Microfluidic Nano-Scale qPCR Enables Ultra-Sensitive and Quantitative Detection of SARS-CoV-2

Round 1

Reviewer 1 Report

The present manuscript presents a more sensitive method to quantify SARS-CoV-2. The data is well presented and the design of the assay is sound. I have only some minor remarks. 

According to the MiQE guidelines, Cq (quantification cycle) should be presented instead of Ct. The authors should also discuss how the Cq value was obtained as different methods exist to calculate Cq values.

It would be interesting to see a comparison of the quantitative outcome of the new assay compared to the clinical diagnostic samples. A Bland Altman analysis for example could reveal the agreement and possible bias between both methods. 

The bar charts in figure 2 should be changed to dot plots which show the actual technical replicates and give the reader a better idea on the real variation between these replicates.

I don't understand why Figure 3 C and Figure4C are showing both the Cq values as well as the viral copies per µl. The latter is calculated from the Cq, hence it is logical that the points fall on a straight line, but it could look as a very strong correlation, which has nothing to do with it. I would suggest to showe his data on a 1D plot with eigher Cq or copise/µl on the single axis. 

Author Response

Response to Reviewer 1 Comments

The present manuscript presents a more sensitive method to quantify SARS-CoV-2. The data is well presented and the design of the assay is sound. I have only some minor remarks.

We thank you for your time in reviewing our manuscripts and appreciate your constructive feedback. Please see our responses to your points:

Point 1:

• According to the MiQE guidelines, Cq (quantification cycle) should be presented instead of Ct. The authors should also discuss how the Cq value was obtained as different methods exist to calculate Cq values.

Response 1:

We have revised all instances of Ct values to Cq values, per your suggestion. Note that in the Fluidigm qPCR platform, the Ct and Cq values are equivalent. We have also added the following explanation in the methods part:

“To calculate the Cq (cycle of quantification) of each sample, a global threshold is automatically calculated and applied to all the samples in the chip. The crossing point at which the amplification curve of each sample crosses the threshold line is determined as the Cq. For comparability with other studies, we note that the Cq is numerically equivalent to the Ct value in our system.“

Point 2:

• It would be interesting to see a comparison of the quantitative outcome of the new assay compared to the clinical diagnostic samples. A Bland Altman analysis for example could reveal the agreement and possible bias between both methods.

Response 2:

 Thank you for this suggestion. We obtained the quantitative data from the clinical diagnostic lab and compared the obtained Ct values from the two methods. We have added our analysis in Supplementary Figure 3, which shows the samples ranked by the Cq values obtained from the microfluidics RT-qPCR, and their comparative Cq values across the two methods. We have also added the following to the manuscript:

“Exploring how the two methods compared, we found a relative consistency between the ranking of Cq values between the microfluidic RT-qPCR test and the clinical diagnostic test, despite the expected difference in Cq values due to the addition of a pre-amplification step (20 cycles) in the microfluidic method (Figure S3).”

“All samples with negative clinical diagnosis showed no detectable signal in the clinical diagnostic qPCR test.”

Point 3:

• The bar charts in figure 2 should be changed to dot plots which show the actual technical replicates and give the reader a better idea on the real variation between these replicates.

Response 3:

We have revised the bar charts in Figure 2 to also show each individual technical replicate with detected Cq Values as well as the variation between replicates. We have also changed our description in the results to reflect how many replicates were detected:

“With preamplification, the viral N gene was detectable in Twist RNA at 0.5 copies/µl (N1 assay, 9 out of 18 replicates) and 5 copies/µl (N2 assay, 18 out of 18 replicates), whereas without preamplification, we were only able to detect viral material in a few replicates at 5000 copies/µl (N1 assay, 3 out of 18 replicates; N2 assay, 2 out of 18 replicates). No viral material was detected below 5,000 copies/µl (Figure 2A)”.

The fewer number of replicates detected in the condition without pre-amplification is a further indication that the chance of detecting SARS-CoV-2 is lower without a pre-amplification step.

Point 4:

• I don't understand why Figure 3 C and Figure4C are showing both the Cq values as well as the viral copies per µl. The latter is calculated from the Cq, hence it is logical that the points fall on a straight line, but it could look as a very strong correlation, which has nothing to do with it. I would suggest to showe his data on a 1D plot with eigher Cq or copise/µl on the single axis.

