To develop an approach to screen cancer risk of post myocardial infarction (MI) patients, we designed LIRA to target miR-210-3p and miR-142-3p (H210 and H142), two cancer-promoting miRNA that are highly expressed in post MI patients. As shown in the Figure 1, LIRA contains a recognition region, a ribosome binding sequence (RBS) and a start codon (AUG). Basically, the recognition region of LIRA is a reverse complimentary sequence of its target RNA. Without interaction with its target RNA, LIRA forms a stem-loop structure, and the RBS and AUG are locked in the stem region, which cannot initiate the translation of the downstream reporter gene. In the presence of the target RNA of LIRA, the recognition region of LIRA could form base-paring interaction with its target RNA, which disrupts the stem-loop structure of LIRA and release the RBS and AUG for translation of the reporter gene.

To design LIRA that target miR-210-3p and miR-142-3p, we basically put a sequence that is reverse complimentary to these miRNAs at the recognition region of LIRA. As shown in the Figure 1, the recognition region of LIRA contains a* and b*. The recognition region b* is designed in the stem structure of LIRA, and the recognition region a* is designed in the loop structure of LIRA. We hypothesized that the ratio of the nucleotides of recognition region that are in the stem structure of LIRA to the nucleotides of recognition region that are in the loop structure of LIRA (stem-loop ratio) is critical for the function of LIRA, as it determines how effective LIRA could interact with its target RNA. To test our hypothesis, we designed LIRA that targets to miR-210-3p or miR-142-3p with different stem-loop ratios. For miR-210-3p, which has 22 nucleotides, we designed LIRAs with stem-loop ratios of 2/20, 7/15 and 10/12, and for miR-142-3p, which has 23 nucleotides, we designed LIRAs with stem ratios of 3/20, 8/15 and 11/12 (Table 1). We would like to test whether stem-loop ratio might influence the function of LIRA. In addition, these 6 LIRAs have different free energies, indicating that they might have different stabilities (Table 1). Therefore, we would like to test whether free energies of LIRA alone might influence the function of LIRA as well.

We cloned these 6 LIRA sequences into plasmid pCOLADuet-1 (Figure 2). To enhanced expression of LIRA system, we chose T7 promoter as our promoter. The expression of reporter gene would be increased by IPTG induction so that we can control reaction time. In addition, a EGFP reporter gene was cloned downstream of LIRA. Therefore, a functional LIRA would initiate the expression of EGFP in the presence of its target miRNA. For the expression of miRNA, we cloned miR-210-3p and miR-142-3p into the plasmid pET-15b and pCDFDuet-1 (Figure 3). Plasmid pCOLADuet-1, pET-15b and pCDFuet-1 carry different antibiotic marker genes for selection, and they also have mutually compatible origin of replication. Therefore, they can be transformed into bacterium simultaneously, and desired colonies can be selected by using different combinations of antibiotics. All the constructs were synthesized by GenScript.

Constructs with LIRA (H142 or H210) were transformed into competent BL21-DE3 bacterium either alone or with constructs with miRNA (input142 or input210), and the transformed bacterium were plated on LB agar plates with corresponding antibiotics to select successfully transformed colonies (Figure 4).

Next, we picked up colonies from the plates and grew these colonies in LB medium. Plasmid DNA were extracted from these colonies and digested with endonuclease. The digested plasmid DNA were analyzed by Agarose Gel Electrophoresis, which confirmed that plasmids from these colonies have correct sizes (Figure 5-6).

IPTG-induced expression of EGFP protein were conducted in these transformed clones by culturing in LB with 0.1mM IPTG for 4h. After IPTG-induction, these bacteria were collected by centrifugation and disrupted by ultrasonic. We measured the concentration of protein in cell lysate and then diluted them into 1.8μg/μL. 100μL cell lysate was added to 96-well plate and fluorescence from EGFP was measured by Multi-Mode Microplate Reader (Ex:488 Em:525).

H142 3/20 and H142 8/15 showed no difference with or without the input142. H142 11/12 showed significantly increased fluorescence with input142. Therefore, H142 11/12 performed much better than H142 3/20 and H142 8/15 on targeting miR-142-3p (Figure 7).

While H210 7/15 and H210 10/12 showed no difference with or without input210, H210 2/20 showed significantly increased fluorescence with input210. Therefore, H210 2/20 performed much better than H210 7/15 and H210 10/12 on targeting miR-210-3p (Figure 8).

From the results of our test, we can draw the following conclusions.

1. For LIRA to miR-142-3p, the LIRA with stem-loop ratio of 11/12, which is the LIRA with the highest stem-loop ratio, performs the best. For LIRA to miR-210-3p, the LIRA with stem-loop ratio of 2/20, which is the LIRA with the lowest stem-loop ratio, performs the best. Therefore, it seems that stem-loop ratio might not have big influence on the function of LIRA to miRNA.

2. For LIRA to miR-142-3p, the LIRA with free energy of -41.1, which is the LIRA with the highest free energy, performs the best. For LIRA to miR-210-3p, the LIRA with free energy of -66.61, which is the LIRA with the lowest free energy, performs the best. Therefore, it seems that free energy of LIRA might not have big influence on the function of LIRA to miRNA.

