Introduction

After identifying the optimal RPA primers-crRNA pair (ModF-GalR-crRNA1) and assessing the preliminary limit of detection (LOD), our next step was to quantify the SHERLOCK system to determine if the fluorescence signal corresponds to the amount of target Prymnesium parvum DNA and, consequently, the number of algal cells. To achieve this, the following steps were performed:

1. Ensuring RPA Proportionality

When comparing the amount of the target DNA before and after RPA for different initial target DNA concentrations the proportion of amplified DNA concentration to initial DNA concentration should be the same for a wide range of initial target DNA concentrations.

2. Check if SHERLOCK tests are:

   • Repeatable (this means that the fluorescence signal for the same concentration of target DNA falls within a specific range, which remains consistent across all SHERLOCK tests;

   • Proportional (that means more target DNA gives a more intense fluorescence signal).

3. Determining the Number of ITS1-5.8S-ITS2 Regions in the Prymnesium parvum Genome

This information is crucial for correlating the amount of target DNA to the number of algal cells. The ITS1-5.8S-ITS2 region is a multicopy gene, and its copy number varies among species and even within individuals of the same species.

Step 1: Make RPA proportional

INTRODUCTION

An assay was conducted to compare how varying primer concentrations impact the amplification rate and fluorescence intensity in SHERLOCK detection. The method used is based on the approach presented in the Jonathan S. Gootenberg article. [1]

EXPERIMENT

Methods:

1. RPA reactions were set up as advised in Kellner’s protocol [2], with different primer and input DNA concentrations, alongside a positive control with SynDNA. For template dilutions, a PCR product coming from an algal culture was used. Dilutions of 200 nM, 2nM, 20 pM and 200 fM were tested, alongside the following primer concentrations: 960 nM, 480 nM, 240 nM and 120 nM. 

The following samples were prepared:

Table 1. SHERLOCK reactions components – RPA quantification.

2. SHERLOCK was conducted according to Kellner’s protocol on a 384-well plate. [2] (one change: the assay was conducted in two instead of four replicates for each sample)

Results analysis – protocol:

1. Averaging Fluorescence Intensity

   • Each sample was measured in duplicates to ensure accuracy and reliability. The final fluorescence intensity for each sample was then calculated by averaging the duplicate measurements.

   • These averaged results were then used to create representative bar graphs, which are presented in Figures 2, 5, 8, and 11.

2. Subtracting Negative Controls

   • For each sample, the fluorescence intensity of the negative control was subtracted from the average intensity obtained in the previous step.

   • Based on these adjusted results, Figures 3, 6, 9, and 12 were created and a logarithmic function was fitted.

Results:

*note: The assay needed to be manually extended, which caused a bump on the graph (fluorescence intensity versus time), but this does not impact the final results.

Figure 1. SHERLOCK - 960 nM.

Figure 2. Final fluorescence intensity to Prymnesium parvum DNA concentration - 960 nM

Figure 3. Final fluorescence intensity to log(Prymnesium parvum DNA concentration) - 960 nM - function fit.

Figure 4. SHERLOCK - 480 nM.

Figure 5. Final fluorescence intensity to Prymnesium parvum DNA concentration - 480 nM

Figure 6. Final fluorescence intensity to log(Prymnesium parvum DNA concentration) - 480 nM - function fit.

Figure 7. SHERLOCK - 240 nM.

Figure 8. Final fluorescence intensity to Prymnesium parvum DNA concentration - 240 nM

Figure 9. Final fluorescence intensity to log(Prymnesium parvum DNA concentration) - 240 nM - function fit.

Figure 10. SHERLOCK - 120 nM. 

Figure 11. Final fluorescence intensity to Prymnesium parvum DNA concentration - 120 nM

Figure 12. Final fluorescence intensity to log(Prymnesium parvum DNA concentration) - 120 nM - function fit.

Conclusions:

Based on the SHERLOCK plots (Figures 1, 4, 7, and 10), neither 200 fM nor 20 pM target DNA was detected at a 120 nM RPA primer concentration. This is the first primer concentration we rejected, as it did not provide sufficient sensitivity.

