Team:TUDelft/Results
Results
On this page, we present the results of our experiments, including engineering of the vitamin-binding aptazyme through in vitro evolution, cell-free expression of our genetic construct, and the absorbance measurements conducted with our dedicated read-out device.
AptaVita
AptaVita aims to develop a quantitative, accessible, and modular biosensor that is able to detect vitamin deficiencies from blood samples. Throughout our project, we engineered our biosensor consisting of a ligand-regulated self-cleaving aptazyme that is fused to a reporter gene to give a colorimetric read-out of the vitamin concentration. This genetic construct was expressed in a cell-free system on paper support and the vitamin concentration was quantitatively analyzed by our dedicated hardware through absorbance measurements.
Engineering aptazymes for vitamin detection
Introduction
To construct our biosensor, we first engineered ligand-regulated self-cleaving aptazymes that detect vitamins through a novel iterative in vitro evolutionary method, De novo Rapid In Vitro Evolution of RNA biosensors (DRIVER) [1]. Each round of DRIVER consists of a transcription, reverse transcription, ligation, and PCR steps. For these evolutionary experiments, we alternated the transcription conditions between the absence and presence of ligands. This results in selection of aptazymes that selectively self-cleave in the absence of the ligands, and remain uncleaved in their presence. We chose folate (vitamin B9), thiamine pyrophosphate (B1), riboflavin (B2), pyridoxal 5′-phosphate (B6), and cobalamin (B12) as our target ligands.
Setup of DRIVER protocol
Validation of transcription protocol
To validate the first step of our DRIVER protocol, we decided to test an already known theophylline-binding aptazyme, gB-Theo(BBa_K3806008) This 109-bp aptazyme binds selectively to theophylline and was discovered by Townshend et al. [1], the same authors that developed the DRIVER method. We run a Urea-PAGE gel to visualize the transcription products of the gB-Theo aptazyme with and without the addition of theophylline at different concentrations. The bands corresponding to the uncleaved and cleaved products of the theophylline aptazyme are expected to be located at 109 bp and 89 bp, respectively. ImageJ was used to quantify the intensities (I) of the gel bands, and the cleavage fraction (f) was determined by:
We observed that the cleavage fraction decreases when increasing theophylline concentration: 71 % at 0 mM theophylline (Fig. 1, lane 3), 39 % at 0.5 mM theophylline (lane 4), and 4 % at 10 mM theophylline (lane 5). With this experiment, we successfully validated our transcription protocol and confirmed the ligand-dependent cleavage activity of the gB-Theo aptazyme [1]. Therefore, we used gB-Theo as a positive control throughout our entire DRIVER experiment.
Compatibility of target ligand solution in transcription
To engineer aptazymes for vitamin detection, target ligands are supplemented to the transcription reaction. We learned from our engineering cycle that potential interference of the ligand addition with transcription can be prevented by finding a compromise between ligand solubility and pH of the solution. The establishment of the reaction conditions for the transcription step is explained in detail on our Engineering Success page. In brief, to find this compromise, we prepared folate solutions at different pH and supplemented the transcription reaction, we used gB-Theo as the DNA template and ran a Urea-PAGE gel to visualize the transcription products.
We found that the addition of folate solution adjusted to a pH range of 8 to 9 does not affect transcription. This can be observed in Fig. 2 (lanes 3 to 6) as the intensity of the transcript products was comparable to that of the positive gB-Theo control (lane 7, no folate solution added).
The pH of the vitamin mixture (B1, B2, B6 and B12) was already within a range compatible with transcription and therefore used to perform DRIVER rounds.
DRIVER - in vitro evolution of aptazymes
As the transcription protocol was validated, also confirming its compatibility with target ligands, we started the in vitro selection rounds using the DRIVER method. Two high-diversity libraries of approximately 1012 potential biosensor sequences were used as inputs to DRIVER. The first library consisted of sequences with a large loop of 30 random nucleotides (N30) and the second library consisted of sequences with a large loop of 60 random nucleotides (N60). Throughout the aptazyme engineering process, two ligand(s) conditions were used: a folate solution and a mixture of vitamin B1, B2, B6, and B12. Combination of the two libraries and the two ligand conditions results in a total of four samples (N30-F, N30-V, N60-F, and N60-V). These four samples were used alongside a positive (gB-Theo) and negative (no DNA template) control on every evolution round. For further explanation on these design choices, visit our Design page.
