A One-Step RT-PCR-Coupled Cysteamine-Functionalized Gold Nanoparticle Assay for Colorimetric Detection of Tobacco Mosaic Virus

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This manuscript established a one-step RT-PCR coupled cysteamine-functionalized gold nanoparticle colorimetric detection platform for rapid visual identification of Tobacco Mosaic Virus. The authors systematically optimized nanoparticle modification, primer design, reaction dilution, incubation temperature and time, evaluated the assay’s sensitivity, specificity and practical detection performance via field crop samples, and built a quantitative regression model relying on CIELAB color parameters. The research integrates nucleic acid amplification and nano-colorimetric sensing, avoids expensive electrophoresis equipment, shortens detection turnaround time, and provides a low-cost on-site detection scheme for plant viral diseases, which has prominent application value in agricultural plant protection. The experimental design is complete, data charts are sufficient, and the logical framework is clear. However, multiple defects need to be thoroughly revised before acceptance. Detailed review comments are listed below:

1.The comparison between bare citrate gold nanoparticles and cysteamine-modified Au@Cys only provides visual images and CIELAB A value statistics, but lacks quantitative spectral support of LSPR peak shifts under positive/negative samples. The authors should supplement the UV–Vis absorption peak difference data of bare AuNPs and Au@Cys after reacting with RT-PCR products and salt induction, to more intuitively explain the superiority of cysteamine surface modification.

2.In Figure 3, four groups of thiolation-modified primer combinations were tested, and dual thiolated primers achieved the optimal color discrimination effect. However, the manuscript did not explain the cost difference of thiol-labeled primers and the repeatability of synthesis batches. The authors need to discuss the economic feasibility of dual thiolated primers for field popularization, and supplement repeatability verification data of three batches of synthesized thiolated primers.

3.The linear quantitative regression equation built by CIELAB mean A value only verified plasmid standard templates, lacking recovery test data of actual field leaf matrix. The authors should conduct spike recovery experiments by adding known copy numbers of TMV plasmids into tobacco/tomato leaf RNA extracts, calculate recovery rate and coefficient of variation to verify the quantification accuracy of the colorimetric method in complex plant crude extract matrix.

4.The manuscript claims the LOD of Au@Cys colorimetric assay is 102 copies/reaction, while traditional agarose gel electrophoresis only reaches 103 copies/reaction. However, the two detection systems use independent readout modes; the authors need to set unified template gradient and consistent RT-PCR system, conduct parallel comparison test, and systematically analyze the reasons for the higher sensitivity of nanoparticle colorimetry compared with gel imaging.

5.All optimization experiments use 3 M NaCl as the aggregation inducer, but the interference of ions in plant RNA extraction buffer and RT-PCR reaction buffer on nanoparticle aggregation is not evaluated. The authors should design control groups containing residual RLT buffer, dNTP and Mg²⁺ to exclude matrix ion interference and prove the anti-interference ability of the optimized Au@Cys detection system.

6.The incubation condition optimization only investigated room temperature, 37 °C, 45–60 °C and 5–30 min incubation time, but did not explore the storage stability of prepared Au@Cys working solution. The authors need to supplement the colorimetric detection performance of Au@Cys stored for 1, 3, 7 and 14 days, and clarify the valid storage period of the nanoparticle reagent for on-site detection.

7.The field sample test results show all 73 tobacco and tomato samples are TMV positive, lacking negative field control samples collected from TMV-free crop planting areas. The authors should add leaf samples from TMV non-host crops or healthy virus-free seedlings as field negative controls to further verify the assay’s field specificity and eliminate false positive interference from plant endogenous substances.

8.The structural characterization of AuNPs and Au@Cys only provides TEM, zeta potential and UV–Vis data in supplementary Figure S1, without dynamic light scattering hydrodynamic particle size change data after combining with RT-PCR dsDNA amplicons. The authors should supplement DLS particle size distribution curves of Au@Cys before and after adsorbing target DNA to visually confirm the formation of DNA protective layer on nanoparticle surface.

9.The manuscript only used EAPV and TSWV as heterologous virus controls for specificity verification, lacking common co-infecting plant viruses in local tobacco and tomato fields. The authors should add other prevalent tobamoviruses or cucumber mosaic virus as cross-reaction controls to fully prove the high specificity of the TMV coat protein targeting primer pair.

10.The digital color extraction code is uploaded to GitHub, but the manuscript does not describe the repeatability of color value extraction under different light intensities and shooting angles. The authors should perform repeated color value extraction under three different lighting environments, calculate intra-group and inter-group variation coefficients, and verify the stability of digital quantitative color analysis.

