Summary
We developed a fiber-optic-based ultra-sensitive biosensor for the rapid detection of three OPXVs proteins in the environment. Our results demonstrated the following:
- The insertion of B5 (1st sushi and all 4 sushi ectodomains) into the pET23a plasmid
- No observed synthesis of B5 protein
- Successful expression of L1, A27, A33 proteins
- Functionalization of anti-L1, anti-A27, Anti-33 antibodies on optic fiber
- Detection of L1, A27 and A33 antigens by our optic-fiber biosensor in PBS and sewage water.
Molecular cloning
Amplification
Figure 1 shows the picture of agarose gel of PCR products encoding 1st and 4 sushi domains of B5R protein. The forward primer was the same for sequences encoding both 1st sushi and 4 sushi domains, while different specific reverse primers were used for each of these. Gel electrophoresis provided confirmation of the successful amplification of genes encoding 1st sushi domain of protein B5R (A) and all 4 sushi domains of B5R protein (B). Simulation run on Snapgene corresponds to the results obtained in the lab: 290 bp of PCR product encoding 1st sushi and 756 bp of 4th sushi of B5R protein.
Figure 1 Results for gel electrophoresis of PCR products A: Amplified PCR product encoding 1st sushi domain of B5R protein B: Amplified PCR product encoding all 4 sushi domains of B5R protein. On the right, the simulation of gel electrophoresis by SnapGene.
Restriction digestion of pET23a plasmid was performed using restriction enzymes NdeI and XhoI. Figure 2 gel electrophoresis shows the successfully digested pET23a plasmid, as lighter fragments have uniformly separated from the double-digested linearized pET23a plasmid.
Both sequences encoding 1st sushi domain and 4 sushi domains were double-digested using restriction enzymes NdeI and XhoI, after which PCR purification of both restriction digestion products was performed.
Figure 2 Results for gel electrophoresis of pET23a plasmid double-digested by NdeI and XhoI A - H: Products of restriction digestion of pET23a plasmid. On the right, the simulation of gel electrophoresis by SnapGene.
E. coli DH5 cells were transformed with the ligation products of double-digested pET23a plasmid and sequences encoding 1st sushi domain (Figure 3A) and 4 sushi domains (Figure 3B). For the positive control, cells were transformed with the pUC control plasmid (Figure 3C). For the negative control, the cells were not transformed prior to plating (Figure 3D). The cells were grown in a selective medium of Luria Bertani agar and Ampicillin (100 mg/mL). Plasmid extraction was performed to amplify the number of ligation products, which are later to be used to transform E. coli BL-21(DE3) cells for the expression of protein B5R.
Figure 3 Transformation of E. coli DH5α cells with ligation products: A) E. coli DH5α cells transformed with ligation product of pET23a plasmid and sequence encoding 1st sushi domain B) E. coli DH5α cells transformed with ligation product pET23a plasmid and sequence encoding 4 sushi domains C) Positive control: E. coli DH5α cells transformed with pUC control plasmid only D) Negative control: Non-transformed E. coli DH5α cells.
Colony PCR was performed to screen bacterial colonies for a successful construct. Colonies were obtained from plates with E. coli DH5α transformed with corresponding ligation products. Since we are using the T7-based expression vector system, forward and reverse T7 primers were used in the Colony PCR. Colony PCR products of pET23a ligated with the sequence encoding either 1st sushi domain (Figure 4D, E) or 4 sushi domains (Figure 4F, G) travel less distance due to the presence of T7 primers, and therefore appear higher in the gel than control runs (Figure 4A, B). Therefore, both E. coli DH5α colonies (pET23a + 1st sushi domain) and only the first E. coli DH5α colony (pET23a + 4 sushi domains) have been identified for successful ligation products. The F colony PCR product was sent to sequencing.
