Team:GreatBay SCIE/Results
Overview
Our results mainly consist of 3 parts: Firstly, the aptamer experiments: The verification of binding of previously-reported alpha-amanitin aptamers with alpha-amanitin, the selection of beta-amanitin aptamers through SELEX and their verifications of binding by ELONA, dot blot, and gel shift. Secondly, the microbial production of the scFv polypeptide and the verification of binding towards alpha-amanitin by ELISA. Finally, the hardware construction from a diverse range of aspects including concept verification by competitive magnetic beads assay, RPA amplification test, Au nanoparticle reporting tests along with lateral-flow test strip tests.
Aptamer Experiments
First and foremost, the core concept of our project is based on the specific binding between the aptamer and its target, amatoxin. Hence, before selecting new aptamers of different subtypes, we decided to demonstrate and characterize the entire mechanism of binding first, using existed aptamers of alpha amanitin developed in previous literature. [3]
According to the state of art in aptamer binding [4], we design an experiment based on the principle of Gel Shift(i.e lower rate of motion in electrophoresis when aptamer is bound to the toxin due to a conformational change). By mixing excess amatoxin determined by the calculation of dissociation constant and aptamer in certain reaction times, we expect to see 2 bands representing the bound and inbound aptamer respectively. For examination of specificity, we also mix a random ssDNA pool where we select aptamer from and compare the band with the correct aptamer. For a high resolution, we used the TBE PAGE gel for DNA, 12%.
Fig.1 The TBE-PAGE we use for Gel Shift.
Here best 2 is the 80bp aptamer selected by the researchers while best 1 is the optimized aptamer by the elimination of 10bp on 3'. Unfortunately, the band is not clear, without differentiation in-band locations. We assume that the unusual width of the band compared to marker indicates a deviation in transfer rate, however, since amatoxin is only a cyclopeptide with 8 amino acids, the conformational change on aptamer is too subtle to be detected. Yet we can still see a trend between random and best aptamers, in which the random aptamer is transferred more rapidly compared to the rest, crossing the border of TBE PAGE gel. The difference between the best 1 aptamer and the best 2 aptamers is owing to the difference in length. It gives us the insight that the binding might be specific for the best 1 and 2 aptamers, not for aptamers in the random pool.
Next, we aimed to identify the aptamer concentration that is sensitive to toxin concentration changes, which is suitable for our detection toolbox. The test of specificity between random and best groups is incorporated repetitively, too. Hence, we construct a concentration gradient of aptamer with fixed toxin concentration immobilized in magnetic beads for both random and best 2 ssDNA(the selected aptamer that is going to be used at the hardware part). The aptamer attaches to the beads is identified by bead elution and PCR.
Fig.2
Test on the suitable concentration for an aptamer to bind to amanitin. The lanes above are respectively:
DNA marker, the PCR product from beads elution in concentration gradient experiment of ssDNA (random ssDNA with concentration from 800nm to 80fm,
decreasing by 10 times between each group, and Best2 aptamer from 800nm to 80fM, decreasing by 10 times between each group) against amanitin.
From the graph above, we can first see a trend within the group, which is a decreased fluorescence as the aptamer concentration decreases, indicating that the successful binding between the aptamer and its target is dependent on the aptamer concentration. Moreover, a huge brightness shift can be seen from a concentration of 800nM to 80nM for random ssDNA, and the band after 80nM is only pure primer dimer. However, this shift of brightness is delayed in the case of the best 2 aptamers to range from 80nM to 8nM. This implicates a higher efficiency of binding between selected aptamer compared to random, verifying the specificity.
The gel shift is inspiring and convenient, yet the difference in fluorescence is only observed by the naked eye without quantification. Thus we attempted different methods to allow quantification and increase the strength of cumulative evidence on the specific binding between aptamer best 2 and alpha amatoxin. We found a research work on another aptamer developed for alpha amatoxin, in which the researchers characterize the aptamer through 2 methods based on the principle of ELISA for antibodies, ELONA, and Dot blot [5].
By firstly using the aptamer mentioned in the literature, we attempt to test the feasibility of ELONA and repeat the work done. We incubated the amatoxin overnight at 96 well plates. Then the surface is blocked by skimmed milk, following by the addition of biotin-labeled H06 aptamer (selected by the researchers). Then, streptavidin HRP conjugate is added, followed by the addition of TMB substrate. Washing is conducted every round to eliminate the unbounded aptamers.
Fig.3 Results of the ELONA on H06.
Fig.4 Results of the ELONA on Best1.