As a first step on our evolution cycles, we run two cleavage selection rounds without the addition of ligands, to enrich the aptazymes that are able to cleave. We then run an Urea-PAGE after each round to visualize the transcript products. After the transcription step of the second round, transcription products could be observed (Fig. 3), suggesting that all the steps of the first round: transcription, RT, ligation, and PCR, were successful. Additionally, by comparing the two gels, it can be seen that in the second round, there is a slight enrichment in the cleavage fraction as the lower band intensifies (lanes 7 and 8 vs lanes 3 and 4). From these results, we conclude that the DRIVER method successfully enriched self-cleaving sequences during the first cleavage selection round.
We continued running DRIVER for 62 rounds, altering between +ligand (uncleavage) and -ligand (cleavage) rounds. With this, we enriched the population with the responsive aptazymes to our target vitamins. We run an Urea-PAGE gel after each round to confirm the transcription products. The intermediate results of the selected Urea-PAGE gels can be found in the DRIVER intermediate rounds-supplementary data below. One should notice that Urea-PAGE gels display high variability between rounds due to handling variation, thus the DRIVER method is highly suggested to be robot-automated. Yet, the use of Urea-PAGE was an important tool through the whole of our evolutionary process, which allowed us to follow the progression of the rounds. For instance, the N60 libraries were progressively lost, and no transcript products could be seen after 43 rounds of selection. This outcome is coherent with the results in the study of Townshend et al [1], where all but one of the aptazymes discovered presented a larger loop with less than 60 nucleotides. We hypothesize that the increase in the length of the large loop results in a large change of the aptazyme structure that hinders aptazyme self-cleavage, and therefore the selection process. Considering the aforementioned, we put our expectations on the N30 libraries and proceeded with NGS.
Library preparation for next-generation sequencing
To identify enriched sequences that exhibit a higher cleavage fraction in the absence of the target vitamins and lower fraction in their presence, DRIVER-evolved libraries should be screened using next-generation sequencing (NGS). Before NGS, libraries were prepared using the product of round 62 to run two distinct rounds: in the presence and absence of the ligands of interest (Fig. 4). For each condition, cleaved and uncleaved populations were amplified enabling computation of the cleavage fraction for each sequence in the mixture after NGS.
Aptazyme hit identification through next-generation sequencing
The NGS data analysis was performed with our sequence analysis pipeline, developed based on the work of Townshend et al. [1]. For details on the development of this sequencing pipeline visit our Software page. Our sequence analysis pipeline analyzes the sequences based on their ability to remain uncleaved in the presence of the target vitamin and to cleave in its absence. A scatter plot is generated to portray how the population of potential aptazyme sequences distribute based on the fraction that underwent cleavage. Potential aptazyme candidates are expected to be located in the bottom right quadrant of the scatter plot, displaying a large cleavage fraction in the absence of the targets, while simultaneously having a low cleavage fraction in their presence. On the other hand, ligand-insensitive sequences are expected to have the same cleavage fraction in the presence and absence of the ligand, and therefore be located along the diagonal of the scatter plot.
The scatter plots for both the folate and vitamin mixture libraries are shown in Fig. 5. As can be noted from this figure, most of the sequences do not exhibit a cleaved fraction in the presence of their ligand and are located on the x-axis rather than on the trend line (dashed line). We hypothesize this could be due to low efficiency in prefix replacement during the NGS library preparation in the presence of the ligand. Nevertheless, a high number of sequences are located in the bottom right quadrant of the scatter plots, which suggests the possible presence of potential aptazyme sequences.
Out of all the potential aptazyme candidates, we selected two sequences from the folate library and three from the vitamin mixture library as the five sequences to further investigate. We based this selection on two criteria: (i) the fold change of cleavage between – and + ligand conditions, which should be ideally equal or higher than 2 [1], and (ii) the preservation of the initial aptazyme’s structure. The first candidate from the folate library presents a 2.56 fold change. Meanwhile, the rest of the five candidate sequences display a fold change below 2. These moderate fold changes values could be explained by the limited number of iterative rounds we performed, 63 rounds, close to the minimum of 57 rounds that was required to find a functional aptazyme in the study of Townshend et al [1]. For this reason, we hypothesize that our candidate sequences present a low affinity for the ligand, which is reflected in the fold change. For all candidates, a qualitative and quantitative characterization is required to validate their nature as aptazymes.
