Difference between revisions of "Team:NYU Abu Dhabi/Results"

Optimised the protocol for RPA to reduce the reaction volume up to 10 μL. Click here for more details.

Optimised PCR, LAMP and RPA reactions to achieve the limit of detection of 0.1 ng/μL. Click here for more details.

Sample collector: Developed sterile, easy to use sample collection mechanism for the end user which may be modifiable for any form of diagnostic test. Click here for more details.

Microfluidic chip: Developed cheap chips with longer shelf life with the ability to test for multiple pathogens (five) at once. Click here for more details.

Visualisation: Optimised Blue light to view fluorescence after running the reaction- drastically reducing cost for all diagnostic devices that utilise fluorescence, also ensuring availability of detection for people of all backgrounds. Click here for more details.

The team started with performing PCR for the plasmids as it is a general technique and the results would serve as a standard to compare the results of LAMP and RPA, the other relatively new amplification techniques we used for the project.

We performed PCR, LAMP and RPA reactions to characterize our plasmids and determine if amplification happens with our designed primers. We wanted to test how specific each of these amplification techniques is by running each plasmid with its primers and the primers of other fragments. Our results show that LAMP is the most specific amplification technique which is consistent with results from the literature that show that LAMP has very high specificity (1). PCR had the lowest specificity which is also consistent with results from literature (2). RPA has been shown to be the best amplification technique currently available in terms of parameters like speed, complexity, and user-friendliness. However, our results show that RPA is less specific than LAMP. Daher et al (2015) (3) showed that mismatches can occur if extra precautions are not taken during primer design to eliminate this. In our case, our primers could be the reason for RPA being less specific than LAMP.

We also tested how sensitive LAMP and PCR are by checking for the lowest concentration of DNA past which amplification is lost. We visualized amplification for both techniques by gel electrophoresis. We visualized the results for LAMP alone by adding SYBR green to the reaction post-amplification and visualizing with UV light. The results show that both LAMP and PCR have a similar sensitivity of up to 0.1 ng/μl of DNA. however, results from literature (4)(5) have shown that LAMP is significantly more sensitive than PCR.

As we would be using SYBR green in our device for visualization of amplification, we performed experiments to determine the optimal concentration of SYBR green to be added to our LAMP reaction and the wavelength of UV light that allows for the best visualization. We found that 1000X SYBR green in 25 μl total volume of LAMP reactants visualized with 254 nm UV light gave the best results.

To test if the results obtained from intra-lab amplification using miniprep DNA would work in our device which would use samples of putatively contaminated food or water, we tested detection of lmo0733 gene from Listeria Monocytogenes in beef. Ground beef was spiked with lmo0733 and E. Coli (as control for specificity) and direct swabs of prepared beef samples were used to run a LAMP reaction using NEB WarmStart colorimetric mastermix and lmo0733 primers. We observed a distinct yellow color in reactions with samples spiked with lmo0733 15 minutes after the reaction which confirms amplification; while samples from unspiked beef (negative control) and beef grown with E. Coli remained bright red. Gel electrophoresis was also used to confirm the results of the colorimetric amplification.

PCR

PCR reactions were run with designed primers to confirm that the primers amplify the gene of interest and also characterize the DNA fragments we would be working with. The agarose gel (1%) shows bands at expected lengths for lmo0733 (430 bp), gbpA (1019 bp) and invA (818 bp). The negative controls show no bands i.e. no amplification which is what was expected. PCR reactions were run to test the specificity of the primers for this technique. Each gene was run with its primers and the primers of other gene fragments

Sensitivity (lmo0773, invA and hipO)

In order to determine the PCR reaction sensitivity, the reaction was run with serial dilutions of miniprepped plasmid with the gene of interest. The reactions were set up according to an optimized protocol used in the laboratory.

