Based on the previous results, it has been confirmed that EXPAR lacks robustness, TWJ-SDA has excellent specificity but insufficient amplification efficiency, and multistep-SDA can achieve high amplification efficiency while minimizing NC compared to conventional linear SDA and TWJ-SDA.
However, for multistep-SDA, amplification has been almost undetectable for low-concentration targets, making it unsuitable as a standalone amplification system for POIROT. Therefore, we pursued the establishment of an amplification system that achieves robustness, specificity, and sufficient amplification efficiency by using the amplification products of TWJ-SDA as the target for multistep-SDA.
To start, we used a TWJ-SDA template that produces the sequence to be the target for multistep-SDA to attempt amplification starting from biomarker1. The sequences used were those modified with a quencher at the 3' end for multistep-SDA and templates suggested to be suitable based on the designed sequence. For the polymerase and nickase, we utilized the conditions tuned in the TWJ_2cycle, specifically 0.063 U/µL Bst LF and 0.12 U/µL Nb.BbvCI.
Additionally, due to the subtle differences in optimal conditions for Cas and nickase, we used a custom-formulated buffer called UTokyo Buffer, which was designed by referencing the compositions of both.
For more Details on the preparation of UTokyo Buffer, see Experiments.
For a discussion of orthogonality with Cas, see CRISPR-Cas.
We refer to the TWJ template as template0, and the templates for the first, second, and third stages of multistep-SDA as template1, template2, and template3, respectively. Experiments were conducted using the TWJ-SDA product as the target under conditions where the first, second, and third stages of SDA were combined. Additionally, for comparison, we also performed experiments with only TWJ, TWJ > 2step-SDA, and 2step-SDA alone.
The fluorescence changes were plotted below.
Even when multiple stages were combined, similar changes in fluorescence intensity were observed as in the case of TWJ-SDA alone. This suggests that the amplification products of TWJ are not effectively connecting to the subsequent stages.
Analysis from the Dry Lab indicated that the amplified ssDNA is more stable when attached to the 5' side of the previous template, suggesting that the process may not advance to the next step until all previous converters are fully occupied on the 5' end.
To address this bottleneck, we redesigned the template to enhance its stability at the 3' end. By conducting experiments with the redesigned template, we expect to resolve the connectivity problems between TWJ-SDA and multistep-SDA. Furthermore, since it is believed that increasing the template concentration in later stages would also be effective, experiments should be conducted to verify this as well.
Using the redesigned sequences, we performed amplification reactions with only TWJ-SDA and with the connection of TWJ-SDA to 3step-SDA. This allowed us to verify whether the problems in the connectivity between TWJ-SDA and multistep-SDA had been resolved.
The fluorescence changes were plotted below.
Note that fluorescence was measured with SYBR Green Ⅰ for TWJ and MB for the 3step-SDA-linked one.
In systems where a large number of templates and primers coexist, as confirmed by multistep-SDA, it is expected that MB can be used to track specific chemical species 1.
In the case of TWJ-SDA, a linear amplification curve was obtained, whereas when TWJ-SDA was connected to 3step-SDA, an increase in amplification speed was observed around the 40-minute mark.
Compared to the 1. preliminary experiment, it was confirmed that the redesign of the template enabled the successful functioning of multistep-SDA using the products of TWJ-SDA as primers.
Previously, template concentrations had been template1: 10 nM, template2: 50 nM and template3: 100 nM, according to the multistep-SDA reference paper 2. However, due to the changes in the buffer and the connection from TWJ-SDA, it is necessary to retune the template concentration. This system, being a combination of multiple stages of SDA, makes the ratio of template concentrations a critical factor for success, as mentioned previously.
2step-SDA was linked after TWJ-SDA. template0 (template of TWJ-SDA) concentrations were 1, 5 and 10 nM, with template2/template1 and template1/template0 concentration ratios of 2, 3, 4 and 5 fold, respectively. Experiments were performed at 10 pM Biomarker1 and 0 M and fluorescence intensity was measured using SYBR Green Ⅰ.
The fluorescence changes were plotted below.
Here, the \(r\) in the caption indicates the ratio of concentrations between templates.
As the concentration of template0 increases and the ratio of template concentrations becomes larger, the slope of the amplification curve increases for both the presence and absence of the target. In other words, the amplification speed is greater.
When the total template concentration is high, the background fluorescence intensity also increases.
When the concentration of template0 is relatively high at 10 nM, the differences between the presence and absence of the target become unclear, or it is observed that the fluorescence intensity and amplification speed are greater in the absence of the target. This indicates that non-specific amplification for the target is dominating.
We decided to adopt conditions where the concentration of template0 is set at 5 nM and the ratio between the templates is 2, as these conditions facilitate faster amplification in the presence of a biomarker at 10 pM compared to the absence of a biomarker, without overly increasing the template concentration. That is, 5 nM template0, 5 nM helper, 10 nM template1, 20 nM template2, 40 nM template3.
The multistep-SDA we are using allows for flexibility in the number of stages that can be added. While increasing the number of stages is expected to improve amplification efficiency, it is essential to avoid overly complex mechanisms.
Based on the experimental results above, it is expected that connecting TWJ-SDA with 3step-SDA yields the desired amplification efficiency within approximately 30 to 60 minutes. Under the tuning conditions, experiments were conducted using a mechanism that connects the TWJ-SDA with 2step and 3step-SDA, adding microRNA of fM order.