Figure 4 Results for gel electrophoresis of colony PCR products
A: PCR product encoding 1st sushi domain of B5R protein
B: PCR product encoding all 4 sushi domains of B5R protein
C: pET23a plasmid used as a control
D: Colony PCR product of pET23a + 1st sushi domain sequence
E: Colony PCR product of pET23a + 1st sushi domain
F: Colony PCR product of pET23a + 4 sushi domains
G: Colony PCR product of pET23a + 1st sushi domains
Colony PCR product of pET23a ligated with the sequence encoding 1st sushi domain was repeated for better results. According to Figure 5, the colony PCR products in wells B, C, D, and F appear higher than the control sample (A). The colony PCR products of pET23a ligated with sequence encoding 1st sushi domain obtained from the successful three (B, C, D) colonies were sent to DNA sequencing analysis.
Figure 5 Results for gel electrophoresis of colony PCR products A: PCR-amplified and purified nucleic acid sequence encoding 1st sushi domain of B5R protein span class="c1">B - F: Colony PCR product of pET23a + 1st sushi domain (Figure 6A)
E. coli BL-21(DE3) cells were transformed with the amplified ligation products of double-digested pET23a plasmid and sequences encoding 1st sushi domain (Figure 6A) and 4 sushi domains (Figure 6B). For the positive control, cells were transformed with the pUC control plasmid (Figure 6C). For the negative control, the cells were not transformed prior to plating (Figure 6D). The cells were grown in a selective medium of Luria Bertani agar and Ampicillin (100 mg/mL).
Figure 6 Transformation of E. coli BL-21(DE3) cells with ligation products: A) E. coli BL-21(DE3) cells transformed with ligation product of pET23a plasmid and a sequence encoding 1st sushi domain B) E. coli BL-21 cells transformed with ligation product pET23a plasmid and a sequence encoding 4 sushi domains C) Positive control: E. coli BL-21(DE3) cells transformed with pUC control plasmid only D) Negative control: Non-transformed E. coli BL-21(DE3) cells
Sequencing
3 samples (reverse and forward for each colony) with 1st sushi domain and 1 sample with 4 sushi domains were sent for DNA sequencing analysis, based on the colony PCR results.
Figure 7 shows the sequencing chromatogram with results from the sequencing run of 1st sushi domain DNA. From the chromatogram, it can be seen that the results are clear, with evenly-spaced peaks and no significant baseline noise.
Figure 7. Part of the sequencing Chromatogram of reversed 1st sushi domain from colony A.
The sequencing results were checked for alignment with the original plasmid DNA for the presence of 1st sushi and all 4 sushi domains’ parts in the colony's DNA for further plasmid extraction. As can be seen from Figure 8 all samples have aligned with the original DNA sequences encoding 1st sushi and all 4 sushi domains. The sequencing chromatogram confirms the proper cloning of 1st sushi and all 4 sushi domains into the pET23a plasmid.
Figure 8. Sequence alignment with plasmid DNA for 1st sushi and all 4 sushi domains. Image created in SnapGene.
Protein expression
SDS-PAGE electrophoresis
For the test expression was checked 48 different conditions for the negative control, 1st sushi domain, and all 4 sushi domains in order to determine the optimal: incubation time (2h, 4h, 6h, and overnight (ON)); temperature (19°C, 37°C, 37°C); IPTG concentration (0mM, 0.2mM, 0.5mM, 1mM) (Figure 9). After expression, bacteria were centrifuged and pellets were resuspended in a lysis buffer with freeze-thaw cycles, followed by homogenization. The presence of B5 was tested using Coomassie blue 15% SDS-PAGE.
The mass of the 1st sushi domain of B5 protein is ~6 kDa. Despite the positive results from sequencing and a wide range of test conditions, the expression of the protein was not observed (Figure 9). Smaller IPTG concentrations (0 mM and 0.2 mM) were not checked due to lack of time, but it is predicted that there is less probability of protein expression without IPTG induction.