| ID | Sequence | Library | Fold change | P-value |
|---|---|---|---|---|
| A | GCTGTCACAGGTCTTCTGTCTGACGAGTCCTTGGCCGGACGAAACAGC | Folate | 2.56 | 1.37 · 10-13 |
| B | GCTGTCACCGGAAATCTTTTTGTACGCTACGAGTCACCGTTATCCGGTCTGAAGAGTCCATCGGTGGGACGAAACAGC | Folate | 1.48 | 9.57 · 10-8 |
| C | GCCTGTCATCGGATTCCATCCGGTCTGACGAGTCCTTGGCGGGACGAAACAGC | Vitamins | 1.88 | 1.68 · 10-16 |
| D | GCTGTCACCGGAAGTCTTTTTGTACGCTACGAGTCACCGTTATCCGGTCTGAAGAGTCCATCGGTGGGACGAAACAGC | Vitamins | 1.67 | 8.75 · 10-49 |
| E | GCTGTCACCGGAGTATTCTTGTTTGCTAATTGCGTCATCCGTTCCGGTCTGATGAGTCCAGTGGGGGACGAAACAGC | Vitamins | 1.63 | 7.87 · 10-12 |
Achievements
- DRIVER reaction conditions were established and 63 rounds were reached.
- A high-throughput assay based on next-generation sequencing was used to characterize enriched sequences in the library mixtures. Five potential aptazyme sequences were selected based on their fold change of cleavage between - and + ligand conditions, and the conservation of the aptazyme structure.
Future prospects
Characterization of the potential vitamin-sensitive aptazymes
Validation of the NGS results is still necessary as this could not yet be performed. We suggest performing an Urea-PAGE assay in the absence and presence of the ligands for all the selected sequences. Aptazymes evolved using the vitamin mixture should be tested for each of the vitamins in the mixture, independently. Binding affinity can be further validated using a surface plasmon resonance assay as described by Townshend et al. in the original DRIVER manuscript [1]. Moreover, the binding selectivity of the aptamer sequences can be challenged by including structurally related molecules to the vitamin mixture. The chosen aptazymes should not respond to these analog molecules.
Improvement of aptazyme binding affinity
Based on the results obtained from these previous suggested experiments, additional evolution rounds can be performed to increase the sensitivity towards the target ligands. The ratio of cleavage and uncleavage selection rounds can be adjusted depending on the final application of the aptazyme. For instance, repeating two cleavage rounds after one uncleavage round can be done to bias selection towards producing sensors with higher cleavage fractions.
Partnership iGEM Vilnius-Lithuania 2021
The goal of our project is the creation of an accessible, quantitative, and modular rapid diagnostic test capable of diagnosing the diversity of vitamin deficiencies. Our project currently targets water-soluble vitamins: B1, B2, B6, B9, and B12. Yet, new biosensors are required for vitamins A, C, D, E, and K, in order to include the detection of these protein-bound, non-water-soluble vitamins in our test. We partnered with the iGEM Vilnius-Lithuania 2021 team, who created an aptamer prediction software, to generate sequences with potential affinity for vitamin A’s carrier protein retinol binding protein 4. We used electrophoretic mobility shift assays (EMSA) under physiological compatible conditions in order to evaluate the binding capability of potential ssDNA and RNA aptamer sequences.
Our partnership with iGEM Vilnius-Lithuania 2021 provided us the opportunity to deepen into the modular aspect of our test, by working together in the development of aptamer sequences with affinity to the vitamin A carrier retinol binding protein 4, (RBP4). By experimentally validating the capability of their predictive software, we would be able to include the detection of protein-bound, non-water-soluble vitamins, in our diagnostic test. For them, our experimental results could validate their predictive software and generate experimental data on affinity for future improvements.
The Vilnius team modified their software in order to be able to generate RNA aptamer predictions, and fed it with the crystallographic structure of the human RBP4. They generated both ssDNA and RNA aptamers for us to experimentally validate. For this, we used an electrophoretic mobility shift assay to potentially demonstrate and quantify the affinity of these sequences for RBP4.
Computational analysis of RBP4 apo- and holo-proteins
Before conducting affinity experiments, we performed a 50-ns molecular dynamic (MD) simulation of our protein in NAMD [2]. We addressed two potential limitations:
- Aptamers were predicted for the apo-RBP4 form. Since we are interested in the holo-RBP4 form of the protein (vitamin bound), can an apo-form binding aptamer also bind to the holo-form?
- Prediction is based on x-ray crystallography structures of the target protein. Crystallographic structures reveal only a static picture of the protein and can impose non-realistic physicochemical conditions. We therefore asked whether the protein allows for a proper aptamer binding under more realistic biological conditions?
Results showed very low mobility of both the apo- and holo-protein forms. Supplementary Fig. 5A shows the root mean square distances (RMSD) between the initial structural conformation and the succeeding frames of the simulation. For both the apo- and holo-forms the average RMSD was below 2 Å, 1.301 ± 0.195 Å, and 1.106 ± 0.195 Å, respectively. These low values confirmed the low protein mobility in any of their forms, suggesting that the protein-aptamer interactions computed using the crystallographic structure of the RBP4 protein, would probably not be affected by the protein mobility. Additionally, Supplementary Fig. 5B shows the superposition of the apo- and holo-RBP4 structures after the simulation, the RMSD computed was 1.56 Å (< 2 Å), supporting that the binding of vitamin A has a low impact on the structure of RBP4. This suggests that the aptamers generated or experimentally confirmed to bind to the apo-form could potentially form stable interactions with the retinol-bound form of the protein.