Response 4:

We have now revised both Figures 3C and 4C to represent the data using 1D plots. Figure 3C now shows violin plots with a boxplot to highlight the difference in viral copies between samples with positive and negative clinical diagnosis. Figure 4C is now a 1D point plot which shows the viral load for each sample in the original experiment, as well as how the classification of each of those samples had changed after one or two freeze cycles.

Reviewer 2 Report

Title: Microfluidic nano-scale qPCR enables ultra-sensitive detection of SARS-CoV-2

Author: Xin Xie et al.

The authors of the article have presented great research and shown through their results that their approach using preapplication step has reduced the standard LOD and made possible the detection of very low viral load. However, some of the claims of the author needs experimental observation or restated in absence of them. I suggest minor changes and request following comments to be fulfilled in order to accept the article for publication. These are the comments I request author to consider for enhancing the succinctness and readability of this research for the user of this journal in my opinion.

 Comments:

1. Please consider using wording as commercially available microfluidic platform instead of directly branding Fluidigm microfluidic platform as it would provide weightage to science along with devices.
2. Cost effective and low false negative are two main point authors claimed in their article. They have successfully shown ability to detect low viral load but cost effectiveness has not shown aptly.
3. Please add the comment about 86/87 positive. How the one sample not detected positive is corelated with the modified three step approach.
4. The range of Ct value of clinically diagnosed negative sample and and ct value range of same sample using authors technique will be necessary to mention for clear comparison as clinical diagnostic standard recommend range of Ct value as a threshold. If not available then please make a statement.
5. Author mentioned that “By performing a large number of replicates, our method can robustly detect SARSCoV-2 and therefore identify samples that are false negatives by the clinical diagnostic procedure”. Please provide the impact of processing large number of replicates (cost and time).
6. Authors shown that” additional freeze cycles may compromise the power of our method to detect low viral loads”. It is not surprising but It is not shown if freezing cycle results in degradation of viral load by starting with load much higher that LOD and then presenting the load remain after the application of each freezing cycle.
7. Authors claims are contradictory to their own use of nanoscale in title as they said” stochastic variation in the number of viral RNA molecules present in the small reaction volumes used for RT”. Please suggest if this could potential limitation of working with small sample volume reagents.
8. Authors stated that Their approach is cost-effective strategy with several advantages but did not compared the reagent requirement and associated cost difference between standard technique and their approach.
9. The use of pre amplification to enhance the LOD is being known previously and published before June 2020. I would suggest to add the necessary references and distinction so that readers can see relative advantages of authors’ approach.

Author Response

Response to Reviewer 2 Comments

The authors of the article have presented great research and shown through their results that their approach using preapplication step has reduced the standard LOD and made possible the detection of very low viral load. However, some of the claims of the author needs experimental observation or restated in absence of them. I suggest minor changes and request following comments to be fulfilled in order to accept the article for publication. These are the comments I request author to consider for enhancing the succinctness and readability of this research for the user of this journal in my opinion.

We thank you for your time in reviewing our manuscripts and appreciate your constructive feedback that helped us improve the paper and enhance its fit for the readership of Processes. Please see our responses to your comments:

Comment 1:

1. Please consider using wording as commercially available microfluidic platform instead of directly branding Fluidigm microfluidic platform as it would provide weightage to science along with devices.

Response 1:

Thank you for this comment. We have replaced any direct branding of “Fluidigm” with a “microfluidic RT-qPCR” throughout the main text, and are now only referring to “Fluidigm” in the materials and methods part.

Comment 2:

2. Cost effective and low false negative are two main point authors claimed in their article. They have successfully shown ability to detect low viral load but cost effectiveness has not shown aptly.

Response 2:

Please see response to comment #8.

Comment 3:

3. Please add the comment about 86/87 positive. How the one sample not detected positive is correlated with the modified three step approach.