Figure 9. SDS-PAGE analysis of B5 1st sushi domain protein. A)19°C. Lane 1- 2h and 0.5 mM IPTG. Lane 2- 4h and 0.5 mM IPTG. Lane 3-6h and 0.5 mM IPTG. Lane 4- ON and 0.5 mM IPTG. Lane 5- absent, Lane 6- 2h and 1 mM IPTG. Lane 7- 4h and 1 mM IPTG. Lane 8-6h and 1 mM IPTG. Lane 9- ON and 1 mM IPTG. B) same lanes but at 30°C. C) same lanes but at 37°C, Lane 5- control
The mass of all 4 sushi domains of B5 protein is ~25 kDa. The expression of these domains also was not observed (Figure 10). The possible reasons for the problem with expression might include the following: the expression system is not working (promoter could be defective); occurrence or absence of the stop codon; problems with protein folding in the bacterial cytoplasm.
Figure 10. SDS-PAGE analysis of B5 all 4 sushi domains protein. A)19°C. Lane 1- 2h and 0.5 mM IPTG. Lane 2- 4h and 0.5 mM IPTG. Lane 3-6h and 0.5 mM IPTG. Lane 4- ON and 0.5 mM IPTG. Lane 5- absent, Lane 6- 2h and 1 mM IPTG. Lane 7- 4h and 1 mM IPTG. Lane 8-6h and 1 mM IPTG. Lane 9- ON and 1 mM IPTG. B) same Lanes but at 30°C. Lane 5- control C) same lanes but at 37°C.
In contrast to the B5 protein, the other 3 proteins were successfully expressed using the same methods. An example of the successful expression of the L1 protein can be observed in Figure 11, the mass of the L1 is ~27 kDa and the highly concentrated accumulation of the protein near this mass can be seen in lanes 2 and 8 in red boxes.
Figure 11SDS-PAGE analysis of L1 protein. IPTG concentration- 1 mM, temperature 30°C, 2h of incubation
Western blot
The Western blot was conducted for several IPTG-induced samples of B5 1st sushi and all 4 sushi domains. Despite the small-expression shown on the SDS-PAGE, the western blot showed that the sample cells have not expressed the required parts of the B5 protein - 1st sushi and all 4 sushi domains. The expression was successful for the other 3 recombinant proteins A27, A33, and L1. Figure 12, 13, and 14 represent the western blot of samples taken from SDS-PAGE analysis for L1, A27, and A33 proteins, respectively. Results from Western blots indicate that proteins L1, A27 and A33 were properly folded and successfully bound to their antibodies, thus antibodies can be used for biosensor functionalization.
Figure 12. Western blot analysis of L1 protein
Figure 13 Western blot analysis of A27 protein
Figure 14 Western blot analysis of A33 protein
Aptamer design
We created aptamer sequences for L1, A27, and A33 proteins using an open-source software MAWS:
Table 1. DNA aptamer sequences for L1, A27, and A33
|
Protein |
DNA aptamer sequence |
Aptamer ID |
|
L1 |
ATCGTGAGGAAGCGGCGGGA |
L1_1 |
|
|
ATCGTGAGGAAGCG |
L1_2 |
|
|
GCGGGGGGGTGCGAAGGGG |
L1_3 |
|
|
GCGGGGGGGTGCGAA |
L1_4 |
|
|
ATCGTGAGGAAGCGGCGGGATTTGG |
L1_5 |
|
A27 |
GGAAGGGGGGTGGCCGGCGA |
A27_1 |
|
|
GTAGTCGGGGTAGCCGGGAG |
A27_2 |
|
|
GGTAGTAAGGCGGGGATAGA |
A27_3 |
|
|
GGTAGTAAGGCGGG |
A27_4 |
|
|
GGAAGGGGGGTGGCC |
A27_5 |
|
A33 |
GAGGGGCTGGGGTTGGTTTG |
A33_1 |
|
|
GAGGAGGGACTGGGGGGGAG |
A33_2 |
|
|
GUGGGCAGGGUAAA |
A33_3 |
|
|
GTGGGCAGGGTAAAG |
A33_4 |
|
|
GAGGAGGGACTGGGG |
A33_5 |
To make optic fiber biosensors with aptamers, we attached a sequence ‘TTTTT’ to the 5’ of the DNA sequence, so aptamers can bind to the organic layer on the optic fibers.