EMSA
In total, four 30 nucleotides sequences were computed and synthesized: a random ssDNA, a random RNA, a potential RBP4 ssDNA aptamer, and a potential RBP4 RNA aptamer. Random sequences were used as controls to show that aptamer-target interactions are sequence-specific. Incubation was carried at 37 °C for 40 minutes in a binding buffer formulated to mimic physiological ion composition. Details of our EMSA protocol can be consulted in our Experiment & Protocol page and the Notebook.
Native polyacrylamide gel shows no apparent binding of ssDNA aptamer sequence for RBP4. Supplementary Fig. 6 shows the resulting native polyacrylamide gel for ssDNA sequences. Lanes 4 and 5 show the expected migration patterns for the free RBP4_ssDNA and rand_ssDNA oligos, in the absence of target protein. For all samples no shift in the migration pattern is observed when compared to the free oligo, indicating the lack of binding affinity between RBP4 and the potential aptamer sequences under the used conditions. The same results are seen for random_RBP4 sequences.
Similar results were observed for the EMSA of RNA sequences in Supplementary Fig. 8. All samples migrated below the 50 nucleotides marker, with no visible shift for any of the ratios. Lane 3 and 4, containing rand_RNA oligos, are barely perceptible. We attribute this to improper homogenization of the oligos before addition to the reaction mix. These results show that under the used conditions, the potential RNA aptamer sequences displayed no affinity for RBP4.
Future prospects
The desired aptamer:protein binding interaction was not observed through EMSA’s at the used conditions. The EMSA methodology is highly protein and sequence dependent. As such, many physicochemical variables can have a strong impact on the experimental outcome. Therefore, different binding buffer compositions, oligos:protein ratios, incubation times and temperatures, and running times and voltages should be explored. During the development of the partnership, key limitations and improvement opportunities were identified for the predictive software of the iGEM Vilnius-Lithuania team. For more details on this and how our partnership developed through the extent of the iGEM competition, visit our Partnership page.
Aptazyme-regulated reporter gene expression using a cell-free system on paper support
Introduction
In keeping with our goal of making AptaVita an accessible and low-resource vitamin detection sensor, we engineered a genetic circuit to couple ligand-regulated cleavage of the aptazyme to a colorimetric output. A detailed explanation of the mechanism behind this genetic circuit can be found in our Design page. In short, the lacZ reporter gene, encoding for the β-galactosidase, was fused to an aptazyme sequence. Self-cleavage of the aptazyme in the absence of the ligand results in liberation of the ribosomal binding site (RBS), allowing for translation initiation of the β-galactosidase protein, which catalyzes the colorimetric substrate conversion of CPRG (yellow) to CPR (red). In the presence of the ligand, the aptazyme is stabilized preventing self-cleavage, the RBS is sequestered and translation is prohibited. As a proof-of-concept, we performed experiments to test the aptazyme-regulated lacZ expression in a cell-free system by using a known theophylline binding aptazyme [1], which can eventually be replaced by the vitamin-responsive aptazymes engineered through DRIVER.
Discovery of the suitable cell-free system for AptaVita
First of all, we identified the appropriate cell-free system for our desired substrate, CPRG. After several iterations of the engineering cycle, we found that the PUREexpress cell-free system is compatible with our envisioned CPRG colorimetric read-out. Throughout the engineering cycle of finding this compatible cell-free/substrate combination, we also learned that PUREfrex2.0, another PURE expression system, is incompatible with CPRG because of undesired side reactions that degraded the substrate even without the presence of the construct. Nonetheless, we also discovered that PUREfrex2.0 can be used in combination with an alternative β-galactosidase substrate X-gal. In this results page, some of the experiments were performed using PUREfrex2.0 and X-gal because of our limited resources in PUREexpress. For more information about the above mentioned engineering cycle visit our Engineering success page.
Enhancement of the biosensor performance through a new genetic construct
Comparing expression levels of the aptazyme-regulated construct to a control
To get a better understanding of our system, we compared the expression levels of our aptazyme-regulated construct (cRBS, BBa_K3806014) to a control part without an aptazyme switch. We used the BBa_I732017 part from the iGEM repository containing a lacZ gene and we fused a T7 promoter to enable expression. We performed the experiment in the absence of theophylline, which is expected to give the highest output signal of the system. The image shown in Fig. 6A, was taken after 2 hours of expression at 37 °C, and incubation with X-gal at room temperature overnight. The amount of X-gal produced was measured via absorbance at 600 nm (Fig. 6B).