Response 3:

We obtained the Cq values from the clinical diagnostic analysis (Cq value, or quantification cycle value, is the standardized terminology by MiQE guidelines; per the suggestion of reviewer 1, we have replaced all instances of Ct with Cq in our manuscript, as in our system, the two are numerically equivalent), and checked this specific sample. The sample had Cq = 36.4 (N gene) and Cq = 38.2 (ORF1 gene), which indicates a very low viral load. Thus, the failure of detection for this sample may be due to RNA degradation during transportation or the heat inactivation process, which could bring the viral load below our LoD. We have incorporated the following in our results section to address this point:

“The one Pos_Neg sample had a relatively high Cq value (36.4 for the N-gene and 37.2 for the ORF1 gene) in the clinical diagnostic test; therefore, it is likely that sample degradation during transport or the heat inactivation process may have brought its viral load below the LoD.”

Comment 4:

4. The range of Ct value of clinically diagnosed negative sample and and ct value range of same sample using authors technique will be necessary to mention for clear comparison as clinical diagnostic standard recommend range of Ct value as a threshold. If not available then please make a statement.

Response 4:

According to the data from the clinical diagnostic lab, all the relative samples showed no amplification in the standard qPCR and had no detectable signal, and thus have no valid Cq values. To address this point, we added a comment in our result section:

“All samples with negative clinical diagnosis showed no detectable signal in the clinical diagnostic qPCR test. “

Comment 5:

5. Author mentioned that “By performing a large number of replicates, our method can robustly detect SARSCoV-2 and therefore identify samples that are false negatives by the clinical diagnostic procedure”. Please provide the impact of processing large number of replicates (cost and time).

Response 5:

Please see response to comment #8.

Comment 6:

6. Authors shown that” additional freeze cycles may compromise the power of our method to detect low viral loads”. It is not surprising but It is not shown if freezing cycle results in degradation of viral load by starting with load much higher that LOD and then presenting the load remain after the application of each freezing cycle.

Response 6:

We thank you for pointing this out. We recognize that the way this statement was phrased: “additional freeze cycles may compromise the power of our method to detect low viral loads” can be misleading, as there is no direct link between additional freeze cycles and the power of this method to detect viral loads. What we aim to show in this repeated freeze-thaw experiment is that we can reproducibly detect most of the ‘false negative’ samples, except for those with extremely low viral loads. This is consistent with the idea that additional freeze-thaw cycles result in RNA degradation which may increase the false negative rate, especially when the viral load is low. We have now edited the text to reflect this.

In fact, the adverse effects of freezing-thaw cycles on viral detection have been reported by previous study: a small increase in Ct value (0.41 to 0.82 cycle) after one freezing-thaw cycle was observed in a high viral load sample, indicating sample degradation [1]. The small changes can compromise the detection of the extremely low viral load samples that are close to the LoD. We also incorporated this reference, which suggests that the failure of detecting samples with extremely low viral load in repeated experiments can be due to the freeze-thaw process, in our edit. The overall change in this section is as follows:

“This suggests that additional freeze cycles may lead to an increased false negative rate, likely due to degradation of viral RNA resulting in copy numbers below the LoD. This is consistent with prior observations of increased Cq values following one freeze-thaw cycle [27]”

Comment 7:

7. Authors claims are contradictory to their own use of nanoscale in title as they said” stochastic variation in the number of viral RNA molecules present in the small reaction volumes used for RT”. Please suggest if this could potential limitation of working with small sample volume reagents.

Response 7:

We believe there is a misunderstanding of the context in this sentence. The original description is: “This suggests that samples with extremely low viral titers close to the LoD could fail to be consistently detected, likely due to factors such as sample degradation or stochastic variation in the number of viral RNA molecules present in the small reaction volumes used for RT.” There is no contradiction to the use of nano-qPCR since we are not referring to the nano qPCR step here. The stochastic effect refers to the RT (reverse transcription) step, which is the same as the standard qPCR diagnosis procedure. Supposing we have a very small copy number of viral material (e.g. 0.2 copy/ul in 50 ul of RNA extract), when 5 ul of RNA extract are used for the RT reaction, there is a higher chance for not pipetting the viral material. This is common for every method when dealing with extremely low viral load samples.

Comment 8:

8. Authors stated that Their approach is cost-effective strategy with several advantages but did not compared the reagent requirement and associated cost difference between standard technique and their approach.

