Overview

Problem

Heart disease has been one of the most prevalent diseases in the United States, with 1 in 4 deaths caused by cardiovascular disease (Centers for Disease Control, 2022). According to the Centers for Disease Control and Prevention, heart disease has contributed to over 25% of all deaths in Georgia, Lambert iGEM’s home region, in 2018 - that is 5% greater than the overwhelmingly large national statistic. Historically, the southern region has had issues concerning obesity, high blood pressure, smoking, and diabetes, all of which are exacerbated by a family history of coronary artery disease (CAD). These problems only increase with the “coincidental” ethnic and racial minorities present within these low-income populations (Hames and Greenlund). However, with a process that considers both affordability and accuracy, we can proliferate the means of diagnosing CAD. This starts with inclusive initiatives targeting these overlooked demographics.

Methodology

Last year, Lambert iGEM created CADlock, an accessible point-of-care diagnostic tool to both detect and quantify microRNA (miRNA) by utilizing rolling circle amplification (RCA). The rolling circle product (RCP) is a continuous transcription of the team’s designed padlock probe, which consists of many interspaced repeats of the reporting mechanism and target miRNA’s complementary sequence (Shi et al., 2012; Song et al., 2010; Ouyang et al., 2019; Ye et al., 2019; Zheng et al., 2016). We identified two miRNAs found to be upregulated to CAD, hsa-miR-1-3p (BBa_K4245006) and hsa-miR-133a-3p (BBa_K4245009), to serve as an early indication of the disease (Kaur et al., 2020).

Initially, we considered Reverse Transcription Quantification PCR (RT-qPCR) which was appealing due to its regularity with current screening, especially with the COVID-19 tests. However, through conversations with experts and further analysis of protocols, we identified the dangers of this method due to the high rate of false positive mismatches and determined that RCA held the most advantages. Not only does RCA eliminate the use of primers–which RT-qPCR requires–but it has no data concerning any mismatch or inconsistent results from our literature studies (Shi et al., 2012; Song et al., 2010; Ouyang et al., 2019; Ye et al., 2019; Zheng et al., 2016).

Moreover, diagnostic tools such as echocardiograms and angiograms are expensive – costing upwards of $1000 – and not readily available at most primary care facilities, preventing a large portion of the population from receiving effective screening (Hansing C. E.). However, our RCA assay costs approximately $9, which is less than 1% of most common diagnostic procedures.

Padlock Design

A padlock probe is a single-stranded DNA sequence (30-150 nucleotides in length) designed to hybridize to a specific RNA/DNA target– in this case, our target miRNA (Jonstrup et al., 2006). The ends or “arms” of the probe are complementary to the target sequence, which allows the miRNA to bind antiparallel to the padlock arms (see Fig. 1). We designed these ends to have similar annealing temperatures to maximize their binding efficiency with the miRNA. In between the ends is the reporter sequence, which can be utilized to quantify the initial miRNA concentration (see Fig. 1). Furthermore, we added a phosphate group modification to the 5’ end of the padlock sequence (Jonstrup et al., 2006).

Figure 1. Padlock probe formation and arrangement with miRNA

The following steps are for generating a padlock probe by hand. Users will need a software tool such as SnapGene that displays melting temperatures of sequences.

  1. Paste target sequence in software of choice.
  2. Take the reverse complementary sequence of the target.
  3. Split the sequence in half. Use the first half of the sequence as the end of the padlock probe and the second half as the beginning of the probe.
  4. Insert the desired reporter sequence in between the two halves of the sequence.
  5. Calculate the annealing temperature of both halves and move nucleotides one at a time from one end of a half to the end of the other half until the difference between the annealing temperatures of the two arms is minimal.

Last year, Lambert iGEM developed Probebuilder: a novel, intuitive, and user-friendly program for designing padlock probes. After validating the hsa-miR-451a padlock probe generated by Probebuilder, the team has continued to use the software to design padlocks for RCA reactions (see Lambert iGEM Wiki Software, 2022).

Probebuilder’s source code is available on the 2022 Gitlab repository

Process

The RCA process relies on the continuous replication of padlock probes to produce RCP that can be quantified and correlated to the target miRNA concentration. There are three main steps in the RCA process detailed by Jonstrup et al. in their 2006 paper: hybridization, ligation, and amplification. By designing the padlock probe to detect a specific miRNA target, the miRNA and padlock probe arms form a DNA-RNA hybridization (see Fig. 2a, 2b). The arms are then in closer proximity, and the enzyme SplintR ligase (which only ligates near RNA-DNA hybridization) utilizes a phosphate modification on the 5’ arm and ATP to circularize the padlock probe (Avantor Staff) (see Fig. 2c). After ligation, phi29 DNA polymerase uses the miRNA hybridized to the padlock arms as a primer to initiate and perform amplification (see Fig. 2d, 2e). The resultant rolling circle product (RCP) contains interspaced repeats of the middle sequence of the padlock probe (see Fig. 2f). This sequence of the RCP is complementary to the reporters (linear probes or lettuce) used to correlate fluorescence to varying concentrations of the target miRNA.