Although the absorbance values of the cRBS construct indicated expression after 2 hours, it was 4.4 times lower compared to the control construct (Fig. 6). This means that the dynamic range of our biosensor was limited to this maximum expression level, while having a higher dynamic range would help to distinguish smaller differences in the ligand concentration.
Additionally, modeling the aptazyme-regulated lacZ expression in the cell-free system suggested that lowering the DNA concentration has a positive impact on the target ligand concentration discernibility. However, the model also showed that lowering the DNA concentration slows down the expression. Given the low expression levels of our cRBS construct under the current DNA concentration, we considered that decreasing it even further could impose an additional limitation on the system’s speed and maximum expression level. Therefore, to keep AptaVita as a rapid and quantitative test, we designed a new genetic construct that could lead to a higher and faster expression, which we intended to have.
Design of a new genetic construct
We hypothesized that slow expression of cRBS construct was caused by the antisense strand remaining attached to the RBS even after the aptazyme self-cleavage. Therefore, we designed a new construct in which only part of the RBS is sequestered by the antisense strand (semi-cRBS, BBa_K3806016). For a detailed explanation of this design, visit our Design page. We expressed both the cRBS and semi-cRBS constructs in the absence and presence of theophylline in PURExpress using CPRG as substrate. The product formation was determined by measuring the absorbance at 575 nm (Fig. 7).
Expression profiles after 4 hours of expression indicated that the semi-cRBS was better expressed than our standard cRBS construct (Fig. 7). We also noticed that, as expected, the presence of theophylline results in reduced expression compared to the absence of theophylline in both of the designs.
As an indication of both (i) the dynamic range represented by the absorbance difference between the smallest and highest ligand concentration (A0mM - A5mM), and (ii) the response time after 4 hours of expression, we calculated the area enclosed between the absorbance curves corresponding to 0 and 5 mM theophylline for both cRBS and semi-cRBS. For the cRBS design, the calculated area was 0.40 h x absorbance units (Au), whereas for the semi-cRBS the area was 1.03 h x Au, meaning a 2.57-fold increase. This indicates a larger dynamic range and faster response time when using the semi-cRBS construct. In addition, we determined the decrease in the expression level caused by the addition of 5 mM theophylline after 1.5 hours of expression as:
The semi-cRBS already shows a 33.3% decrease of expression after 1.5 hour, in contrast to the 6.1% decrease in the cRBS design, which confirms the improved response time of the system when using the semi-cRBS.
To further validate our results, both designs were also expressed in triplicates using PUREfrex 2.0 and X-gal as substrate (Fig. 8). We again omitted the addition of theophylline as this gives the highest signal output. The product formation was measured at 600 nm.
As expected, the expression level was significantly higher for the semi-cRBS design, confirming that the results were consistent even if a different PURE cell-free system and substrate were used. Overall, we improved the performance of our biosensor by modifying our initial design enabling a more rapid and quantitative read-out.
Ligand-responsive aptazyme cleavage in PURE
Testing the impact of theophylline addition on PURE
Our final goal is to determine the ligand concentration by a colorimetric read-out measured via absorbance. Therefore, we tested the impact of theophylline on the colorimetric output of the system. Concerned that addition of theophylline in millimolar concentrations could have an impact on the expression system, we first performed a control experiment to analyze whether the decrease in expression observed after theophylline addition may result from interference of theophylline with the PURE expression system rather than from regulation by the ligand-responsive aptazyme. For this purpose, we again expressed the BBa_I732017 part, containing a lacZ gene without an aptazyme sequence, in PUREfrex2.0 using X-gal as substrate and increasing concentrations of theophylline. With this, we could assess the impact of theophylline on the PURE expression system independently of the aptazyme switch influence.
After two hours of expression at 37 °C with and without theophylline added (Fig. 13), we observed a decrease in the color intensity when increasing theophylline concentrations (Fig. 9, samples 3-5). This suggested that higher concentrations of theophylline have an inhibitory effect on the PURE expression system. To examine whether the ligand responsive expression was only due to theophylline inhibition or a combined effect of inhibition and aptazyme cleavage, we set up our next experiment.