Additionally, a new MB was designed to detect the products of the 2step-SDA, and in this experiment, fluorescence was measured using the MB for both the 2step and 3step processes.
The fluorescence changes were plotted below.
As expected, the 3step-SDA showed superior amplification efficiency. Although the 2step mechanism is simpler and more desirable, its amplification was slow and diminishes the difference in target concentration. Therefore, we decided to connect the TWJ-SDA and the 3step-SDA as the amplification system for POIROT.
Furthermore, to ultimately produce dsDNA that activates Cas3 or Cas12a, it is necessary to perform an elongation reaction using the final products of the multistep-SDA to produce dsDNA.
To activate Cas3 or Cas12a in the final step, it is necessary to produce dsDNA. We designed the following mechanism to produce the desired dsDNA through an elongation reaction using the final products of the multistep-SDA as primers.
In this mechanism, similar to the elongation reactions using previous templates, the primer hybridizes to the 3' end of the template, and then the DNA polymerase elongates the 3' end of the primer. The key difference is that the product does not contain a nicking site, allowing for the production of dsDNA without introducing nicks.
The ssDNA that serves as the template for this mechanism is referred to as ds-template.
Cas3 or Cas12a recognizes a double-stranded PAM sequence and a single-stranded spacer sequence for activation. It has been reported by Yoshimi et al. that even partial PAM sequences combined with a complete spacer sequence can still elicit some level of activity 3. Therefore, a sequence complementary to the spacer sequence was incorporated into the ds-template.
In other words, the ds-template sequentially contains a complementary sequence to the primer, a PAM sequence, and one strand of the complementary spacer sequence, starting from the 3' end. The final product is dsDNA, and tracking it using the fluorescence intensity of MB is not feasible in principle. Therefore, we conducted the experiments using SYBR Green Ⅰ. In the final stage, we performed two types of experiments: elongation only after 1step-SDA and elongation using the ds-template.
The fluorescence changes were plotted below.
In both experimental conditions, the background fluorescence intensity was high, and subsequent fluorescence changes were not monotonic. Particularly in experiments using only the ds-template, only elongation reactions occurred. Fluorescence measurements using SYBR Green, which emits fluorescence through non-specific nucleic acid recognition, suggested that changes in fluorescence intensity were buried in background noise, making accurate quantification difficult. There are no known simple methods for sequence-specific recognition using fluorescent dyes or aptamers. By connecting to the Cas system, we expect to be able to quantify dsDNA in a sequence-specific manner for the first time. Therefore, we decided to perform the tuning for the part that produces dsDNA in section CRISPR-Cas. Please refer to the details here.
In previous experiments, nucleic acids diluted with TE were used as targets. POIROT targets miRNA in tear fluid, and in order to further increase the reliability of POIROT as an amplification system, we conducted amplification using miRNA diluted in artificial tear fluid as a target, taking into account the effects of differences in solution types on amplification. The conditions were those that had been tuned so far. In addition, comparative experiments were conducted using nucleic acids diluted with TE.
The fluorescence changes were plotted below.
When miRNA diluted with artificial tear fluid or TE was used as a target, the following amplification curve was obtained.
Under the tuned conditions, we were able to detect target miRNAs down to about 1 fM.
In particular, it was confirmed that amplification whose speed was depending on the target concentration occurred in the range of 1 fM to 100 fM.
When comparing artificial tear and TE, amplification occurs faster in artificial tear at all target concentrations. This is presumed to be due to changes in enzyme activity due to differences in salt concentration. The composition of the artificial tear is 13 nM KCl, 68 nM NaCl (pH 7.0 - 8.0) 4. However, it was confirmed that the relative relationship did not change and that the amplification reaction occurred without any problem in the artificial tear.
Through the experiments conducted thus far, the overall structure of the amplification system for POIROT has been determined. Specifically, we initiate with the product of TWJ-SDA, followed by 3step-SDA, and use the final product as a primer for the elongation reaction, which produces dsDNA that activates either Cas3 or Cas12a.
At this stage, by using a sequence complementary to the spacer in the ds-template, we ensure that dsDNA is produced only when the elongation reaction occurs, which in turn activates Cas. Additionally, we expect that using the Cas-based method allows for accurate quantification of the final product, dsDNA.
Wang, C., & Yang, C. J. (2013). Application of molecular beacons in real-time PCR. In M. D. Teintze (Ed.), Molecular Beacons (pp. 45-59). Springer. https://doi.org/10.1007/978-3-642-39109-5_3
Komiya, K., Noda, C. & Yamamura, M. (2024). Characterization of Cascaded DNA Generation Reaction for Amplifying DNA Signal.New Gener. Comput. 42, 237-252. https://doi.org/10.1007/s00354-024-00249-2
Yoshimi, K., Takeshita, K., Kodera, N., Shibumura, S., Yamauchi, Y., Omatsu, M., ... & Mashimo, T. (2022). Dynamic mechanisms of CRISPR interference by Escherichia coli CRISPR-Cas3. Nature Communications, 13 (4917), 1-11. https://doi.org/10.1038/s41467-022-32618-0
Santen Ct,.Ltd. (n.d.). Soft Santear. https://www.santen.com/content/dam/santen/japan/pdf/
healthcare/eye/products/otc/soft_santear.pdf