Engineering Success

Investigating how to design toehold switches of interest, we encountered GENOSWITCH, a software created by iGEM City of London UK 2023 able to output a toehold switch with given microRNA inputs. The City of London UK 2023 iGEM team was unable to verify toehold switches created in silico with in vivo or in vitro wet lab work. As a result, we strived to assess the capability of a past iGEM team’s software to accurately engineer toehold switches that respond to complementary microRNAs within our project. Sourcing microRNA nucleotide sequences from miRBase Entry and the Atlas of Genetics and Cytogenetics, hsp-microRNA-155 and hsp-microRNA-21 were input into GENOSWITCH, outputting toehold switches responsive to each microRNA.

GENOSWITCH is additionally capable of generating toehold switches responsive to multiple microRNA inputs. To craft our AND toehold switch, a construct complementary to hsp-microRNA-21, hsp-microRNA-155, and an AND gate linking the two microRNAs, miRNA-21 and miRNA-155 were entered into the software. Each toehold switch nucleotide sequence output was imaged in the RNAFold web server in order to evaluate the RNA folding, Gibbs Free energy, and position of the ribosome binding site (RBS) in the toehold switch

GENOSWITCH software output for miRNA-21. Upon inputting the gene sequence for a microRNA, the software provides a toehold switch sequence and Boltzmann diagram of the predicted RNA structure.

RNAfold web server predicted RNA folding of miR-21 toehold switch. Visualizing the miR-21-repsonsive toehold switch was vital for understanding how toehold switches are structured, and gaining insight into how our AND toehold switch would work.

Choice of promoter was influenced by the viability of certain promoter systems in a cell-free system and within experimental design. Genes encoding microRNAs and toehold switches were placed under separate promoter systems, the rhaBAD promoter inducible by (L-rhamnose) and the T7 promoter (inducible by IPTG) respectively. Differentiating the promoters that each type of construct was produced under would allow for experimental control over the production of microRNAs and toehold switches in bacterial cultures, permitting for future testing over the effect of inducer on total fluorescence output by cultures producing a toehold switch with its complementary miRNA. Moreover, the T7 promoter system is compatible with lysates used in cell-free systems, and is compatible with common RNA purification protocols. This would allow for downstream use of our constructs in in vitro transcription and testing of a cell-free experimental set up. Our selection of the rhaBAD promoter system for microRNA-encoding constructs was informed by its status as a well-documented, non-leaky inducible promoter.

Our goal to achieve in vivo testing of the toehold switches necessitated a method for producing microRNAs and toehold switches within our chassis of choice, E. Coli. To do so, both plasmids encoding each microRNA and plasmids encoding each designed toehold switch were designed. After thoughtful analysis, pSB1K3 was selected as the plasmid vector for each gene encoding a microRNA sequence. The presence of mRFP, a red fluorescent protein in pSB1K3, allowed for a form of preliminary screening that indicated whether or not DNA assembly was successfully performed to ligate a microRNA-encoding gene insert into the linearized vector. If bacterial colonies expressed red fluorescence, the pSB1K3 vector remained unchanged and DNA assembly was unsuccessful. If bacterial colonies lacked fluorescence, successful assembly occurred: each microRNA-encoding gene was correctly assembled in the vector such that it would replace the mRFP gene in pSB1K3. pNP1 served as the plasmid vector for each gene encoding each toehold switch. Our decision to use pNP1 was based on the presence of superfolder GFP (sfGFP) in the plasmid. Each toehold switch-encoding gene could be assembled into pNP1 such that sfGFP was directly downstream of the riboregulator, allowing for the fluorescent protein to act as the reporter for toehold switch activity.

Our goal to achieve in vivo testing of the toehold switches necessitated a method for producing microRNAs and toehold switches within our chassis of choice, E. Coli. To do so, both plasmids encoding each microRNA and plasmids encoding each designed toehold switch were designed. After thoughtful analysis, pSB1K3 was selected as the plasmid vector for each gene encoding a microRNA sequence. The presence of mRFP, a red fluorescent protein in pSB1K3, allowed for a form of preliminary screening that indicated whether or not DNA assembly was successfully performed to ligate a microRNA-encoding gene insert into the linearized vector. If bacterial colonies expressed red fluorescence, the pSB1K3 vector remained unchanged and DNA assembly was unsuccessful. If bacterial colonies lacked fluorescence, successful assembly occurred: each microRNA-encoding gene was correctly assembled in the vector such that it would replace the mRFP gene in pSB1K3.