After the addition of stopping buffer to stop the enzymatic reaction, a yellow color is formed based on the amount of HRP, which is based on the amount of biotin-labeled aptamer. The absorbance at 450nm demonstrated that the color for alpha amanitin is 0.308, which is significantly higher than the control groups below 0.1. The aptamer H06 should be binding amatoxin specifically and the method based on the ELISA principle is feasible. Details of experiments on verification of binding are explained on our Measurement page.
Visit our Measurement Page >
Then, we attempt Dot Blot based on the exact same principle for our Best 2 aptamer, in which the only difference is the immobilization of toxin on NC membrane (which resembles more our hardware) and the reporting color is blue.
Fig.5 Results of Dot Blot on Best 2.
Again, the binding specificity is verified as the blue color for only alpha amatoxin is visible by the naked eye, compared to controls. At this point, we can draw a conclusion and characterize our best 2 aptamers for future detection.
After the thorough examination and characterization of the existed aptamer for alpha aptamer, we also planned to select a new aptamer for beta amatoxin based on magnetic beads SELEX, give that some mushroom contains poisonous beta amatoxin only and the aptamer found cannot bind beta aptamer specifically. Before commencing SELEX, we notice that the beta amatoxin cannot be immobilized on our carboxyl magnetic beads due to the lack of a primary amine group. According to the chemical conjugation method mentioned in previous literature(10.3390/toxins1007026), we activate the carboxyl group presented in beta amatoxin by EDC and NHS, conjugating it with BSA through the formation of the ester bond. Then, the BSA beta amanitin conjugate can be bound to magnetic beads. We then measure the success of this conjugation by the absorbance curve in Nanodrop and agarose electrophoresis for protein(which has higher resolution in comparing slight weight charge change of proteins, though with a complex protocol compare to SDS PAGE). Those methods are found in previous research. [6]
Fig.6 Test on the conjugation of β-amanitin and BSA.
We set the range of absorbance between 250nm and 450nm, with an interval of 1nm for 4 solutions: Beta amatoxin on its own(1), BSA-beta conjugate(2), BSA itself dissolved in EDC/NHS solution(3), and a simple mixture of BSA and beta amatoxin(4), all with the same concentration. Consistent with previous experiments, the bound of BSA and beta amatoxin(in orange) increases the peak of BSA (in blue), which is lower than the peak of amatoxin it alone, indicating the binding.
The result of protein agarose electrophoresis is not that significant, but also mentionable.
Fig.7 Our protein agarose electrophoresis.
Due to the specific orientation requirements of this test, the bottom of the image represents the starting point of the sample. Coherent with the current experiment, the conjugate is transferring slightly faster due to the conformational change of BSA. However, this change is small, probably owing to the size of beta amatoxin, in which 8 amino acids cannot bring about significant charge change of the sample.
Up to this point, the preparation for magnetic beads SELEX [7] is done. We plan to conduct this iterative process for 15 rounds, gradually increasing the strictness of selecting by decreasing the incubation time of toxin and aptamer every round. It is noticeable that the first four round is mainly for counterselection, eliminating the aptamer sequences that bind only BSA magnetic beads. The product of each round of SELEX is monitored by control PCR, a PCR from the product in elucidating of BEAMing (see design). All of the control PCR indicates a correct sequence amplified with acceptable specificity of amplification, and we select one of them for demonstration here.
Fig.8 One of the control PCR experiments (2&4) of SELEX.
The resultant product of each round is about 80bp, as we expected, and the band is bright. As the SELEX progresses, the band seems to be broader, indicating a nonspecific amplification, which might be owing to the increase in the concentration of aptamer with the nonspecific binding site of primer, due to its affinity with the toxin on beads. We thus increase the revealing temperature and the band becomes sharp again. At the last round(round 15), we used all beaming elucidate from round 14 to prepare 2 samples of aptamer pool, 15-1 & 15-2, to ensure an unbiased resultant aptamer pool.
Terminating the process, we found out that the sequence of the aptamer is too short (80bp) to be sequenced. To overcome this dilemma, we purify the PCR product from round 15 and attempts to insert that ssDNA into plasmid PSB1C3 by Gibson assembly.
Fig.9 Colony PCR of our bacteria that carries the sequence of the aptamer through Gibson Assembly.
However, as we can tell from the graph, the plasmid digestion is not effective so that sample in lane 15 and 16 might be the template plasmid directly transferred to bacteria BL21. Then, we also found out even those ones with seemingly correct length have an incorrect sequencing result, indicating that the 80bp aptamer is not assembled to the plasmid and the length is the pure plasmid part being amplified. We attribute this failure as a low assembling efficiency of Gibson assembly specific to a short DNA sequence, hence we assemble the plasmid and aptamer by an alternative method, golden gate.