Among the other tested concentrations (240, 480, and 960 nM), the 960 nM concentration emerged as the most favorable option. This conclusion is based on the observation that it exhibited the highest R² value, suggesting that the relationship between initial DNA concentration and fluorescence intensity is more proportional at this concentration compared to the lower concentrations of primers. 

Additionally, while all tested concentrations showed signs of fluorescence signal saturation at 200 nM, the saturation effect was least pronounced at 960 nM. This indicates that the 960 nM concentration may provide a broader dynamic range for fluorescence detection, allowing for more accurate quantification of target DNA concentrations.

It is worth noting that although the 200 fM concentration was not detected in the 960 nM set, we decided not to be overly concerned about this result. It is possible that an error occurred during the preparation of the RPA mix, which could have led to the absence of a detectable signal.

An additional test will be conducted to further validate if 960 nM primer concentration is optimal and to determine the range of DNA concentrations for which the relationship between log₂[DNA] and fluorescence intensity is linear.

Step 2: SHERLOCK – test in duplicate

INTRODUCTION

In this assay, we wanted to check:

• Repeatability

To achieve this, we conducted the SHERLOCK assay in duplicate across two separate runs. Both runs utilized the same RPA mixtures, ensuring that any observed differences in fluorescence signals would be attributable solely to the kinetics of the SHERLOCK reaction. The goal of this test was to evaluate whether the average final fluorescence intensity corresponding to a specific target DNA concentration could reliably quantify target DNA in test samples, potentially eliminating the requirement for a standard curve in each assay.

• Proportionality (that means more target DNA gives a more intense fluorescence signal)

And also to determine the final LOD and range of DNA concentrations for which the relationship between log₂[DNA] and fluorescence intensity is linear.

To cover the points above we used the following target DNA concentrations: 200 nM, 20 nM, 2 nM, 200 pM, 40 pM, 20 pM, 10 pM, 5 pM, 2 pM, 1 pM, 200 fM, 20 fM and 2 fM. Later, we compared the fluorescence intensity in measurement points and the final fluorescence intensity.

EXPERIMENT

Methods:

1. RPA reactions were set up, with different template DNA concentrations: 200 nM, 20 nM, 2 nM, 200 pM, 40 pM, 20 pM, 10 pM, 5 pM, 2 pM, 1 pM, 200 fM, 20 fM and 2 fM. We used 960 nM RPA primers concentration.

2. The SHERLOCK assay was conducted as in Step 1 on the following samples in 2 separate runs (duplicate of the test):

Table 2. SHERLOCK reactions components – SHERLOCK tests

Results

Figure 13. SHERLOCK - 1a.

Figure 14. SHERLOCK - 1b.

Figure 15. SHERLOCK - 2a.

Figure 16. SHERLOCK - 2b.

Conclusions

1. The SHERLOCK reaction demonstrated poor repeatability, with inconsistent fluorescence intensities observed in samples from the same RPA mix. The LOD also varied between test duplicates, with one test showing an LOD of 1 pM and the other 200 fM. As a result, it is not possible to establish a universal standard curve for the assay.

Example: 

In Figure 13 going from most intense fluorescence, we have target concentrations: 20 nM, 40 pM, 2 nM, 200 pM

In Figure 15: 2 nM, 20 nM, 200 nM, 40 pM, 200 pM

2. Despite optimizing the RPA primer concentration, it remains challenging to define a range of target DNA concentrations that correlate proportionally with the final fluorescence intensity or at any other measurement point. As shown in the SHERLOCK plots (Figures 13-16), the curves often intersect, and the slopes of the series corresponding to specific target DNA amounts do not consistently align. Consequently, at this stage, reliable quantification of the assay is not feasible.

If a range of concentrations proportional to fluorescence intensity could be established (point 2), then creating a standard curve for each test would provide reference points for quantifying DNA concentration.

However, we assessed the final limit of detection (LOD) to be 1 pM. Additionally, we successfully detected Prymnesium parvum in the genomic sample isolated from an algal culture. These findings suggest that SHERLOCK can be used for screening the presence of Prymnesium parvum in water with a detection limit of 1 pM.