pNP1 served as the plasmid vector for each gene encoding each toehold switch. Our decision to use pNP1 was based on the presence of superfolder GFP (sfGFP) in the plasmid. Each toehold switch-encoding gene could be assembled into pNP1 such that sfGFP was directly downstream of the riboregulator, allowing for the fluorescent protein to act as the reporter for toehold switch activity

pNP1 plasmid containing gene sequence encoding toehold switch for hsp-microRNA-21

pSB1K3 plasmid containing gene sequence encoding hsp-microRNA-21

MicroRNA and toehold switch genes were ordered from IDT DNA. Due to the repetitive nature of toehold switch sequences, the toehold switches were difficult to synthesize as a gene block and so we ordered them as multiple overlapping oligos to be annealed in lab. We designed those oligos by using the Primerize tool [1]

The main mechanism of building our constructs was High-Fidelity (Hi-Fi) DNA assembly (New England Biolabs). For each toehold switch or microRNA-encoding gene to be ligated with a plasmid vector, we created a pair of outward-facing primers that contained 12-20 base pair overlaps with that insert’s plasmid vector. Inward-facing primers were designed to amplify each vector.

The process of building each toehold switch-encoding plasmid consisted of first annealing the oligos forming the toehold switch and then assembling the primers via a standard 25-cycle PCR protocol. We then created a linearized pNP1 vector with inward primers via a standard 25-cycle PCR protocol and performed Hi-Fi DNA assembly as directed by New England Biolabs. Assembled plasmids were first transformed into DH5-α E. Coli due to its high transformation efficiency, which would let us increase our supply of DNA at hand for future experiments. Plasmids for microRNAs and single-input toehold switches were later transformed into (DE3) BL21 for downstream testing.

Succesful transformations of Hi-Fi DNA assembly products for the toehold switch responsive to miRNA-21 (top right), the toehold switch for miRNA-155 (bottom right), the AND toehold switch (top left), and a positive control for Hi-Fi DNA assembly (bottom left) into DH5-α E. Coli.

A colony PCR screening protocol was developed to ensure that our constructs contained the desired gene sequence, and that each gene of interest was inserted into the correct position on the plasmid. After undergoing successful PCR screening, each plasmid underwent full-plasmid nanopore sequencing by Plasmidsaurus.

The PCR screening process made use of a reverse primer that would anneal onto the plasmid vector and a forward primer annealing onto a unique sequence found on the insert, such as a sequence on the toehold switch for miRNA-21. The resulting PCR product would have a certain expected length in base pairs; the colony PCR product could thus be visualized in a electrophoresis gel to determine its size and serve as an indicator as to whether or not DNA assembly was successful.

Gel Electrophoresis for PCR screening of Toehold Switch Plasmids. The 1 kb plus NEB ladder was used, such that the top band is 10 kilobases and the lowest band is 0.1 kilobases. Top Lanes 2-9 are PCR Screening products for the plasmid containing the toehold switch for miRNA-21. With an expected size of 592 base pairs, the PCR screening indicates that the miRNA-21 toehold switch plasmid was successfully assembled

The final build stage of our project was transforming a pair of complementary plasmids, one containing a toehold switch and the other encoding a complementary miRNA (eg. toehold switch for miRNA-21, and the miRNA-21-encoding gene), into (DE3) BL21 E. Coli. The selection of the BL21 (DE3) E. Coli strain was informed by its having T7 RNA polymerase. This would prove useful for expression of our toehold switch constructs which are under control of the T7 promoter system.

Production of both a toehold switch and microRNAs within the same cell provides insight into toehold switch-miRNA interactions and affinity. For example, if a toehold switch and a non-complementary miRNA were to be produced in the same cell, no fluorescence would be expected. However, fluorescence would be expected if a toehold switch was exposed to a complementary microRNA produced in the same cell. Testing non-matching toehold switches and miRNAs can demonstrate a toehold switch’s specificity to its complementary microRNA.

In this way, maintaining production of both parts of the system indicates whether the constructs were correctly complementary to each other, and lacked the ability to interact with non-complementary structures. This helps to experimentally validate or disprove the ability of GENOSWITCH software to produce toehold switches complementary to an input microRNA.