Figure 1. Agarose gels (1%) corresponding to the PCR reaction with serial dilutions of miniprepped lmo0773, invA and hipO DNA. (a). lmo0773 serial dilutions 363 ng/µl, 200 ng/µl, 100 ng/µl, 50 ng/µl, 25 ng/µl, 10 ng/µl, 1 ng/µl, 0.5 ng/µl, 0.1 ng/µl. (b). invA serial dilutions 295.5 ng/µl, 200 ng/µl, 100 ng/µl, 50 ng/µl, 25 ng/µl, 10 ng/µl, 1 ng/µl, 0.5 ng/µl, 0.1 ng/µl. (c). hipO serial dilutions 159.8 ng/µl, 100 ng/µl, 50 ng/µl, 25 ng/µl, 10 ng/µl, 1 ng/µl, 0.5 ng/µl, 0.1 ng/µl

Results obtained clearly shows that PCR is sensitive up to 0.1 ng/µl, lowest concentration tested, for all three plasmids tested. Visually, all bands appear to be similarly thick, which shows that, despite the changes in concentration, PCR amplified each DNA in a similar manner.

Published research has reported PCR to be sensitive up to 3 pg/µl (6). This shows that PCR is a sensitive technique that is able to amplify DNA even at low DNA concentrations.

Specificity (lmo0773, invA and gbpA)

Two sets of experiments were carried out to test the specificity of each amplification technique used. In the first set of experiments, the genes used were kept constant, while the primers were varied, while in the second set, the primers were kept constant while the genes were varied. As can be seen in the figure below, PCR was not found to be specific, as the DNA for a specific gene was amplified by primers designed for another gene as well as with primers specific for that gene.

Figure 2. Agarose gels (1%) corresponding to PCR specificity reactions carried out on three different genes (a) lmo0733, (b) invA and (c) hipO. The first set of reactions for each gene is done by keeping the gene constant while varying the primers, while the second set of reactions are carried out by varying the gene used while keeping the primers constant.

LAMP

Loop-mediated isothermal amplification was performed using primers designed with PrimerExplorer for lmo0733, invA, hipO and gbpA. The reaction was run using miniprep DNA and transformed E. Coli colonies to assess if amplification can occur with whole cells. The Agarose gel (1%) shows amplification in all the lanes with miniprep DNA.

Figure 3. Agarose gel (1%) showing LAMP amplification of invA, gbpA and lmo0733 miniprep DNA with designed LAMP primers (PrimerExplorer). Amplification is seen for lmo0733 and gbpA but not invA when gene transformed E. Coli colonies were used. (Lane 1) 500 bp ladder; (Lane 2) invA miniprep + invA LAMP primers; (Lane 3) Nuclease-free water + invA LAMP primers; (Lane 4) invA transformed E. Coli colony + invA LAMP primers; (Lane 5) gbpA miniprep + gbpA LAMP primers; (Lane 6) Nuclease-free water + gbpA LAMP primers; (Lane 7) gbpA transformed E. Coli colony + gbpA LAMP primers; (Lane 8) lmo0733 miniprep + lmo0733 LAMP primers; (Lane 9) Nuclease-free water + lmo0733 LAMP primers; (Lane 10) lmo0733 transformed E. Coli colony + lmo0733 LAMP primers.

Figure 4. Agarose gel (1%) showing LAMP amplification of gbpA with non colorimetric reaction mastermix (MM) (Optigene) with either hydroxy naphthol blue (HNB) or SYBR green added and with colorimetric reaction mastermix (NEB). (Lane 1) 500 bp ladder; (Lane 2) gbpA + Optigene MM + gbpA LAMP primers + HNB; (Lane 3) nuclease free water + Optigene MM + gbpA primers + HNB; (Lane 4) gbpA + Optigene MM + gbpA LAMP primers + SYBR green; (Lane 5) Nuclease free water + Optigene MM + gbpA LAMP primers + SYBR green; (Lane 6) gbpA + NEB MM + gbpA LAMP primers; (Lane 7) Nuclease free water + NEB MM + gbpA LAMP primers.

SYBR Green Optimization

SYBR Green was used to visualize amplification of miniprep DNA in the presence of UV light. 1 ul of SYBR Green was added to 25 ul of LAMP reactants (using NEB Master Mix, designed primers, & miniprep DNA for positive controls and water for negative controls). The samples were visualized at different wavelengths of UV light to determine the optimal wavelength for visualization and to optimize the concentration of SYBR Green in 25 ul of LAMP reactants.

Figure 5. Visualization of SYBR green at 302 nm and 365 nm for lmo0733 LAMP reaction.

No fluorescence was detected in the absence of SYBR green. Background fluorescence was observed in the negative controls. A clear distinction was observed between positive and negative controls.