To detect miR-210-3p and miR-142-3p simultaneously, we designed LIRAs with a double-arm structure that can target two target RNA (Figure 9). Double-arm LIRA has two recognition regions, and each recognition region could form base pairing with a target RNA. Ideally, only when both recognition region of the double-arm LIRA forms base pairing with its target RNA, it can destabilize the stem-loop structure of LIRA to expose RBS and AUG to initiate the expression of downstream reporter genes.

We used sequences that were reverse complimentary to miR-210-3p and miR-142-3p as recognition region, and utilized NUPACK to designed 100 double-arm LIRAs. We simulated the hybridization process of these LIRA sequences with miR-142-3p and miR-210-3p, and excluded the LIRA sequences that would not meet the expectation of our design. To optimize LIRA design, we introduced multiple indicators to calculate the score of expectation of LIRA, and finally chose 3 LIRA sequences for experiment test (Please see our Model section Optimization of double-arm LIRA for detailed information). While double-arm LIRA 2 and double-arm LIRA 5 have highest score of expectation, double-arm LIRA 1 have a medium score of expectation.

As the cloning of single-arm LIRA described in the previous engineering cycle, We cloned the double-arm LIRA sequences into plasmid pCOLADuet-1 (Figure 10). To enhanced expression of LIRA, we chose T7 promoter as the promoter of LIRA. A EGFP reporter gene was cloned downstream of LIRA. Therefore, a functional double-arm LIRA would initiate the expression of EGFP in the presence of both miR-210-3p and miR-142-3p. All the constructs were synthesized by GenScript.

We transform constructs with double-arm LIRA and miRNA to E coil BL21-DE3 by heat shock and selected transformed bacteria colonies by selection on plates with combined antibiotics (Figure 11-13).

To verify whether our transformation was successful, we purified plasmid DNA from colonies by maxiprep and digested the plasmid DNA by restriction endonuclease. The digested plasmid DNA was analyzed by agarose gel electrophoresis, which confirmed that the transformed colonies have the right size plasmid DNA(Figure 14-16).

Next, all clones were conducted IPTG-induced expression of EGFP protein by culturing the transformed clones in LB with 0.1mM IPTG for 4h. After IPTG-induction, the transformed clones were disrupted by ultrasonic. We measure the concentration of protein in cell lysate and then diluted them into 1.8μg/μL. 100μL cell lysate is added to 96-well plate and measured by Multi-Mode Microplate Reader (Ex:488 Em:525). Double-arm LIRA 1 showed no difference with or without input210 and input142 (Figure 17). Double-arm LIRA 2 showed increased fluorescence with input142 alone, indicating that Double-arm LIRA 2 could be activated by miR-142-3p alone (Figure 18). Double-arm LIRA 5 showed significantly increased fluorescence only in the case that both input142 and input210 were present, indicating that the activation of Double-arm LIRA 5 requires both miR-210-3p and miR-142-3p (Figure 19).

The double-arm LIRA 5, which has the highest score of expectation, performed best, which demonstrates that our optimization of double-arm LIRA with indicators works.

In the future, we would like to apply our LIRA-based screening approach to screen blood samples from post-myocardial infarction patients, so as to identify those patients with high risk of cancer. Since it is unlikely that we could transform miRNAs from blood samples into bacteria, we need to adapt our LIRA-based screening approach into a cell-free system. In addition, for a screening approach, it would be much easier to detect changes of color than expression of EGFP, which requires machine to measure fluorescence. Therefore, we set to make two improvements of our LIRA-based screening approach. First, we change the reporter gene from EGFP to LacZ, which could hydrolyze CPRG (yellow) into CPR (purple). Second, we use a cell-free system to express the LacZ cloned in our LIRA constructs.

To set up the cell-free system for LIRA, we first tried H01 LIRA that we have successfully used in our experiments at the beginning. We replaced the reporter gene from EGFP to LacZ in H01 LIRA construct, and have it synthesized by Genscript company (Figure 20). For the cell-free system, we chose the PURExpress® In Vitro Protein Synthesis from New England BioLabs. We used Input01 construct for the expression of input RNA and engineered H01 LIRA construct for the expression of LacZ following the protocol provided by the manufacture.

We first added Input01 plasmid to the In Vitro transcription system provided by the T7 High Yield RNA Synthesis Kit (Yeasen) and incubated it at 37℃ for 4h to get Input01 RNA. Then, we added the H01 LacZ plasmid with or without purified Input01 RNA to the cell-free system provided by the PURExpress® In Vitro Protein Synthesis and incubated it at 37℃ for 2h. To validate that H01 LIRA can work in the cell-free system, we added CPRG. The results could be visualized or quantified by measuring OD₅₆₂ with a multi-mode microplate reader. While the sample with H01 plasmid alone could not change the color of CPRG, indicating that LacZ was not expressed, the sample with both H01 plasmids and Input01 RNA changed the color of CPRG, indicating that LacZ was expressed in this sample (Figure 21). This result demonstrated that input01 RNA activated H01 LIRA and initiated the expression of LacZ.

This engineering cycle validated our idea of combining LIRA with a cell-free system, which paved the way of applying our engineered LIRA that targets miR-142-3p and miR-210-3p into a cell-free system.