To experimentally assess the interaction between complementary microRNAs and their toehold switches, two experimental designs were developed. Firstly, promoter induction experiments were performed. These tests aimed to determine the amount of fluorescence produced by a toehold switch and its complementary microRNA. Due to time limitations, the AND toehold could not undergo PCR screening and sequencing in time to begin the testing phase.

In addition to single-input toehold switches, a pNP1 toehold switch and a T7 control plasmid were tested. The pNP1 toehold plasmid, complementary to the NP1 microRNA, was sourced from Addgene; the NP1 microRNA plasmid was constructed by designing a gene sequence encoding the pNP1 microRNA trigger and assembling it with the pSB1K3 vector. The T7 control plasmid was designed via site-directed mutagenesis, where the toehold switch part of pNP1 was eliminated, leaving superfolder GFP under a T7 promoter system; this was a control for proper T7 induction.

Sourced from [2], the pNP1 toehold switch was experimentally confirmed in past literature to be complementary to the NP1 microRNA, and to permit expression of sfGFP as a reporter protein when the NP1 trigger is co-expressed. For this reason, the pNP1 toehold switch and NP1 microRNA served as positive controls for toehold switch-microRNA interaction and affinity.

The constructed T7 control plasmid acted as a positive control for the T7 promoter system, demonstrating the level of fluorescence expressed by sfGFP under T7 when no toehold switch is present in the system.

Experimental Design for First Promoter Induction

A time course experiment was performed to test fluorescence over time when a set concentration of inducer was provided or withheld to cell cultures containing both a toehold switch and its complementary miRNA. In each time course, fluorescence and absorbance readings were performed at 30 minutes to 1 hour intervals. The amount of inducer added to induced cultures was held constant, and provided to the bacterial cultures once they grew to 0.3-0.5 A, measured at 600 nanometers. As a negative control, absorbance and fluorescence of LB media was measured at each time point.

Absorbance v. Time graph for first time course experiment for (DE3) BL21 E. Coli culture transformants with miRNA-21 and its complementary toehold switch. Across all three time points, the uninduced culture experienced higher absorbance relative to the induced culture.

Fluorescence v. Time graph for first time course experiment for (DE3) BL21 E. Coli culture transformants with miRNA-21 and its complementary toehold switch.

During our first time course experiment for promoter induction, the negative control consisting of only media, with no bacterial culture, frequently had relatively high fluorescence compared to bacterial cultures. Over time, the cultures induced for expression of the miR-21-responsive toehold switch and miR-21 itself demonstrated an increasingly higher fluorescence compared to the uninduced culture.

Difference in Fluorescence v. Time graph for first time course experiment for (DE3) BL21 E. Coli culture transformants with miRNA-21 and its complementary toehold switch, between induced and uninduced cultures. Over time, the difference in fluorescence increased between induced and uninduced cultures.

This increasing difference in fluorescence suggests, when induction of the T7 and rhaBAD promoters occurs, microRNA-21 is readily expressed and able to trigger translation of the downstream sfGFP on its complementary toehold switch. Without promoter induction, fluorescence is comparatively low, denoting that fluorescence is resulting from co-expression of a toehold switch and its complementary microRNA in the same system.

To account for the high fluorescence in LB media seen in our first experiment and better characterize the difference in fluorescence between induced and uninduced miRNA-21 cultures, we completed an additional promoter induction experiment. While the first time course only involved reading the fluorescence of 100 microliters of a culture at each time point for each bacterial culture, the second time course measured the fluorescence of 5, 100 microliter replicates of each bacterial culture in order to control for random error and variance in measurement by the plate reader used.

Second Promoter Induction Test

Absorbance v. Time graph for the second time course experiment for (DE3) BL21 E. Coli culture transformants with miRNA-21 and its complementary toehold switch. Here, absorbance was measured on a spectrophotometer, and the negative LB control was set as the reference for all absorbance measurements.

Fluorescence v. Time graph for second time course experiment for (DE3) BL21 E. Coli culture transformants with miRNA-21 and its complementary toehold switch. The average fluorescence of each of the five replicates for each culture was taken and graphed.

In our second time course experiment, the LB media displayed less fluorescence compared to the first promoter induction time course and thus served as a more accurate negative control for induction. The induced culture demonstrated a higher amount of fluorescence compared to the uninduced culture, relative to the first time course test. At the third time point, the fluorescence of the induced culture for miR-21 and the miR-21-reponsive toehold switch reaches a maximum of 6.00 x 10⁵ relative fluorescence units, whereas the induced culture in the first time course had the highest fluorescence at exactly half the amount, 3.00 x 10⁵ relative fluorescence units.