Figure 6. Visualization of SYBR green at 302 nm and 365 nm for invA LAMP reaction

Results obtained matched the experiment performed with lmo0733. 1000X SYBR Green was determined to be the optimal concentration and 365 nm seemed to produce the best images for visualization of LAMP amplification.

SYBR Green Visualization

SYBR Green (1000X) was used to visualize the results of the LAMP reaction as determined by the optimization process detailed above. The gbpA gene was amplified using the LAMP method and fluorescence was detected in the positive sample at UV 254 nm as well as under blue light. The negative control showed background fluorescence, possibly due to the addition of primers or as a result of the SYBR Green itself. However, a clear distinction in fluorescence was observed.

Figure 7. Visualization of gbpA LAMP reaction with SYBR Green under UV (254 nm) and Blue Light

Colorimetric Visualization

NEB WarmStart Colorimetric Master Mix was used as a colorimetric dye to detect color change under visual light when the samples were amplified using the LAMP technique. The originally pink colored mixture turned yellow as a result of the amplification.

Figure 8. Colorimetric results from WarmStart Colorimetric Master Mix reactions with invA gene.

Colorimetric Test: Real Sample Swab and Amplification

To test the working principles of the Pathogene pathogen project, it had to be established that the intra-lab amplification techniques would be effective on real world samples of contaminated food, water, surfaces etc. Two samples of beef were prepared for the purposes of determining the colorimetric visualization of results for LAMP, in particular the use of NEB WarmStart Colorimetric Mastermix. A sample of untreated store-bought beef was prepared alongside a sample of DH5-alpha cells transformed with the lmo0733 gene from Listeria Monocytogenes. Direct swabs were taken from each sample and used in the LAMP reactions. Reactions lacking the target gene appear a bright red colour whereas reaction mixes containing the amplified gene appear salmon to yellow in colour.

Figure 9. Colorimetric results from WarmStart Colorimetric Master Mix reactions immediately after extraction from thermal cycler.

Figure 10. Colorimetric results from WarmStart Colorimetric Master Mix reactions 15 minutes after extraction from thermal cycler.

Figure 11. Agarose gel (1%) (left to right) : (a) 500bp ladder, WarmStart reaction mix with treated beef sample, WarmStart reaction mix with untreated beef sample, WarmStart reaction mix with nuclease-free water (b) 500bp ladder, Optigene reaction mix with treated beef sample, Optigene reaction mix with untreated beef sample, Optigene reaction mix with nuclease free water.

Both the visual colorimetric results immediately following and 15 minutes after the reaction show the treated sample of beef as having a distinctly lighter colour than the other two unamplified samples. The gel electrophoresis confirms the amplification of the lmo0733 gene from the whole swabbed bacterial cells and no amplification in samples without the target gene, assuring LAMP’s specificity.

LAMP Sensitivity (lmo0773, invA and hipO)

In order to determine the LAMP reaction sensitivity, the reaction was run with serial dilutions of miniprepped plasmid with the gene of interest. The reactions were set up according to Optigene or NEB LAMP kit protocols.

Figure 12. Agarose gel (1%) corresponding to the LAMP reaction with serial dilutions of miniprepped lmo0773, invA and hipO DNA. (a). lmo0773 serial dilutions 363 ng/µl, 200 ng/µl, 100 ng/µl, 50 ng/µl, 25 ng/µl, 10 ng/µl, 1 ng/µl, 0.5 ng/µl, 0.1 ng/µl. (b). invA serial diltioons 295.5 ng/µl, 200 ng/µl, 100 ng/µl, 50 ng/µl, 25 ng/µl, 10 ng/µl, 1 ng/µl, 0.5 ng/µl, 0.1 ng/µl. (c). hipO serial dilutions 172.5 ng/µl, 100 ng/µl, 50 ng/µl, 25 ng/µl, 10 ng/µl, 1 ng/µl, 0.5 ng/µl, 0.1 ng/µl

The sensitivity test corroborated that LAMP is a sensitive technique that can detect the DNA up to very small concentrations. Results obtained show that lmo0773 and hipO plasmids are sensitive up to 0.1 ng/µl, while invA plasmid is sensitive up to 0.5 ng/µl. Visually the amplification is comparably visible for all concentrations for lmo0773 and hipO plasmids, with miniprepped plasmid band being visible up to 25 ng/µl for all plasmids. The invA plasmid seems to not be as sensitive, however, as the literature reports LAMP to be sensitive up to 33 ng/µl, it is very likely that the invA plasmid is an outlier (6). The sensitivity of invA plasmid could have been affected by the improper set up of the reaction e.g. inaccurate serial dilutions, mistakes in the protocol, etc. Therefore, the test showed that LAMP is a good alternative technique, which is comparably sensitive to PCR.

In order to determine if visualization will be possible with the SYBR Green fluorescent dye even at low DNA concentrations, a LAMP reaction was run with hipO plasmid serial dilutions with 1 µl of 1000x SYBR Green dye added into the 25 µl LAMP reaction. Additional negative controls were included, such as SYBR Green + water and SYBR Green, water + hipO primers to determine the background fluorescence.

Figure 13. The SYBR Green fluorescence of the hipO serial dilutions represented under (a) UV light (365 nm) (b) Blue light (c) under Blue light with overexposure demonstrates.

The fluorescence results under UV and Blue light confirmed that the reaction can be visualized up to 0.1 ng/µl. Under UV light you can clearly see the difference between positive and negative controls. There is also a trend of decreasing fluorescence with decreasing plasmid concentration with the highest fluorescence at 172.5 ng/µl and lowest at 0.1 ng/µl. This is corroborated by the reaction vessels under the Blue light, which show the same trend. The overexposure option of visualization under the Blue light allows to show that even at 0.1 ng/µl there is more fluorescence than the background fluorescence present in the negative controls. Therefore, this test shows that even at low concentrations the SYBR Green is effective at showing the successful DNA amplification.

LAMP Specificity (lmo0773, invA and gbpA)

The same two experiments done with PCR were done with LAMP. The results obtained indicate that LAMP is highly specific as every gene was only amplified by its primers and not by any other primers. LAMP was found to be the only completely specific technique out of PCR, LAMP and RPA.

Figure 14. Agarose gels (1%) corresponding to LAMP specificity reactions carried out on two different genes lmo0733 and invA. The first set of reactions for each genes, (a) for lmo0733 and (c) for invA is done by keeping the gene constant while varying the primers, while the second set of reactions, (b) for lmo0733 and (d) for invA are carried out by varying the gene used while keeping the primers constant.

RPA

Reaction Volume Optimization

The original TwistDx RPA Basic kit protocol that is specified for a total reaction volume of 50 ul was optimized for the volumes of 25 ul and 10 ul. The amounts of reagents in the protocol stated for 50 ul were brought down proportionally and the reactions genes were amplified successfully as shown in the agarose gel (3%) images. The reactions were run with miniprepped plasmid with the gene of interest. Optimization of RPA reactions to lower volumes helped save RPA reagents in the experiments as well as in the microfluidic chips used for the amplification of pathogens in the Pathogene device.

Reaction volume: 50 uL reaction

Figure 15. Agarose gel (3%) showing RPA amplification of lmo0733, invA and gbpA miniprep DNA and transformed E. coli colonies in 50 ul volume reactions. The light bands seen in the negative control lanes are primer dimers and proteins from the RPA reaction. (Lane 1) 100 bp ladder; (Lane 2) lmo0733 miniprep + lmo0733 RPA primers; (Lane 3) lmo0733 transformed E. coli colony + lmo0733 RPA primers; (Lane 4) lmo0733 negative control; (Lane 5) invA miniprep + invA RPA primers; (Lane 6) invA transformed E. coli colony + invA RPA primers; (Lane 7) invA negative control; (Lane 8) gbpA miniprep + gbpA RPA primers; (Lane 9) gbpA transformed E. coli colony + gbpA RPA primers; (Lane 10) gbpA negative control; (Lane 11) 500 bp ladder

RPA reactions worked successfully for 50 ul reactions. To save lab reagents in experiments and to make the microfluidic chips economical in terms of resources, further experiments were carried out to test RPA reactions at lower volumes, i.e. 25 ul and 10 ul.

Reaction volume: 25 uL reaction

Figure 16. Agarose gel (3%) showing RPA amplification of invA and gbpA miniprep DNA and negative controls in 25 ul volume reactions. The light bands seen in the negative control lanes are primer dimers and proteins from the RPA reaction. (Lane 1) 100 bp ladder; (Lane 2) invA miniprep + invA RPA primers; (Lane 3) invA negative control; (Lane 4) gbpA miniprep + gbpA RPA primers; (Lane 5) gbpA negative control.

The reaction volume for RPA was successfully optimized to a total volume of 25 ul. This allowed for economical use of reagents in lab experiments and for use in microfluidic chips.

Reaction volume: 10 uL reaction

Figure 17. Agarose gel (3%) showing RPA amplification of gbpA, invA, and lmo0733 miniprep DNA and negative controls in 10 µl volume reactions. The light bands seen in the negative control lanes are primer dimers and proteins from the RPA reaction. (Lane 1) 100 bp ladder; (Lane 2) gbpA miniprep + gbpA RPA primers; (Lane 3) invA miniprep + invA RPA primers; (Lane 4) lmo0733 miniprep + lmo0733 RPA primers; (Lane 5) gbpA negative control; (Lane 6) invA negative control; (Lane 7) lmo0733 negative control.

The reaction volume for RPA was successfully optimized to a total volume of 10 µl. The agarose gel (3%) shows brighter bands for gbpA and lmo0733 compared to invA. This allowed for economical use of reagents in lab experiments and for use in microfluidic chips.

SYBR Green Optimization

SYBR Green was used to visualize amplification of miniprep DNA in the presence of UV light. 1 ul of SYBR Green was added to 25 ul of RPA reactants. The samples were visualized at different wavelengths of UV light to determine the optimal wavelength for visualization and to optimize the concentration of SYBR Green in 25 ul of RPA reactants, as was performed for LAMP.

Figure 18. Visualization of SYBR green at 302 nm and 365 nm for invA RPA reaction

No fluorescence was detected in the absence of SYBR green. Minimal background fluorescence was observed in the negative controls. A clear distinction was observed between positive and negative controls. 1000X SYBR Green was determined to be the optimal concentration and 365 nm seemed to produce the best images for visualization of LAMP amplification.

The Collection Device is a portable, sturdy, and easy to use sample collection device. Its current version provides a small form factor that may be further reduced. Sample collection is done by brushing the sterile cotton swab against a sample. The device contains a sterile watertight chamber for TE buffer solution for safe transportation and use. This chamber is sealed with a thin watertight film that maintains the liquid in the chamber until it is needed. A funnel maintains the film in place through pressure with the use of three screws. Once sample collection occurs, TE buffer is released through the press of a plunger, which the funnel guides into the cotton swab with the sample. The sample is washed out of the cotton by the flow of liquid. A lid may be used to direct the liquid flow directly into the microfluidic chip inlet, making the transition between sample collection and sample preparation seamless.

The sample collector is currently manufactured with 3D printed PLA with the exception of the sterile cotton tip swab and the plunger retrieved from standard 10 ml medical syringes. It uses two O-rings (at the base of the sample collector and funnel) to ensure water tightness. The current unit cost of materials and chemicals is $2.89 USD but can be further reduced with design optimization and mass production.

Figure 1. Final Sample Collector Prototype

The heating device provides a platform that sustains the designated temperature under which the RPA, LAMP reaction should run. There are three modes for the device: in the first mode, the green LED light lights up, signaling that the power is connected. The 6 blue LED lights that aids visualization will also be on. In the second mode (for RPA reaction), one red LED light turns on, and the heating board heats up to 40&deg C. In the third mode (for LAMP reaction), two red LED lights turn on and the heating board heats up to 65&deg C. A temperature sensor is closely attached to the heating board and helps ensure that the temperature is maintained as desired. A power bank that consists of a 9V rechargeable battery is used as the power supply for the device. The user can just connect the power bank to the device for the device to start operating.

Figure 2. Final Heating Device Prototype

The microfluidic chip provides chambers for each LAMP/RPA reaction to occur. The chip is designed to have a circular shape with a small a small inlet in the middle. The inlet is connected by 600um width channels to 10 different reaction wells. DNA will then be amplified in each reaction well to determine the presence of harmful pathogens within the food sample. The chip is made with a hydrophilic film and double-sided tape to reduce cost and ensure that the food sample flows easily to each reaction chamber. The material used for this chip has a significantly lower cost compared to the conventionally used PDMS chip. Compared to the method of using silicon mold used to create PDMS chip, Pathogene’s chip is laser cut also contributing toward low production cost. It also has a longer shelf-life and is more stable compared to paper microfluidic chips. With this chip, users are able to expect reliable results with reduced cost.

Figure 3. Final Microfluidic Chip Prototype