Fig.10 Colony PCR of our bacteria that carries the sequence of the aptamer through Golden Gate.
All of our sample from golden gate assembly is in the correct length, and the sequencing result indicates successful assembly for all. Here we put our insight into the sequencing result and it's inspiring that one single sequence occurs repetitively among 15 samples of bacteria PCR, 5 times. This implies a possible high concentration of the sequence after PCR, indicating a possibly high affinity to the toxin during fishing.
We ordered the aptamers from DNA synthesis companies based on the sequences we obtained from sequencing and carried out experiments to verify the specificity of each aptamer towards three targets: beta-amanitin, beta-amanitin-BSA conjugate, and BSA. (The experiments and measurements are described on the measurement page)
We found the following two aptamers which demonstrated binding specificity through ELONA.
Aptamer 1: 5'-CATGCTTCCCCAGGGAGATGTAGCGTCTGAAGCCGTTTCATGCATTGCTACAACCCATGAGAGGAACATGCGTCGCAAAC-3'
Aptamer 2: 5'-CATGCTTCCCCAGGGAGATGGCCCGGGGTAACGTAGCCAGATTAGGTCGTGATCGTGATGGAGGAACATGCGTCGCAAAC-3'
Fig.10 Colony PCR of our bacteria that carries the sequence of the aptamer through Golden Gate.
After the characterization for the alpha aptamers in previous literature and the beta aptamers we screened from the SELEX process, we then focused on the construction of our hardware. The overall design was inspired by previous work on LFIA (lateral flow immunoassay) and ALFA (aptamer lateral flow assay) [8], [9]
Fig.11 A concept model of our hardware.
As mentioned on the design page, the aptamers on the test strip competitively binds to the toxin in the liquid extracted from the mushroom sample and the toxin immobilized on the immobilization pad. The aptamers bound to the immobilized toxin then amplified by RPA and detected by Au particle reporting through AFLA. Hence, the concept of hardware should be verified by 3 aspects: Feasibility of RPA, competitive assay, and Au nanoparticle reporting by binding of DNA complementary strand.
Firstly, we examine the property of the RPA kit we acquired by amplification of best 1 and best 2 aptamers for alpha amatoxin.
Fig.12 The verification of RPA kit.
Best 1 and 2 in lane 2 and 3 respectively, is successfully amplified, though the width of the band indicates a dragging force of impurities on target DNA amplicons since the RPA amplicon is not completely transparent. All controls (3,6,10) have no band, indicating the absence of primer dimer. Thus, RPA can be optimized by finding better ways to clear recombinase on amplicons.
Secondly, the competition between free toxin in excess and toxin immobilized on magnetic beads is conducted and the product is amplified by PCR for higher stability in the lab. The specificity of alpha-toxin attracting aptamer which is then amplified is achieved.
Fig.13 Test on concentration gradient(800nM-0.8nM) of aptamer
Here in couplets from lane 2, there's a decrease of aptamer concentration 800nM, 80nM, 8nM and 0.8nM. The second line of each couplet is the one using alpha beads and the first is using BSA beads. For high concentrations like 800nM and 80nM, the specificity of aptamer binding with alpha amanitin is demonstrated by a significantly brighter band. However, as aptamer concentration decreases, the band difference is not significant since there's too little aptamer to bind either bead.
Apart from verifying the binding specificity, this experiment is also inspiring as we can see a trend of product concentration (fluorescence) through the concentration gradient of the aptamer. Hence, we aimed to find the most sensitive aptamer concentration of competitive assay, with a narrow dynamic range and successful aptamer draining by the free toxin. We examined the concentration around 600-800nM which is calculated by the model and demonstrated in the specificity test.
Fig.14 First competitive test between immobilised amanitin and free amanitin with a concentration gradient(800nM-100nM) of aptamer.
The concentration at 200nM and 400nM indicates that the aptamer concentration is too low, so the band is similar to the negative control of primer dimer in lane 11. At 800nM, the trend between competitive assay and non-competitive (i.e no free toxin) is obvious, which means the free toxin can attract a large quantity of aptamer and verify the competitive assay concept. However, the 2 bands at 600nM are abnormally bright, so we repeat the experiment with concentration centered at 800nM to confirm this result.
Fig.15 First competitive test between immobilised amanitin and free amanitin with a concentration gradient(1000nM-600nM) of aptamer.