Author Response

Reviewer 1:

1. The comparison between bare citrate gold nanoparticles and cysteamine-modified Au@Cys only provides visual images and CIELAB A value statistics, but lacks quantitative spectral support of LSPR peak shifts under positive/negative samples. The authors should supplement the UV–Vis absorption peak difference data of bare AuNPs and Au@Cys after reacting with RT-PCR products and salt induction, to more intuitively explain the superiority of cysteamine surface modification.

Response to reviewer: We sincerely thank the reviewer for this valuable suggestion. In this study, the primary objective of the colorimetric assay is to distinguish positive and negative RT-PCR samples based on the visible color response, which has already been quantitatively evaluated using CIELAB color analysis. Nevertheless, we agree that UV–Vis spectroscopy can provide complementary spectral evidence for the observed color changes.

Following the reviewer's suggestion, UV–Vis absorption spectra of both bare citrate-capped AuNPs and cysteamine-modified Au@Cys were measured under all experimental conditions and are presented in the revised Fig. S1. The spectra show that both nanoparticles exhibit an LSPR peak around 520 nm before salt induction. After NaCl addition, spectral fitting revealed that the LSPR peak of bare AuNPs shifted to 534 nm, whereas Au@Cys exhibited only a slight shift to 528 nm. The smaller red shift of Au@Cys is consistent with the reduced aggregation observed from the colorimetric images and CIELAB analysis, further supporting the improved stability provided by cysteamine surface modification. Accordingly, Fig. S1 and the corresponding discussion in Section 3.1 (lines 211-217) have been revised.

 

2. In Figure 3, four groups of thiolation-modified primer combinations were tested, and dual thiolated primers achieved the optimal color discrimination effect. However, the manuscript did not explain the cost difference of thiol-labeled primers and the repeatability of synthesis batches. The authors need to discuss the economic feasibility of dual thiolated primers for field popularization, and supplement repeatability verification data of three batches of synthesized thiolated primers.

Response to reviewer: We sincerely appreciate the reviewer’s constructive suggestions regarding the economic feasibility of dual-thiolated primers for field popularization and the assessment of synthesis batch repeatability.

Although three independent synthesis batches would provide additional statistical confidence, only two batches were available within the revision period. Therefore, here we evaluated the consistency of the colorimetric assay using two independent batches of thiolated primers synthesized during our study. It is also worth noting that thiol-modified oligonucleotides were synthesized using highly standardized commercial platforms with stringent quality control by Integrated DNA Technologies (IDT).

The experimental data obtained from the two available independent batches demonstrated high reproducibility. Specifically, both batches confirmed that while the dual-thiolated primers provided the most distinct colorimetric discrimination between positive and negative samples, the single thiolated forward primer also yielded highly consistent and clear discrimination (as illustrated in the figure below).

                                                            

Consequently, we have revised Fig. 3 by incorporating colorimetric data obtained from two independent batches of thiolated-primers. Additionally, we have expanded our discussion in the revised manuscript (lines 469-472) to address the economic viability and practical trade-offs of these primers for field application. The added text reads as follows:

“Considering the additional synthesis cost associated with dual-thiol modification, employing a single thiolated forward primer, which has demonstrated clear color discrimination between positive and negative samples, represents a cost-effective alternative for practical field applications.”

3. The linear quantitative regression equation built by CIELAB mean A value only verified plasmid standard templates, lacking recovery test data of actual field leaf matrix. The authors should conduct spike recovery experiments by adding known copy numbers of TMV plasmids into tobacco/tomato leaf RNA extracts, calculate recovery rate and coefficient of variation to verify the quantification accuracy of the colorimetric method in complex plant crude extract matrix.

Response to reviewer: We sincerely thank the reviewer for this valuable suggestion. Spike recovery experiments require a virus-free plant RNA background to accurately evaluate the quantitative performance of the proposed assay in a complex biological matrix. However, despite extensive sampling, all available tobacco and tomato leaf samples, including asymptomatic seedlings, were confirmed to be TMV-positive and were therefore unsuitable as negative matrices for spike recovery experiments.

