √(SSE÷no. of pairs)=√(0.0002257÷6)=0.006133
Limit of detection
CLOD = 3(sy/x)÷b = 3(0.006133)÷0.0009 = 20.44nM
A step closer to real-world application - moving to RNA detection
After conducting a series of experiments to detect DNA, we moved on to RNA detection which is more closely related to the potential real-world application of miRNA detection using DNA nanostructure for disease diagnosis.
Results regarding RNA detection can be found in here.
Appendix - Preliminary Tests
Toehold length optimization
Series of preliminary testing were carried out at the initial stage of our project.
One of them is to determine the optimal toehold length.
In our design, a toehold is a region of unhybridized DNA for the hybridisation of the target by complementary base-pairing and hence the subsequent displacement.
We varied the toehold length by adjusting the number of bases complementary to O5 on O1. This gives to 10 versions of O1.
If the hybridised region is shorter, a longer toehold will be exposed.
We expected that both longer and shorter toeholds would be unfavourable for our target to hybridise to the toehold for the hybridisation between O1 and O5 due to base-pairing between them respectively.
| O1_X | Length (nt) | Length of hybridized region between O1 and O5 (nt) | Toehold length on each end of O5 (nt) |
|---|---|---|---|
| 1 | 30 | 2 | 10 |
| 2 | 32 | 4 | 9 |
| ... | ... | ... | ... |
| 5 | 38 | 10 | 6 |
| ... | ... | ... | ... |
| 10 | 48 | 20 | 1 |
In our design, we expected that after the hybridisation of input to O5 and hence the displacement of O5, O1 will be unhybridized and G-quadruplex will be formed by self-folding.
With an optimal toehold length, G-quadruplex would only be formed in the presence of the input and therefore produced the highest signal-to-background ratio.
Competent Cell Test kit
Before we used our competent cells (E. coli DH10B), LB agarose growth media etc., we made good use of the test kit to ensure that the stuffs and our minds are on the right track.
We followed the protocol availble on the iGEM website with some modifications suggest on the NEB protocol.
E.g. the cell recovery stage is amended into 1 hour instead of 2 hours suggested.
For more details, please visit ourprotocols page.
Results:
| pSB1C3 conc. | 1 | 2 | 3 | Avg | Competency (Avg colonies/conc.x1000) |
|---|---|---|---|---|---|
| 0.5 | 0 | 0 | 0 | 0 | 0 |
| 5 | 1 | 4 | 1 | 2 | 0.0025 |
| 10 | 6 | 3 | 3 | 4 | 0.0004 |
| 20 | 15 | 25 | 21 | 20 | 0.0010 |
| 50 | 17 | 61 | 67 | 48 | 0.0010 |
Plasmid and Parts
| Name (Part number) | Type | Description | Length(bp) | |
|---|---|---|---|---|
| ‽ | BBa_K2054000 | Coding | ALL oligos included in our desired nanostructure | 3641 |
| ‽ | BBa_K2054006 | Coding | Oligo 1 and Oligo 5 beacon | 2641 |
| BBa_K2054001 | Coding | Oligo 1 | 2388 | |
| BBa_K2054002 | Coding | Oligo 2 | 2358 | |
| BBa_K2054003 | Coding | Oligo 3 | 2375 | |
| BBa_K2054004 | Coding | Oligo 4 | 2375 | |
| BBa_K2054005 | Coding | Oligo 5 | 2313 |
Basic Parts
We first digested the G-block fragments we ordered and the linearised p_SB1C3 backbone by PstI and EcoRI.
Then we put onto a ligation at a molar ratio (backbone:insert) of 1:1, 1:3 and 1:10.
1% Agarose Gel to show the ligation (1:3)
The ligated products (backbone-to-insert ratios of 1:10 and 1:3 and 1:1) were then transformed following the protocol available on NEB.
The ligated p_SB1C3 digested with PstI and EcoRI was transform as an negative control, where no colonies were observed on the corresponding plate and those for ratio of 1:1.