Response 8:

Detailed cost analysis is beyond the scope of this paper; however, we now provide the approximate cost of per sample analysis based on the reagents cost in the UAE and the cost of the microfluidic chip. We also add a statement on the advantage of having room for additional assays in the microfluidic chips at no extra cost, which can be utilized to test for various viral and bacterial coinfections. Furthermore, each sample is loaded only once in the microfluidic system, and then analyzed against all the assays simultaneously, meaning that there is no extra cost or time for multiple assaying or for running technical replicates. We have added the following to the discussion:

“Based on our experience, and considering the fact that reagents cost more in the UAE than in Europe or the US, the cost of SARS-CoV-2 testing per sample using the 192.24 microfluidic chip and the design from this study (9 replicates for N1 and N2, 6 replicates for RP) is approximately $17, including reagents and the chip, but excluding labor cost. This is approximately four times cheaper than the current cost of standard SARS-CoV-2 testing in the UAE. Lowering the number to three technical replicates for each of the N1, N2 and RP assays would allow for four additional assays, which can be utilized to diagnose other viral infections using three technical replicates (or six additional assays with two technical replicates) at no extra cost, bringing down the cost per sample per assay even further. Since each sample is loaded only once and then analyzed against 24 assays simultaneously, no extra time is required for performing multiple assays of replicates.”

Comment 9:

9. The use of pre amplification to enhance the LOD is being known previously and published before June 2020. I would suggest to add the necessary references and distinction so that readers can see relative advantages of authors’ approac

Response 9:

Thank you for this suggestion. We have updated the manuscript to cite a prior study that incorporated the use of target pre-amplification in the detection of SARS-CoV [2]. The LoD of the enhanced real-time PCR method was 100-fold higher than the standard real-time PCR assay, which is consistent with our observation. We add a corresponding discussion part to elaborate this. We also added several references in the introduction part, which applied a pre-amplification step in detecting and analyzing various types of low input genomic samples: including ancient DNA [3], circulating tumor DNA [4] and Viral genetic materials [2,5]. This will enrich the background of the introduction. Overall, the changes are as follows:

Introduction 

“A target-specific preamplification step has been successfully incorporated in a number of studies to detect and analyze various types of samples with limited amount of genomic materials including virus in human samples [23] or in drinking water [24], circulating tumor DNA in blood [25], and even ancient DNA samples [26]”

Discussion 

“Using serial dilutions of positive controls, we demonstrate a 1,000-fold improvement in detection sensitivity of the microfluidic qPCR system when adding the preamplification step, consistent with a previous study in which a target preamplification step was found to enhance the LoD by 100-fold in detecting SARS-CoV [23]."

[1] Lin Li, Xiao Li, Zhendong Guo, Zhongyi Wang,Ke Zhang,Chao Li, Changjun Wang, and Shoufeng Zhang. Influence of Storage Conditions on SARS-CoV-2 Nucleic Acid Detection in Throat Swabs. The Journal of Infectious Diseases. 2020, jiaa272.

[2] Lok Ting Lau, Yin-Wan Wendy Fung, Freda Pui-Fan Wong, Selma Sau-Wah Lin, Chen Ran Wang, Hui Li Li, Natalie Dillon, Richard A. Collins, John Siu-Lun Tam, Paul K.S. Chan, Chen G. Wang, and Albert Cheung-Hoi Yu. A real-time PCR for SARS-coronavirus incorporating target gene pre-amplification. Biochemical and Biophysical Research Communications. 2003, 312: 1290–1296.

[3] Stefania Del Gaudio, Alessandra Cirillo, Giovanni Di Bernardo, Umberto Galderisi,Theodoros Thanassoulas, Theodoros Pitsios, and Marilena Cipollaro. Preamplification Procedure for the Analysis of Ancient DNA Samples. The Scientific World Journal. 2013, Article ID 734676.

[4] Jackson JB, Choi DS, Luketich JD, Pennathur A, Ståhlberg A, Godfrey TE. Multiplex Preamplification of Serum DNA to Facilitate Reliable Detection of Extremely Rare Cancer Mutations in Circulating DNA by Digital PCR. Journal of Molecular Diagnosis. 2016, 18(2):235-43.

[5] Parker JK, Chang TY, Meschke JS. Amplification of viral RNA from drinking water using TransPlex™ whole-transcriptome amplification. Journal of Applied Microbiology. 2011,111(1):216-223.