Therefore, total RNA extracted from passion fruit leaves, a non-host plant of TMV, was employed as an alternative plant RNA matrix. The absence of TMV was confirmed by RT-PCR analysis, with no detectable amplification observed on agarose gel electrophoresis. Specifically, 200 ng of total RNA extracted from passion fruit leaves was added to each RT-PCR reaction containing 10⁷, 10⁵, and 10³ copies of the TMV plasmid and to a negative control using nuclease-free water as the template. Three independent RT-PCR trials were performed, and the subsequent colorimetric assay was conducted in duplicate for each run (n=6).

To evaluate the reliability of our linear quantitative regression equation for quantifying viral infection in field samples, the color values extracted from the CIELAB color space of the RT-PCR products were then applied to the regression equation (Fig. 6) to back-calculate the TMV plasmid copy numbers. The recovery rate (R%) and coefficient of variation (CV%) were then determined using the following equations:

Recovery rate R%=[(plasmid copy number calculated from the regression equation/actual plasmid copy number in RT-PCR)] x 100%

Coefficient of variation CV%=(SD/mean) x 100% where SD is standard deviation of calculated plasmid copy number across replicates and mean is average value of calculated plasmid copy number

The results of experiment are summarized in the bellow:

Actual plasmid copy number in RT-PCR (Log10)	Color value (mean ± SD, n=6)	Plasmid copy number calculated from the regression equation (mean ± SD, n=6)  (Log10)	Coefficient of variation (CV%)	Recovery rate (R %)
7	163.01 ± 0.78	7.29 ± 0.22	3.07%	104.2%
5	156.88 ± 1.18	5.55 ± 0.34	6.06%	111 %
3	150.04 ± 1.07	3.6 ± 0.3	8.46%	119.9%
Negative control	136.94 ± 1.54	-0.14 ± 0.44

The color values extracted from the CIELAB color space using our open-source code in above table were not significantly different with color values in Fig. 6 (can see in Fig. S4C of revised manuscript for detail graph) and exhibited a small standard deviation, indicating that the plant RNA matrix does not significantly interfere with the color value results as well as reproducibility of the colorimetric assay. The negative control yielded a calculated plasmid copy number of           10^-0.14 with a minimal standard deviation, confirming that the plant RNA background does not produce false-positive results. Furthermore, the average recovery rates fell within the acceptable range of 80-120% for biological samples (Wang et al., 2023). However, the coefficient of variation (CV%) indicated that the stability of TMV quantification via the regression equation varies with the viral load. Specifically, quantification is highly stable under high TMV infection levels (10⁷ copies, CV% < 5%) and became less stable though acceptable under moderate and low TMV infection levels (10⁵ and 10³ copies, CV% < 10%). To improve viral quantification accuracy in future studies, we intend to construct a regression equation with higher resolution by incorporating intermediate plasmid standard concentrations, specifically supplementing points at 10^7.5, 10^6.5, 10^5.5, 10^4.5, 10^3.5, 10^2.5, 10^1.5 copies. Alternatively, alternative mathematical models may be developed to better capture the relationship between color values and viral loads.

In summary, these data demonstrate that our TMV detection procedure is highly stable and robust for qualitative assessment based solely on the color values; however, precise quantification using the linear regression equation can be acceptable but needs further optimization, particularly at low viral titers. Nonetheless, in practice, our regression equation remains highly valuable for disease management. It establishes a critical diagnostic threshold to identify mildly infected plants that are otherwise difficult to distinguish via visual inspection alone. Furthermore, from an agricultural management perspective, any infected plants—even those harboring low viral loads—must be promptly identified and culled, rendering the exact quantification of low viral titers secondary to its primary detection capability.

According to the new data obtained, a substantial revision has been made to the Results section (Part 3.5) of the revised manuscript.

4. The manuscript claims the LOD of Au@Cys colorimetric assay is 10² copies/reaction, while traditional agarose gel electrophoresis only reaches 10³ copies/reaction. However, the two detection systems use independent readout modes; the authors need to set unified template gradient and consistent RT-PCR system, conduct parallel comparison test, and systematically analyze the reasons for the higher sensitivity of nanoparticle colorimetry compared with gel imaging.

Response to reviewer: We sincerely thank the reviewer for highlighting this important point. We completely understand the reviewer’s concern regarding the comparability of the two analytical methods.

We would like to clarify that in this research we routinely performed parallel analyses using both agarose gel electrophoresis and the colorimetric assay on the exact same amplification products (lines 343-350 in revised manuscript). Specifically, immediately following the RT-PCR reaction, a 1µL aliquot of the RT-PCR product was loaded for agarose gel electrophoresis, while another 1µL aliquot of the identical sample was subjected to the colorimetric assay. This parallel approach eliminates variations in template concentration or amplification efficiency, rendering the comparative sensitivity data highly robust.