Figure 2. RCA process: (a, b) hybridization, (c) ligation, (d, e) amplification, (f) rolling circle product

2022 Results

In 2022, we conducted multiple rounds of experimentation to generate quantifiable results with our product. First, we conducted initial experiments to assess how well our hsa-miR-1-3p biosensor (BBa_K4245200) performs in a controlled laboratory setting by evaluating the effectiveness of rolling circle amplification (RCA) and various reporter mechanisms. All results in Figures 4-11 and13 were quantified in a plate reader at exciation wavelength of 480 nm and emission intensity at 518 nm.

  1. To quantify the rolling circle product (RCP), we first used the fluorescence dye SYBR™ Safe dye which fluoresces when bound to ssDNA. The results were inconsistent, so we tested the potential rolling circle product (RCP) through gel electrophoresis. The gel showed a band near the wells, indicating a long strand of DNA was produced, likely our RCP (see Fig. 3).

    Figure 3. Gel electrophoresis of RCA reactions, depicting bands greater than 1 kB; 2% agarose gel ran for 1 hour at 48V

  2. Through a literature review, we found a lettuce aptamer (BBa_K4245133) and tested it with simulated RCP (BBa_K4245131). This showed an increase in fluorescence compared to the controls (see Fig. 4), indicating that the lettuce aptamer is able to fluoresce when bound to the simulated RCP (VarnBuhler et al., 2022).

    Figure 4. Lettuce aptamer with simulated RCP.

  3. We performed a series of experiments of the split Lettuce aptamer (BBa_K4245134; BBa_K4245135) with actual RCP. The results display a significant increase in fluorescence with RCP and split lettuce as compared to the controls, indicating that the lettuce aptamer was able to successfully bind to the RCP and induce fluorescence of DFHBI-1T (see Fig. 5).

    Figure 5. Graph of fluorescence from left to right: 1) water + dye control, 2) split lettuce + dye control, 3) split lettuce + RCP + dye, 4) split lettuce + padlock&miRNA control + dye.

  4. After consultation with other experts on linear DNA probes, we decided to test these with simulated RCP (BBa_K4245131) to ensure linear probes were an effective and characterizable means of quantifying miRNA (Zhou et al., 2015). The FAM dye tagged probe (BBa_K4245130) and BHQ-1 quencher tagged probe (BBa_K4245132)) are complementary to our RCP. When the linear DNA probes are bound to the RCP, the fluorescence produced by the FAM Dye is quenched by the BHQ-1 quencher which significantly reduces fluorescent intensity (Zhou et al., 2015). As shown in Figure 6, linear DNA probes with simulated RCP have significantly less fluorescence than just the FAM dye, indicating the mechanism would be efficient with actual RCP.

    Figure 6. Linear DNA probes with simulated RCP.

  5. We further characterized the linear DNA probes with various simulated RCP concentrations. Figure 7 below shows a negative logarithmic correlation between the complement concentrations and the relative fluorescence units, a parallel relationship to our ODE model (see Fig. 8).

    Figure 7. Linear DNA probes with various concentrations of simulated RCP.
    Figure 8. Deterministic ODE Model Simulation of RFU output dependent on concentration of linear DNA probe complement concentration.

  6. After we determined the vitality of using linear DNA probes with simulated RCP, we tested these probes with actual RCP. As shown in Figure 9 below, there is a statistically significant decrease in the fluorescence output of the probes with RCP as compared to that of just the FAM-tagged probe. This confirms that we produced our desired RCP.

Figure 9. Fluorescent readout of RCP for miRNA-1-3p and miRNA-133a-3p using linear DNA probes.
  1. We further characterized linear DNA probes to quantify the relationship between various miRNA concentrations and fluorescence. There is a negative logarithmic correlation between the miRNA concentrations and the RFU (see Fig. 10), a parallel relationship to our ODE model (see Fig. 11), validating that RCA coupled with linear probes are an effective means of quantifying miRNA.
Figure 10. Characterization curve for showing a negative logarithmic relationship between RFU from linear DNA probes and various miRNA concentrations.
Figure 11. Deterministic ODE Model Simulation of RFU output dependent on miRNA concentration

Serum Testing

To assess the feasibility of our biosensors in real-world applications, Lambert iGEM experimented with pooled human serum. Upon meeting with Dr. Charles Searles from the Emory University School of Medicine, we tested the ability of our RCA sensors to detect miRNAs in blood serum by spiking pooled human serum with hsa-miR-1-3p and RNase inhibitors (to prevent native RNases from degrading the added miRNA). The gel electrophoresis results show clear bands close to the wells (see Fig. 12), indicating our biosensors were able to produce a long strand, likely the RCP.

Figure 12. Agarose gel with RCP from RCA reactions and negative control reactions that had no added miRNA from serum treated with 10 nM miRNA; 2% agarose gel ran for 1 hour at 48V

To experimentally validate that the long DNA strand that was produced is our desired RCP, we utilized linear DNA probes. As shown in Figure 13, there is a statistically significant decrease in the fluorescent output of the linear DNA probes with RCP as compared to that of just FAM tagged probes, validating the success of the RCA reaction with hsa-miR-1-3p in spiked serum.

Figure 13. Linear DNA probes fluorescence output from RCA with serum extracted miRNA-1-3p.