On the other hand, considerable numbers of colonies were observed on the plates with the plasmids shown above.
It was mini-preped on the following day after another 14-16 hrs of incubation in LB broth.
An 1% agarose gel electrophoresis was performed to show the mini-prep results:
From the above, some appear to have the plasmids of the right size e.g. p_O1 in lane 2, while some appear to have possible contaminations e.g. p_O4 in lane 10, which showed the possibility of wrong colonies picked.
So we did a colony PCR to select the appropriate ones.
Colony PCR
The colony PCR was performed following the NEB protocol.
A 1% agarose gel electrophoresis was also performed right after to determine which colonies contain the right inserts-ligated plasmids.
We picked 4 colonies per plate.
The couple of primers used was insert-specific, which will amplify the inserts, having expected band sizes as follows:
| Plasmid length (bp) | Insert length (bp) | |
|---|---|---|
| p_O1 | 2388 | 411 |
| p_O2 | 2358 | 381 |
| p_O3 | 2375 | 398 |
| p_O4 | 2375 | 398 |
| p_O5 | 2313 | 336 |
1% Agarose of the PCR products, the colonies were labelled.
The upper set was from ligation with backbone-to-insert ratios of 1:3, and 1:10 in the lower set.
Colonies 1a-1d, 2b-2d, 3a-3d, 4b-4d, 5a-5d (1:3 ratio) and 1a-1d, 2a-2b, 3a, 3b, 3d, 4a-4d and 5a-5d appeared to be the appropriate ones.
We selected those with weaker intensities from these for inoculation into the LB broth (with 20ug/mL Chloramphenicol added) and incubated them overnight at 37°C.
Mini-prep was performed followed by a 1% gel electrophoresis.
PCR was also performed following the protocol available in NEB with insert-specific primers such that inserts (approximate sizes) would be amplified.
Below shows the 1% Agarose Gel electrophoresis of the PCR product.
It is expected to have band sizes as on the above table.
1% agarose gel stained with GelRed and visualised under UV light to show PCR products of mini-preped plasmids.
Restriction Mapping
Multiple bands were seen in each lane, and this could be due to different configurations of the plasmids which affect the migration through the gel pores.
E.g. in the 2nd lane, the plasmid p_O1 might exist in super-coiled, linearised and circular (nicked circle) configurations which gave rise to the lower, the middle and the upper bands.
The middle band matches the band sizes of the ladder on the left, which is around 2300-2500bp.
Diagram to show the migration of different configurations of plasmid from literature (above) and our plasmid p_O1 (below)
To double check whether the above was the case and whether the mini-preped plasmids are the real deals, an RE digestion with EcoRI was performed following the protocol from NEB followed by an 1% Agarose Gel Electrophoresis.
Featured Parts - p_O15 and p_O12345
Within the prefix and suffix of our parts, there are 4 restriction sites - EcoRI & XbaI (prefix), SpeI & PstI (suffix). And for XbaI and SpeI RE sites, they are overlapping in the sticky ends:
XbaI RE site:
5’ – T^CTAG_A – 3’
After cutting:
5’ – T and CTAGA – 3’
3’ – AGATC and T – 5’
SpeI RE site:
5’ – A^CTAG_T – 3’
After cutting:
5’ – A and CTAGT – 3’
3’ – TGATC and A – 5’
The sticky ends produced then will anneal such that neither SpeI nor XbaI will recognise.
We can therefore fuse the G-block fragments by cutting one couple of G-block fragments one at a time with one RE applied in each, then ligate it to produce a ‘dimer’. Taking our p_O15 as an example, we will first digest the suffix of O1-fragment with SpeI and the prefix of O5-fragment with XbaI. Then we performed ligation.
After that, we will digest the backbone and ligated O1-O5 fragment with PstI and EcoRI, and ligate them together to make it as a complete plasmid.
The flowchart below illustrates our plan:
