Specificity

While Lambert iGEM has been utilizing rolling circle amplification (RCA) to detect a single isolated microRNA (miRNA), human blood serum contains a total of 204 detectable miRNAs (Wang et al., 2012). Research conducted by Jonstrup et al. in 2006 found that the padlock probe ligates on a perfectly matching RNA template, distinguishing between differences in the target and other sequences. To test whether padlock probes would be able to detect specific miRNA, and therefore be applicable for serum testing, we ran RCA using the hsa-miR-1-3p padlock (BBa_K4245200) in the presence of four different miRNA sequences (see Fig. 1). The first is the original miR-1 sequence (BBa_K4245006), which is expected to hybridize to the padlock and result in the greatest fluorescence decrease. Two sequences with differing single nucleotide variants (SNVs) found from the National Library of Medicine microRNA 1-1 database were utilized to determine the specificity of RCA: one with a single SNV (BBa_K4683003) and one with three SNVs (BBa_K4683004). hsa-miR-133a-3p (BBa_K4245009) was also included to ensure the padlock would not ligate to any miRNA.

Figure 1. Comparison of sequences used to test specificity of hsa-miR-1-3p RCA padlock: 1 bp SNV in the seed region, 3 SNVs outside of seed, and miR-133a-3p

We ran the reactions and control on a gel electrophoresis; only the well with 40.8 pM of miR-1 showed visible bands of DNA near the top of the wells, which is likely our RCP (see Fig. 2) (see Experiments: blueGel™ with RCP).

Figure 2. Gel results: RCA with A: miR-1, B: 1 SNV, C: 3 SNVs, D: miR-133a; 2% agarose gel ran for 1 hour at 48V

We then tested the RCP with linear DNA probes and quantified the resultant fluorescence in a plate reader at an emission wavelength of 480 nm and an excitation wavelength of 528 nm (see Fig. 3) (see Experiments: Linear DNA Probes with RCP. The RCA reaction utilizing the miR-1 padlock probe with miR-1 exhibited significantly less fluorescence than the other miRNAs. Since linear DNA probes produce a negative correlation between fluorescence and miRNA concentration, this result, along with the gel, indicates that RCA is specific to single nucleotide differences.

Figure 3. Comparison of RCA with miR-1, 1SNV, 3SNVs, and 133a fluorescence output using linear DNA probes.

Emory Testing

Dr. Charles Searles is a cardiologist specialist working with biomarkers regarding coronary artery disease (CAD) and an associate professor of medicine at Emory Healthcare Hospital. Last year, he provided us with hsa-miR-451a (BBa_K4245012) - which is unrelated to the pathophysiology of CAD - as a control for our proof of concept experimentation and also offered to test our biosensor in his lab. This year, we continued to communicate with him in order to test whether our biosensor could be practical and applicable as a diagnostic tool. This was done with the help of Kimberly Ann Rooney, a lab technician at the Searles Cardiovascular Lab led by Dr. Charles Searles, who ran the RCA reactions utilizing our protocol and miR-1 padlock probe (BBa_K4245200), and their enzymes (see Fig. 4) (see Experiments: SYBR™ Safe with RCP). They ran RCA on 40.8 pM of miR-1 (BBa_K4245006) with SYBR™ Safe dye, which fluoresces when bound to ssDNA, as the output. Resultant fluoresence was quantified in a plate reader at exciation wavelength of 480 nm and emission intensity at 528 nm. As shown in Figure 4, there was a significant increase in fluorescence in the RCA reaction as compared to that of the controls, therefore validating the application of our biosensor in other labs.

Figure 4. Triplicate of RCA with SYBR™ Safe output done by independent hands

After running our assay twice, Ms. Rooney gave us feedback regarding our protocol such as reducing the number of variables as controls and performing serial dilutions for more precise concentrations of miRNA. Ultimately, she told us that our biosensor lacked practicality and applicability for point-of-care testing due to its long reaction time, which is approximately 15-16 hours‒ more than double the time needed for quantification through qRT-PCR. Subsequently, the next step we took to optimize RCA was to reduce the time required to run the reaction.

Time Optimization

Ligation Time

Lambert iGEM’s 2022 RCA protocol (see Experiments: RCA 2022 protocol) requires samples to be incubated in the thermocycler at 37°C for two hours. However, the properties and usage of SplintR Ligase show that the reaction is successful with a 15-minute ligation time (Avantor Staff). Therefore, we ran RCA utilizing the miR-1 padlock probe (BBa_K4245200) with four different ligation times with miR-1 (BBa_K4245006): 15 minutes, 30 minutes, one hour, and two hours. After amplification, the reactions and controls were run on a gel; the bright bands near the top of the well showed that DNA product was produced for all reactions except for 15 minutes (see Fig. 5). Moving forward, we implemented a 30-minute ligation time (see Experiments: Optimized RCA protocol).

Figure 5. 1: 15-minute ligation, 2: 30-minute ligation, 3: 1-hour ligation, 4: 2-hour ligation, A: control (no enzymes); 2% agarose gel ran for 1 hour at 48V

Amplification Time

Lambert iGEM’s 2022 RCA protocol (see Experiments: RCA 2022 protocol) requires samples to be incubated in the thermocycler at 37°C for eight hours for amplification. To reduce this time, we ran RCA with two different concentrations of miR-1. After hybridization and ligation, we incubated the reactions in the plate reader at 37°C with 4uL SYBR™ Safe (see Experiments: SYBR™ Safe with RCP). The reactions were run overnight and the subsequent fluorescence was quantified in a plate reader (excitation wavelength 480 nm; emission wavelength: 528 nm): 30-minute intervals (see Fig. 6). Over time, the two RCA reactions increased in fluorescence, with no SEM overlap observed between the starting time and 5-hour mark. This suggests that RCP can be produced optimally starting at 5 hours. The significant increase in fluorescence between the RCA reactions and controls shows that SYBR™ Safe can determine the presence of RCP; however, the lack of difference between the 40.8 pM and .41 pM of miR-1 fluorescence indicates that the dye is not sensitive enough to differentiate between miRNA concentrations. Therefore, we did not continue to utilize SYBR™ Safe as a reporter.

Figure 6. RCA amplification optimization reaction with 40.8 pM and 0.41 pM of miRNA. No SEM overlap between starting point and 5-hour incubation time. SEM overlaps between 40. pM and .41 pM show that miRNA concentrations cannot be differentiated by SYBR™ Safe.

Phi29-XT

Lambert iGEM’s RCA protocol (see Experiments: Optimized RCA protocol) utilizes Phi29 DNA polymerase to perform amplification, resulting in an 5 hour amplification time. Phi29- XT DNA polymerase is an optimized enzyme with improved thermostability and sensitivity, which could shorten this time down to 2 hours (Biolabs). We ran RCA following the protocol for Phi29-XT on the New England Biolabs website (see Experiments: Amplification with Phi29- XT). The reactions and controls were run on a gel; no visible bands could be seen on the gel, indicating that RCP was not produced and therefore the reaction was not successful. As a result, we did not pursue utilizing Phi29-XT for further RCA reactions (see Fig. 7).

Figure 7. A: control (no enzymes), X: RCA with phi29-XT DNA polymerase; 2% agarose gel ran for 1 hour at 48V

Overall, we were able to reduce the RCA workflow from around 15 hours to 7 (see Experiments: Optimized RCA Protocol).