Difference in Fluorescence v. Time graph for second time course experiment for (DE3) BL21 E. Coli culture transformants with miRNA-21 and its complementary toehold switch, between induced and uninduced cultures. Over time, the difference in fluorescence strongly increased between induced and uninduced cultures.

The second promoter induction time course illustrated a consistent, stronger increase in difference in fluorescence among the induced and the uninduced bacterial culture over time. The difference in fluorescence between the cultures at the third time point, 2.36 x 10⁵ relative fluorescence units, is more than double that of the prior promoter induction experiment, with a maximum difference in fluorescence of 1.00 x 10⁵ relative fluorescence units.

Analysis

Across the two promoter induction experiments, induced bacterial culture expressing both the toehold switch responsive to microRNA-21 and microRNA-21 exhibited higher fluorescence compared to the uninduced culture.

To determine significance of the data in our first time course, we performed an Analysis of Variance (ANOVA) two-sample test without replication between the fluorescence of the induced and uninduced cultures. This statistical measure was chosen as the fluorescence data is affected by 2 core factors: time of fluorescence reading and presence of inducer. In addition, for the first time course, each bacterial culture was sampled only once for every time point, highlighting the need for a test without replication.

With an alpha value of 0.05, the p-values for variation between rows and columns is less than α, indicating a statistically significant difference between fluorescence among time points and among induced and uninduced cultures.

In an ANOVA two-factor test without replication, variation between rows indicates the difference in fluorescence between the induced and uninduced cultures. Variation between columns signifies a difference in fluorescence across induced and uninduced samples, among different fluorescent reading time points. The p-value for rows was 0.08, whereas the p-value for columns was approximately 0.31. With α = 0.05, this would suggest that promoter induction had a near statistically significant impact on fluorescence and that no evidence was present to indicate that progression of time significantly impacted fluorescence.

To demonstrate significance in our second time course, we performed an Analysis of Variance (ANOVA) two-sample test with replication between the fluorescence of the induced and uninduced cultures. We chose to perform this test as both the induced and uninduced bacterial cultures were sampled multiple times.

With an alpha value of 0.05, the p-values for variation between sample, columns, and interaction is less than α, indicating a statistically significant difference between fluorescence among time points and among induced and uninduced cultures.

An ANOVA two-factor test with replication introduces the factor of interaction into the statistical analysis. The p-value for interaction is a measurement of how significant systematic differences between columns in the same row and between rows in the same column are. For our data, interaction would measure the extent to which promoter induction and progression of time affect each other.

With α = 0.05, the p-value for variation in the sample (p = 1.23 x 10⁻⁷ < α = 0.05), columns (p = 1.57 x 10⁻¹⁸ < α = 0.05) , and in interaction (p = 2.38 x 10⁻¹⁸ < α = 0.05) were all statistically significant. Holistically, the ANOVA two-factor test with replication results illustrate statistical significance in the difference in fluorescence among induced and uninduced (DE3) BL21 cultures with miRNA-21 and the miRNA-21-responsive toehold switch, suggests that promoter induction has a significant effect on fluorescence, relays that progression of time has a significant effect on fluorescence. It additionally indicates that the impact of promoter induction on fluorescence and the effect that progression of time has on fluorescence influence each other.

Testing Other Constructs

To evaluate the toehold-microRNA interactions among the miR-155 constructs, positive control for the T7 promoter system, and the positive control for toehold switch-miRNA interactions, parallel time course experiments were conducted for promoter induction of (DE3) BL21 E. Coli transformed with these constructs. To assess the specificity of a toehold switch to its complementary miRNA, a promoter induction time course test was additionally performed for BL21 E. Coli transformed with the microRNA-21-encoding plasmid and the pNP1 toehold switch-encoding plasmid.

Results from these experiments can be found in the Results page.

Testing Diffusion of MicroRNAs

A central question of our project was seated on the advantages and disadvantages of engineering a cell-free or in vivo at-home liquid biopsy test for Triple Negative breast cancer. An in vivo, cell based test would involve utilizing a chassis transformed with the AND toehold switch, responsive to both miRNA-21 and miRNA-155, that is able to have miRNAs freely diffuse into its cell membrane. The basis of this product design is founded in the research of [3], capitalizing on the diffusion of microRNA-155 into an E. Coli culture. The free diffusion of microRNAs into the chassis is a necessary component of the system, as it allows for the microRNAs to access the toehold switch produced within the cell.