Again, the difference in brightness is most significant at 800nM, and the 1000nM result suggests an excess of aptamer which cannot be drained by the free toxin. The 600nM result is normal here, without significant difference. Thus, we successfully identify the sensitive concentration of aptamer in our hardware, that is, 800nM.
Concerning safety issues and other uncertainties, we determined to firstly adopt another traditional aptamer target match to prove our pathway of signal amplification. We utilized thrombin and its aptamer of 29 bases with proved affinity by convergent evidence, and we amplify the bound aptamer to thrombin on magnetic beads by PCR(polymerase chain reaction). Then we attempted to further decrease the limit of detection by reducing the aptamer concentration, based on our conclusion of model 2. The principle is verified by a significant decrease of LOD, characterized by the difference of electrophoresis band at 100 times lower concentration (13.5uM) compared to the original concentration (1.35mM).
Fig.16 Competitive test between immobilised thrombin and free thrombin with concentration gradients of both aptamer and free thrombin.
Later on, the Au nanoparticle reporting on both test line with the complementary strand to RPA amplicon (another tail complementary to Au particle) and control line complementary to the reporter probe on Au is done. We immobilized amatoxin on NC membrane by mixing with CBS and a positive result is obtained for alpha amatoxin, specifically.
Fig.17 Test on immobilization of amanitins on NC membrane.
scFv Experiments
1. Construction Of Expression Vector
According to Dr. Dong Wei's work [1], we have successfully constructed the vectors that express the modified anti-α-amanitin scFv efficiently. The sequence of scFv was acquired in the article, and we have optimized the codon for E.coli expression. We have constructed a series of strains including BL21(DE3)-pET28b, BL21(DE3)-pET28b-scFv-6*His Tag, BL21(DE3)-pET28b-pelB-scFv-6*His tag and BL21(DE3)-pET28b-sfGFP-His Tag.
Fig.1 The plasmid map of pET28b-pelB-scFv-HisTag.
Fig.2 The plasmid map of pET28b-scFv-His tag.
Verification Of scFv Expression
As this protein is hard to express because the disulphide bond it has, and the need of being secreted into the periplasm and then culture supernatant or it will aggregate and form inclusion body. To check whether we have correctly constructed the scFv expression vector and to see how properly it can work, we induced the expression strain by IPTG to promote the transcription and protein production. We have cultivated the BL21(DE3)-pET28b-pelB-scFv and BL21(DE3)-pET28b-scFv and induced it for verification of scFv production. We used a competitive anti-His Tag colloidal gold test strip and SDS-PAGE to test if the protein exists in our bacterial pellet and supernatant.
Fig.3 the SDS-PAGE showing the expression of scFv in BL21(DE3)-pET28b-pelB-scFv.
Fig.4 The colloidial gold test strip showing that the scFv expressed in BL21(DE3)-pET28b-pelB-scFv.
This result clearly showed that there are clear strong bands on the gel at the correct size, and also a positive result of colloidal gold test strip, which means there is scFv exist in the culture.
Optimization of scFv expression culture
The scFv production has a relatively low productivity which isn't good enough for further development. So we decided to improve the expression condition. According to Dr.Dong Wei [1] and Dr.Melvyn Little's [2] work, they have given out a method to improve the productivity of scFv in Escherichia coli. We have changed the procedure to induce bacteria, and optimized the cultivation conditions such as culture medium, induction time and temperature. The gel showed that the 2YTS medium can greatly increase the protein that secreted into the medium in the soluble form which is exactly what we want. However, the gel below showed there is still a large amount of protein left inside the pellet, so we need to develop a new method to extract the protein inside the pellet.
Fig.5 The SDS-PAGE showing scFv production in different culture mediums.
Fig.6 The SDS-PAGE showing the periplasm extract after arginine extraction.
Extraction of scFv
We have used various methods to extract the protein that secreted into the periplasm and medium. We have used ultrafiltration, salting out, and TCA (trichloroacetic acid) precipitation to collect the protein inside the medium, and we find that there is a lot of scFv preserved in the culture medium. As salting out is a gentle method, we find that it has the best effect in collecting protein inside the medium. TCA precipitation works best when collecting a small amount of protein in medium for verification. When extracting the protein inside the periplasm, we have used ultrasonication, arginine extraction and osmotic shock. We find that the arginine extraction and osmotic shock can mostly extract the protein in the periplasm without reducing the activity of the protein. On the other hand, we find that ultrasonication is very unstable and the effect of it is not satisfying. We recommend Arginine extraction and osmotic shock for extraction of protein in periplasm. Also, to best utilize the inclusion body, which contains lots of target protein, we tried to solubilize the inclusion body in vitro. So we use urea as the denaturant to denature and solubilize the inclusion body.