Cleavage characterization of the aptazyme-regulated genetic construct in PURE
From the DRIVER experiments, we have proven that the theophylline aptazyme shows ligand-regulated cleavage activity in the T7 RNA polymerase reaction buffer (Fig. 1). However, the cleavage activity also needs to be validated in the PURExpress cell-free system. Since we saw expression inhibition by adding theophylline, we decided to decouple transcription and translation on our cell-free system reactions to address whether ligand-dependent cleavage of the atazyme also contributes to the observed decrease of expression. To do so, we first expressed our newly designed semi-cRBS construct in PURExpress for 1 hour (Fig. 2, step 1). The product of this first step was used as input for a reverse transcription (RT) reaction to convert the transcribed mRNA to cDNA (step 2). These cDNA products were then amplified using two primer sets: one that flanks the cleavage site (uncleaved region), and one that flanks a region downstream the cleavage site (control region). In the former case, it should be noted that only the uncleaved mRNA fraction will be PCR amplified (step 3). As a final step, the PCR product was visualized on an agarose gel (step 4).
From this assay, we expected to see a lower band intensity with decreasing concentration of theophylline. This means that cleavage occurred and the forward primer flanking the cleavage site was unable to bind to the 5’ of the cDNA to produce a PCR product (Fig. 10 Scenario 1). However, if the ligand-dependent cleavage activity is not maintained, we expected to see the same band intensity regardless of theophylline concentracion (Fig. 10 Scenario 2). The region downstream the cleavage point was used as control for transcription, RT, PCR and loading in the agarose gel, as the band intensity should be similar between samples regardless of theophylline addition.
The resulting agarose gel after the above mentioned procedure can be seen in Fig. 11. ImageJ was used to quantify the intensity (I) of the gel bands, and the uncleaved fraction was determined by:
The computed uncleaved fractions were 36%, 52%, and 60% for 0, 1 and 5 mM theophylline, respectively. This means that the aptazyme preserves its theophylline-dependent cleavage activity in the PURExpress cell-free system (Fig. 10 Scenario 1). Additionally, the fusion of the lacZ gene to the aptazyme sequence does not hinder aptazyme self-cleavage.
From these results we can conclude that even though we observed theophylline-caused inhibitory effects on the expression system (Fig. 9), the aptazyme presents a ligand-dependent self-cleavage activity in the desired working conditions (Fig. 16). Therefore, we hypothesize that the decrease in absorbance seen in Fig. 14 is caused by a combination of both theophylline inhibition, and aptazyme-regulated expression. Since the intensity of the control band was similar regardless of the theophylline concentration, we hypothesize that inhibition occurs at translation level. Fortunately, vitamin concentrations in human blood are in the nanomolar range, therefore we suspect that inhibition of the system in this range is less likely to occur [3, 4]. This could be confirmed by assessing the impact of the vitamins at human-like concentrations on the PURExpress cell-free system.
Ligand-regulated expression on paper
Next, we tested the ligand-regulated expression system on a paper support. Cell-free system reactions were embedded and freeze dried onto paper discs. After 5 days of storing at room temperature, the paper discs were rehydrated with water or theophylline, and incubated at 37 °C for 24 hours. As expected, directly after rehydration, we did not see any color change of the paper disc for all of the tested conditions (Fig. 12, 0 h). After 24 hours of expression, however, we could see a slight color change from yellow to orange in the disc rehydrated with water and containing the semi-cRBS genetic construct (Fig. 12, 24 h, middle paper disc).
We hypothesize that the moderate color change can be due to the loss of enzymatic activity during the freeze-drying process. This hypothesis is supported by a preliminary test that we performed with the cRBS construct omitting the freeze-drying step, which showed a strong color change for non-theophylline conditions (Fig. 13, 24 h, middle paper disc). Moreover, we did not observe a color change for theophylline conditions (right paper disc). This experiment shows the potential of AptaVita to be used as a paper-based platform for vitamin detection. Further adjustments in the freeze-drying process are needed to achieve the desired expression level to enable sterile and inexpensive transport of the biosensor.
Achievements
- We discovered that PURExpress is a compatible cell-free system to use with our desired substrate CPRG.
- Comparison between our genetic construct (cRBS, BBa_K3806014) and the BBa_I732017 part revealed low expression levels of our design. We created a new design (semi-cRBS, BBa_K3806016) that showed an improved dynamic range compared to our previous design, enabling a more rapid and quantitative read-out.
- From the expression of the BBa_I732017 part in PURE with different theophylline concentrations we concluded that theophylline has inhibitory effects on the expression system. However, we showed that the aptazyme presents a ligand-dependent self-cleavage activity in the desired working conditions. This suggests that even though theophylline causes inhibition on the system, lacZ expression is also controlled by the theophylline-regulated aptazyme.