To determine the feasibility of this design, a time-course experiment where (DE3) BL21 E. Coli containing each single-input toehold switch we designed was exposed to both complementary and non-complementary microRNAs while conducted. The presence of complementary or non-complementary microRNAs was varied among samples and the fluorescence in relative fluorescent units was measured across 1 hour time intervals. Inducer presence or concentration was kept constant, and the T7 control we constructed was additionally exposed to microRNA-155 and microRNA-21 to serve as a positive control. Results of this experiment can be found in the Results page.

Controlling microRNA Expression

A core disadvantage of the experimental design for the promoter inductions we performed is that it fails to pinpoint whether or not fluorescence is the result of toehold switch-microRNA interaction itself by only truly controlling for presence of an inducer in the system. To isolate the effect of the expression of the toehold switch from co-expression with a microRNA, a time course experiment was conducted in order to confirm that the fluorescence seen in our previous promoter induction experiments wasn’t solely due to expression of the toehold switch plasmid. (DE3) BL21 E. Coli containing the plasmid for expression of the miR-21 toehold switch and miR-21, the toehold switch-miRNA pair for miRNA-155, the toehold switch-miRNA pair for the pNP1 plasmid, and a mismatch between a toehold switch and a non-complementary microRNA were induced with IPTG for expression of their toehold switches. Results of this experiment can be found on the Results page.

Across our build stage, promoter induction experiments, and miRNA diffusion tests, we learned that more precise measurements and experimental design is crucial to investigate the specificity and affinity of toehold switches to their complementary microRNAs. Our second promoter induction experiment was able to portray a statistically significant impact of promoter induction and time on fluorescence on bacterial cultures containing both a toehold switch and its matching microRNA. In addition, our final promoter induction that only induced toehold switch expression gave insight into the concept that the fluorescence conveyed in the experiments was not only due to toehold switch-plasmid expression. However, replications of these experiments would be monumental in supporting and providing insight into toehold switch-microRNA interactions as seen in our lab work. Crafting experiments that would allow for direct data collection and assessment of toehold switch-microRNA affinity, without any lingering variables such as cell culture absorbance, would be a future scientific improvement upon this project.

Our design and build stage encouraged us to reflect on efficient methods for constructing repetitive or difficult-to-synthesize DNA sequences. While the single-trigger toehold switches were successfully constructed after one trial of annealing their oligos, we failed to construct the AND toehold switch and AND bridge across two trials, opted to re-design the gene sequences using Primerize and re-order our constructs, and are currently transforming them for future downstream testing. In the future, we may opt to strategically design our AND toehold and target to avoid overly repetitive sequences. For example, opting for a dual-promoter system, ribosomal skipping sequence, or multiple promoters could cut down on the base pair length of our AND target while working around the repetition created by using multiple rhaBAD promoters within the sequence.

More information on the specific successes and trials in our experimental procedures can be found in our Notebook and our Experiments page.

References

[1] Siqi Tian, Joseph D. Yesselman, Pablo Cordero, Rhiju Das, Primerize: automated primer assembly for transcribing non-coding RNA domains, Nucleic Acids Research, Volume 43, Issue W1, 1 July 2015, Pages W522–W526, https://doi.org/10.1093/nar/gkv538

[2] BioBits Explorer: A modular synthetic biology education kit. Huang A, Nguyen PQ, Stark JC, Takahashi MK, Donghia N, Ferrante T, Dy AJ, Hsu KJ, Dubner RS, Pardee K, Jewett MC, Collins JJ. Sci Adv. 2018 Aug 1;4(8):eaat5105. doi: 10.1126/sciadv.aat5105. eCollection 2018 Aug. 10.1126/sciadv.aat5105 PubMed 30083608

[3] Zhao L, Zhou T, Chen J, Cai W, Shi R, Duan Y, Yuan L, Xing C. Colon specific delivery of miR-155 inhibitor alleviates estrogen deficiency related phenotype via microbiota remodeling. Drug Deliv. 2022 Dec;29(1):2610-2620. doi: 10.1080/10717544.2022.2108163.