It is recommended to repeat this test to rule out any potential operator error. If the results remain inconsistent between runs and lack proportionality, it would be advisable to conduct the test using a primer concentration of 480 nM. Unfortunately, this could not be attempted due to time constraints.

According to article [3], there may be other reasons why RPA is challenging to quantify. The authors suggest, based on their model and laboratory experiments, that commercial RPA kits are optimized for rapid amplification, but the amplified DNA might show a more proportional relationship with the initial target DNA concentration when the kit is diluted (e.g., 2x). Therefore, testing with a diluted RPA kit is recommended to explore this possibility further. It would be also advised to check the amount of DNA after RPA using more direct methods (not SHERLOCK where we rely on 2 proteins’ activity – polymerase T7 and Cas13).

In summary, to achieve quantification with the SHERLOCK assay, the following steps are recommended:

• Evaluate RPA proportionality using direct methods, such as fluorescence dyes [3]:

  • Repeat the evaluation using the same primers and target DNA concentrations as in Step 1.

  • Test with a 2x-diluted RPA kit. [3]

• Once a proportional relationship between the initial target DNA concentration and post-amplification DNA is confirmed across a range of initial concentrations, proceed with SHERLOCK tests as outlined in Step 2.

Step 3: Calculating the amount of DNA to the amount of Prymnesium parvum’s cells

INTRODUCTION

Although we were unable to make our system quantifiable, we still aimed to estimate the number of golden algae cells detectable at a limit of detection (LOD) of 1 pM.

We conducted an analysis revealing that Prymnesium parvum from the Oder River has ITS1-5.8S-ITS2 sequences exclusively on chromosome 20, present in two haplotypes: one with 8 repetitions of the ITS1-5.8S-ITS2 sequence and another with 4 repetitions. Statistically, this means that 25% of the Prymnesium parvum population should have 8 copies of the ITS1-5.8S-ITS2 sequence (haplotypes 4 + 4), 50% should have 12 copies (8 + 4), and 25% should have 16 copies (8 + 8). For a detailed analysis, please refer to the entry dated 06.08 in the GenomeLab section of our Notebook page.

CALCULATIONS

Assuming there are 4 Prymnesium parvum cells in 1 L of water, there should be a total of 48 copies of the ITS1-5.8S-ITS2 sequence, distributed as follows: 1 cell with 8 copies, 2 cells with 12 copies each, and 1 cell with 16 copies.

To convert this to moles:

For future teams

We believe that uncovering the kinetics of the RPA reaction and Cas13a collateral activity will enhance the reliability and repeatability of scientific research. By understanding these kinetics, it will become feasible to model the reactions in silico. Examples of such models might include:

• Choosing the optimal RPA primer concentration without the trial-and-error approach we used. Only the best concentrations would be tested in the lab;

• Predicting the LOD for different RPA primer-crRNA pairs;

• Identifying a time point when SHERLOCK becomes quantifiable.

References:

[1] J. S. Gootenberg, O. O. Abudayyeh, M. J. Kellner, J. Joung, J. J. Collins, and F. Zhang, “Multiplexed and portable nucleic acid detection platform with Cas13, Cas12a, and Csm6.,” Science, vol. 360, no. 6387, pp. 439–444, Apr. 2018, doi: 10.1126/science.aaq0179.

[2] M. J. Kellner, J. G. Koob, J. S. Gootenberg, O. O. Abudayyeh, and F. Zhang, “SHERLOCK: nucleic acid detection with CRISPR nucleases.,” Nat Protoc, vol. 14, no. 10, pp. 2986–3012, Oct. 2019, doi: 10.1038/s41596-019-0210-2.

[3] P. Valloly and R. Roy, “Nucleic Acid Quantification with Amplicon Yield in Recombinase Polymerase Amplification,” Anal Chem, vol. 94, no. 40, pp. 13897–13905, Oct. 2022, doi: 10.1021/acs.analchem.2c02810.