- Expression of the genetic circuit on a paper support after freeze-drying resulted in a limited color change. On the other hand, expression without prior freeze-drying led to a larger color change. Adjustments in the freeze-drying procedure are required to enable inexpensive, sterile, and abiotic distribution of the envisioned test.
Future prospects
Incorporation of novel vitamin biosensors
Our theophylline-regulated aptazyme genetic construct provides an insight into the potential of AptaVita as a robust vitamin detection system. New vitamin-responsive aptazymes engineered through DRIVER should be incorporated in the genetic construct. To confirm the desired ligand-dependent colorimetric read-out, expression of the vitamin-responsive apatzyme circuit should be validated in a cell-free system on paper support. After optimization of the genetic circuit and paper preparation process, patient samples could be tested on the biosensor.
Condition optimization to improve colorimetric read-out
Freeze-drying procedure
From the expression experiments using the theophylline construct, we saw that the impact of certain conditions on the aptazyme-regulated expression, for instance the freeze-drying procedure, could still be optimized for a better colorimetric read-out. As was observed in the paper expression results, only a slight color change was visible after freeze-drying of the paper. Adjustments on the freeze-drying procedure such as the freeze-drying time, or the addition of the substrate before freeze-drying or during rehydration, could potentially lead to an improved expression.
Magnesium-dependent aptazyme self-cleavage
The activity of most hammerhead ribozymes is Mg2+-dependent [5]. Preliminary results in our lab showed that the addition of extra rNTPs to the cell-free system reaction in order to chelate free Mg2+, has a positive effect on the stabilization of the aptazyme, resulting in less self-cleavage of the aptazyme in the absence of the ligand. Future experiments in which the Mg2+ concentration is varied could be used to gain knowledge on the optimum Mg2+ concentration to achieve the highest discernibility between vitamin concentrations.
Paper and humidity
Point-of-care tests need to be accurate in any environment in which they are used. This concern was also expressed by local doctors and health organizations. The point-of-care tests should then be robust enough so they can be exposed to different storage and operating conditions and still give reproducible results. In Uganda, our target country for implementation, there are months where the humidity reaches 70%. Health organizations and experts told us that previous tests failed due to their bad performance under humid conditions. Therefore, we developed a detailed set-up for future experiments that recreates the humidity conditions in Uganda. Unfortunately, due to time restrictions, we could not fill in the required safety documents to comply with the iGEM safety and security policies. Therefore, we established a detailed experimental set-up for future purposes without performing the experiment.
The aim of this experiment is to check: (i) how our test performs under humid conditions, and (ii) whether expression starts only after rehydration, and not during the storage time. To simulate the humidity conditions in Uganda, we would place our freeze-dried paper disks, containing the genetic construct, cell-free system, and CPRG substrate in the botanical garden of the Delft University of Technology. This garden has a room with an average temperature of 19-26 °C and 100% humidity. In the most positive outcome, the paper discs would not rehydrate within the storage time frame and would, after rehydration, show a similar expression as the control that is left in the lab (Fig. 14). This would give some first indications on whether our biosensor is able to be used under humid conditions. It is important to consider that this should be done after optimizing the freeze-drying conditions. A second optimization cycle might be needed to adjust the freeze-drying conditions for the paper-based biosensor to perform in a humid environment.
Paper-based microfluidic chip
To improve the user-friendliness of the biosensor, we are interested in detecting multiple vitamins by a single rehydration step of the paper disks. For this purpose, we looked into using a microfluidic approach on a laser-printed paper, which has been reported to be a successful strategy by Ghosh et al. [6]. We performed some pilot experiments to test whether the ink could create a hydrophobic layer by printing a pattern on chromatography paper consisting of two wells connected through a channel.
By heating-up the paper, we expected the ink to melt and impregnate the paper to create a hydrophobic channel that is able to create a microfluidic flow for hydrophilic substances. To easily follow the liquid trajectory, we tested our paper design with blue-colored milli-Q water. We observed that the ink creates a hydrophobic layer that conserves the water in the well and channel (Supplementary Fig. 8). Unfortunately, we did not see the water flow through the channels as we expected it to do. In order to fix this, future experiments should then focus on testing different channel thicknesses, channel length, optimal shape, and patterns and test the fluidics and flow speed of more patient sample related fluids and conditions to allow for a single rehydration step in our biosensor
Hardware
To make our test available at the point of care, we developed a frugal read-out device that measures absorbance over time to quantify the color-based outcome of the AptaVita test. For more information on the read-out device, visit our Hardware page
Blocking ambient light
For the measurements, we needed to ensure that the test box is opaque to minimize the ambient light coming through the test box and reaching light sensors. To this end, the test box and the inside of the embodiment were spray painted with black matte spray. We measured the ambient light that reached each sensor for four different conditions (Tab. 2). The results are shown in Fig. 16. For each sensor, no ambient light was measured in the fourth condition, which is the same as the test condition. Hereby, we show that the ambient light is sufficiently blocked.