And as we had limited time, we made 3D structures for aptamers of only L1 (Figure 15 and 16), using a method described in Oliveira et al. and the software dnaTurner that we hacked (see software page for more information).
Figure 153D structure of a DNA aptamer of L1 (ID: L1_1)
Then to check how well they bind with L1, we docked the 3D structures of aptamers and the protein in the HDOCK web server, and used the PLIP web server to analyze the docking model with the least binding energy, and ordered the two aptamers forming the most number of hydrogen bonds for further in-vivo tests (aptamers L1_1 and L1_5).
|
Docking score |
Confidence score |
Ligand rmsd (Å) |
|
-296.11 |
0.9489 |
41.99 |
Figure 16 A docking simulation model for L1 and his aptamer (ID: L1_1)
Table 2. Data of aptamer-L1 interactions
|
Aptamer ID |
Docking Confidence Score (via HDOCK) |
Number of hydrogen bonds in a docking simulation (via PLIP) |
|
L1_1 |
0.9489 |
15 |
|
L1_2 |
0.8577 |
7 |
|
L1_3 |
0.8806 |
12 |
|
L1_4 |
0.8286 |
5 |
|
L1_5 |
0.8493 |
20 |
Testing optic fiber biosensors on antigen detection
Overview
We conducted tests with two types of optic fiber biosensors: biosensors functionalized with single antibodies and multiple antibodies (multiplex biosensors). Biosensors with single antibodies have been tested for detection of their antigen in PBS solution, while the multiplex sensors were tested both in PBS and sewage water.
Biosensors with single antibodies
Biosensors, made via functionalizing optic fiber ball resonators with single antibodies, successfully responded to the presence and change of concentrations of target antigens up to attomolar concentrations and proved specific to their target antigens.
We functionalized 3 optic fiber ball resonators with antibodies of L1, A27, and A33 proteins and used one optic fiber ball resonator functionalized without antibodies as a control, and immersed them in a solution with the proteins L1, A27 and A33 dissolved in PBS. We started at the 10 attoMolar concentration of proteins and increased by the power of 10, up to 10 nanomolar concentration. The sensor responses were recorded 10 minutes of immersing in a solution of concentration, and the measurements for one sensor were consecutive.
Compared to the control sensors, outputs of the optic fiber biosensors functionalized with single antibodies represented as their amplitudes, change as a function of the concentration of antigens in PBS (Figure 17). The graphs also show the average and standard deviation of three measurements, and the Limit of Detection (3ylowest + δmax).The sensors functionalized with single antibodies showed a logarithmically linear trend with R2 > 0.90, and the range of LOD of sensors to detect vaccinia virus proteins were from 0.1pM to 1fM. The tests have not been tested for repeatability and require further validation, however they show a good potential to detect viral antigens in low concentrations with high confidence.
Figure 17 Response of the optic fibers functionalized with antibodies of (a) L1, (b) A33 and (c) A27, and (d) the control optic fiber as a function of concentrations of L1, A27, and A33 proteins in PBS
Figure 18 illustrates how the response of optic fibers is usually obtained - as amplitudes of the sensor at a range of wavelengths and shows how it changes at different concentrations.
Figure 18Spectra showing the change in the amplitudes of optic fibers functionalized with antibodies of L1 in different concentrations of the antigens L1, A27, and A33
To test for the specificity, sensors functionalized with L1, A33 and A27 were immersed in solutions with high concentrations of the target antigen, and Chemokine ligand 5 (CCl5), Interleukin 4 (IL4) and Thrombin proteins were used as Controls. The sensors show increased responses (shown in terms of Refractive Index Unit,) to their corresponding antigens than to other compounds, indicating their high specificity to the target antigens (Figure 19).
Figure 19Specificity of the optic fibers functionalized in detecting the target protein. a- biosensor functionalized with anti-L1 antibodies, b- biosensor functionalized with anti-A27 antibodies, c- biosensor functionalized with anti-A33 antibodies.
Testing Multiplex biosensors in PBS and sewage water
Multiplex biosensors were observed to respond to the presence of antigens and their change in concentration both in PBS and in sewage water samples.