Fig.7 The SDS-PAGE showing the extract by arginine extraction and solubilization of inclusion bodies.
Fig.8 The SDS-PAGE showing the extract of osmotic shock.
Verification of scFv
From previous experiments, we have successfully obtained the total protein extract, so the next step is to purify it to control the exact tilted of the scFv. As when the scFv has a his tag, thus we conducted Ni metal affinity chromatography to the cell extract. However, as His tag may be screened by the protein, the affinity between the protein and resin isn't very strong, so it will be hard to purify our scFv. We are suspicious that there are interference chemicals exist inside the arginine extraction buffers and cell pellet. So we use His tagged GFP to verify the purification process. However, we successfully obtained two purification results, but the amount of protein is too low, and the imidazole contained in the solution will influence the concentration measurement, so we cannot know how exactly the protein is.
Fig.9 The SDS-PAGE showing the purification of his tagged sfGFP by improved protocol.
Fig.10 The SDS-PAGE showing the purification of osmotic shock exract.
Fig.11 The SDS-PAGE showing another purification of Osmotic shock product.
ELISA
As there is not enough epitomes in amatoxins, we have to use competitive detection to detect the toxin in samples. However, the IMAC purification have not produced enough scFv for further development, so we used the cell lysate to use as the antibody and successfully verified the binding of scFv and α-amanitin. Also, we have used ic-ELISA to calculate the IC50 (the toxin concentration for 50% inhibition) of our scFv. The lower the IC50, the higher the binding sensitivity. The result successfully quantified the toxin concentration in the sample, and the IC50 of 13.27ng/ml is significantly lower than Dr. Dong Wei's work.
Fig.12 The result showing scFv ELISA when toxin is present and not present.
Fig.13 The result showing that scFv can quantify the toxin in the samples.
Engineering Success
Click the following link to view our Engineering Success page:
Engineering Success >
Contribution
Click the following link to know more about our contribution to the future iGEM teams and the field of synthetic biology:
Contribution >
References
1. Zhang X, He K, Zhao R, Feng T, Wei D. Development of a Single Chain Variable Fragment Antibody and Application as Amatoxin Recognition Molecule in Surface Plasmon Resonance Sensors. Food Anal Methods. 2016. doi:10.1007/s12161-016-0509-3
2. Kipriyanov S.M. (2002) Bacterial Expression, Purification, and Characterization of Single-Chain Antibodies. In: Walker J.M. (eds) The Protein Protocols Handbook. Springer Protocols Handbooks. Humana Press. https://doi.org/10.1385/1-59259-169-8:1035
3. Muszyńska K, Ostrowska D, Bartnicki F, et al. Selection and analysis of a DNA aptamer binding α-amanitin from Amanita phalloides. Acta Biochim Pol. 2017;64(3):401-406. doi:10.18388/abp.2017_1615
4. Koussa MA, Halvorsen K, Ward A, Wong WP. DNA nanoswitches: A quantitative platform for gel-based biomolecular interaction analysis. Nat Methods. 2015;12(2):123-126. doi:10.1038/nmeth.3209
5. Han Q, Xia X, Jing L, et al. Selection and characterization of DNA aptamer specially targeting α-amanitin in wild mushrooms. SDRP J Food Sci Technol. 2018;3(6):497-508. doi:10.25177/JFST.3.6.2
6. He K, Mao Q, Zang X, Zhang Y, Li H, Zhang D. Production of a broad-specificity monoclonal antibody and application as a receptor to detection amatoxins in mushroom. Biologicals. 2017;49:57-61. doi:10.1016/j.biologicals.2017.06.008
7. Hu T, Wessels H, Fischer C, et al. Aptamers Using Magnetic Separation and BEAMing. Anal Chem. 2014;86:10940-10947. doi:10.1021/ac503261b
8. Bever CS, Barnych B, Hnasko R, Cheng LW, Stanker LH. A new conjugation method used for the development of an immunoassay for the detection of amanitin, a deadly mushroom toxin. Toxins (Basel). 2018;10(7):11. doi:10.3390/toxins10070265
9. Jauset-Rubio M, Svobodová M, Mairal T, et al. Aptamer Lateral Flow Assays for Ultrasensitive Detection of β-Conglutin Combining Recombinase Polymerase Amplification and Tailed Primers. Anal Chem. 2016;88(21):10701-10709. doi:10.1021/acs.analchem.6b03256