| Closed embodiment | Closed test box | Test cassette in test box | |
|---|---|---|---|
| Condition 1 | - | - | - |
| Condition 2 | - | - | + |
| Condition 3 | - | + | + |
| Condition 4 | + | + | + |
Proof-of-concept for concentration measurements
Our final goal is to measure the CPR concentration over time during a test. The measured concentration curve can be linked to the vitamin concentration of a sample. In this section, we demonstrate the ability of our device to measure colorimetric differences between CPR concentrations at the pH of blood plasma (pH = 7.4) [7].
First, we measured the absorbance spectrum of CPR on a paper disc in a plate reader to validate that CPR absorbs light in the range of wavelengths emitted by the LEDs (550 - 600 nm) of our read-out device. This was the case, as we observed an absorbance peak at 579 nm (Fig. 17).
To measure the absorbance at a range of CPR concentrations, we linearly diluted the CPR to 0.25 mM, 0.5 mM, 0.75mM, 1mM, and 1.25 mM (see Fig. 17). For each concentration, the absorbance was measured in duplicate in the plate reader, and triplicate in the read-out device. As a blank measurement, we used a paper disc that was hydrated with the same buffer as used for the dilution. In the plate reader, we observe a linear relation with an R2 of 0.97 (Fig. 18a). This shows that the Beer-Lambert law is valid for this range of CPR concentrations. In the read-out device, we also observed a linear relation between the absorbance and the CPR concentration with an R2 of 0.95 (Fig. 18b). This demonstrates that the device is capable of distinguishing different CPR concentrations through absorbance measurements.
Achievements
- We show that our device is able to reduce the ambient light to a level which cannot be measured by the read-out device.
- We demonstrate that the hardware is capable of distinguishing different CPR concentrations through absorbances measurements.
Future Prospects
- By improving the positioning of the LEDs and reorganizing the cables, the amount of light that reaches the sensor can be increased, so that a larger part of the range of the sensor is used, thereby increasing the precision and sensitivity of the measurements.
- Although we showed that our device is capable of performing measurements on the same disc over a time period, we have not tested if the device can measure a change of CPR concentration over time. This would be the next step in testing our read-out device. In the end, the measured temporal evolution of the CPR concentration could be compared to reference curves that correspond to known vitamin concentrations. The vitamin concentration of the test sample could be inferred from the comparison.
References
- Townshend, B., Xiang, J. S., Manzanarez, G., Hayden, E. J. and Smolke, C. (2021). A multiplexed, automated evolution pipeline enables scalable discovery and characterization of biosensors. Nat Commun, 12, 1437.
- Phillips, J. C., Braun, R., Wang, W., Gumbart, J., Tajkhorshid, E., Villa, E., Chipot, C., Skeel, R. D., Kalé, L., & Schulten, K. (2005). Scalable molecular dynamics with NAMD. Journal of computational chemistry, 26(16), 1781–1802. https://doi.org/10.1002/jcc.20289
- Center for Disease Control and Prevention, World Health Organization, Nutrition International, UNICEF. Micronutrient survey manual. Geneva: World Health Organization; 2020. Licence: CC BY-NCSA 3.0 IGO. https://www.who.int/publications/i/item/9789240012691
- Albaugh, V.L., Williams, D.B., Aher, C.V., Spann, M.D., & English, W.J. (2021). Prevalence of thiamine deficiency is significant in patients undergoing primary bariatric surgery. Surgery for Obesity and Related Diseases, 17(4), 653-658. https://doi.org/10.1016/j.soard.2020.11.032
- Naghdi, M. R., Boutet, E., Mucha, C., Ouellet, J., & Perreault, J. (2020). Single Mutation in Hammerhead Ribozyme Favors Cleavage Activity with Manganese over Magnesium. Non-coding RNA, 6(1)., 14. https://doi.org/10.3390/ncrna6010014
- Ghosh, R., Gopalakrishnan, S., Savitha, R., Renganathan, T., & Pushpavanam, S. (2019). Fabrication of laser printed microfluidic paper-based analytical devices (LP-µPADs) for point-of-care applications. Scientific reports, 9(1), 1-11.
- Aoi, W., & Marunaka, Y. (2014). Importance of pH homeostasis in metabolic health and diseases: crucial role of membrane proton transport. BioMed research international, 2014, 598986. https://doi.org/10.1155/2014/598986