The optic fiber ball resonators were functionalized with three proteins simultaneously, making a biosensor that can detect several antigens, and then were tested with the same method in PBS and sewage water. The tests in PBS were conducted to observe if the antibodies on the surface of the sensors bind to the antigens in solutions where external interference is minimized. Then the sensors were tested in synthetic sewage water samples to detect target proteins in a more challenging and interference-rich environment, to test the capabilities of the sensor to work in medium-rich environments.
Figure 20Response of (a) the multiplex sensor and (b) the control optic fiber, in solutions of different concentrations of L1, A27, and A33 proteins in PBS
Figure 21Amplitude change of multiplex sensors (a and b) and control sensor (c) as a function of the concentration of the antigens in sewage water
The multiplex sensors showed a log-linear trend with R2 > 0.92 in PBS and R2 > 0.88 in sewage water samples (Figures 20 and 21). The Limit of Detection for multiplex sensors were in the range of 1fM and 1pM in sewage water samples, indicating that the multiplex sensors can detect target antigens in low concentrations with high confidence (Figure 21).
Figure 22Spectra showing the change in the amplitude of the multiplex sensor to different concentrations of the antigens L1, A17, and A33
Figure 23Spectra showing the change in amplitude of a multiplex optic fiber biosensor to different concentrations of antigens in sewage water
Overall, the optical fiber ball resonators functionalized with single antibodies were able to detect vaccinia virus proteins in low concentrations with high specificity in PBS. The multiplex biosensors have shown the ability to detect vaccinia virus proteins in low concentrations both in PBS and sewage water samples.
For a more thorough interpretation of the results, please see the Proof of Concept page.
Successes and failures
Successes
- Designed B5 1st and all 4 sushi constructs
- Expressed L1, A27, A33 properly folded and bound to their antibodies
- Optic fiber biosensors detect the presence of target proteins in PBS solution and sense changes in concentrations of the target proteins.
- Optic fiber biosensors detect the presence of target proteins in sewage water and sense changes in concentrations of the target proteins.
- Optic fiber biosensors showed high sensitivity to target antigens.
- We designed aptamers for L1, A27, and A33 using MAWS software, and modeled and analyzed their interactions in-silico.
Failures
- The B5’s 1st and all 4 sushi domains were not expressed due to either defect expression system, occurrence or absence of stop codon, or improper folding in the cytoplasm.
- Designed aptamers have not been tested in the laboratory for complete verification of their binding with the target proteins.
- In the process of implementing our project, the machine that makes ball resonators from ordinary optic fibers broke, and thus, because we couldn’t make new optic fiber biosensors, we have been limited in the number of sensors and experiments on the sensors we can conduct, which to a noticeable extent constrained the quality of conclusions we can make. For example, it is hard to elaborate on the repeatability of the sensors because we cannot make new sensors to test.
Future tests
- We started the cloning with another plasmid for B5 expression - pet22b with the inserted signal sequence for Sec (secretion system) which enables secretion within the periplasm. However, due to the lack of time and long primer delivery to Kazakhstan, we were not able to start testing for the expression. So, we are planning to test all possible solutions for B5 protein expression.
- We plan to test the multiplex biosensor in real environmental conditions to have reliable results about the duration and stability of the biosensor.
- In the future we plan to order and test aptamers for L1, A27, A33, and B5 (if successfully expressed).
Conclusion
Our multiplex biosensor based on optic-fiber functionalized with antibodies for 3 vaccinia virus proteins (A27, A33, L1) demonstrated the ability to detect target proteins with high specificity, concentration-dependent change and stability of response over time both in PBS and sewage water. We successfully designed aptamers for the desired proteins, but due to a lack of time, we did not have time to test them on a fiber-optic sensor for protein detection. Also, due to a defective or inappropriate plasmid system that we have chosen or due to the toxicity of the protein to BL-21 E.coli, the B5 protein was not synthesized, so the biosensor was not functionalized for this protein. Thus, we need to test whether the insertion of the Sec system into the B5 sequence and the usage of different plasmids will allow the expression of B5.