The underlying mechanisms explaining why the colorimetric assay exhibits superior sensitivity over conventional gel electrophoresis have been thoroughly elaborated in the Discussion section of the revised manuscript (lines 478-485). The added text reads as follows:

“The superior sensitivity of the colorimetric assay stems from the unique optical properties of gold nanoparticles (AuNPs). AuNPs possess exceptionally high extinction coefficients, up to a million times greater than those of organic dyes used in agarose gel electrophoresis, allowing them to intensely absorb and scatter light even at nanomolar concentrations [22,23]. Furthermore, while agarose gel electrophoresis is a separation-based method that can cause diffusion and smearing of DNA molecules (resulting in signal loss), the colorimetric assay is performed in-tube. This in-tube approach allows the reaction to maintain equilibrium without diluting the signal.”

5. All optimization experiments use 3M NaCl as the aggregation inducer, but the interference of ions in plant RNA extraction buffer and RT-PCR reaction buffer on nanoparticle aggregation is not evaluated. The authors should design control groups containing residual RLT buffer, dNTP and Mg²⁺ to exclude matrix ion interference and prove the anti-interference ability of the optimized Au@Cys detection system.

Response to reviewer: We sincerely thank the reviewer for valuable comments and suggestions. We would like to clarify that negative controls, using nuclease-free water as the template in the RT-PCR reactions, were systematically included in all experimental batches. These negative controls contained the identical components as the test samples (including dNTPs, Mg²⁺, primers and the enzyme) and were subsequently analyzed via the colorimetric assay using the exact same protocol. Therefore, any potential contribution from RT-PCR reagents is inherently accounted for in the negative controls and does not affect the comparison between positive and negative samples.

Regarding the RNA extraction process, the isolated RNA was eluted in nuclease-free water, and its quality was verified using UV-absorbance spectrophotometry. The A₂₆₀/A₂₈₀ ratio was approximately 2.1, confirming high purity and the absence of significant contamination. Furthermore, a minimal volume of RNA input was utilized (200 ng of RNA, corresponding to approximately 0.1-2µL of the eluate per 10µL RT-PCR reaction), resulting in a high dilution factor of the original sample matrix. Consequently, the potential interference from residual RLT buffer is negligible and can be safely discounted.

6. The incubation condition optimization only investigated room temperature, 37 °C, 45–60 °C and 5–30 min incubation time, but did not explore the storage stability of prepared Au@Cys working solution. The authors need to supplement the colorimetric detection performance of Au@Cys stored for 1, 3, 7 and 14 days, and clarify the valid storage period of the nanoparticle reagent for on-site detection.

Response to reviewer: We sincerely thank the reviewer for this valuable suggestion. Following the reviewer's recommendation, we evaluated the colorimetric performance of Au@Cys stored at 4 °C for 1, 3 and 7 days. The corresponding data were presented in the figure below as mean ± s.e.m. (n = 3 replicates derived from three independent RT-PCR trials). As demonstrated in the results, no significant deterioration in colorimetric performance was observed over this period, demonstrating good short-term storage stability. Based on these results, the prepared Au@Cys working solution can be stored at 4 °C for at least 7 days without compromising assay performance.

Evaluation beyond 7 days requires real-time storage and could not be completed within the revision period. A systematic investigation of extended storage stability under various storage conditions is currently underway and will be reported in future work.

 

7. The field sample test results show all 73 tobacco and tomato samples are TMV positive, lacking negative field control samples collected from TMV-free crop planting areas. The authors should add leaf samples from TMV non-host crops or healthy virus-free seedlings as field negative controls to further verify the assay’s field specificity and eliminate false positive interference from plant endogenous substances.

Response to reviewer: We sincerely thank the reviewer for the constructive comments and suggestions. During the assay development, the use of negative field samples to verify specificity and rule out potential interference from endogenous plant substances was thoroughly considered. However, we were unable to obtain true negative samples from the field, and even commercially sourced, asymptomatic seedlings tested positive. To address this, we utilized field leaf samples from a non-host plant of TMV as an alternative control. As shown in Supplementary Figure 4A-B, we performed RT-PCR on field sample of passion fruit leaves infected with East Asian Passiflora virus AO strain (EAPV-AO). The resulting RT-PCR products were evaluated using both agarose gel electrophoresis and the colorimetric assay. The data demonstrate that our primer pair does not cross-amplify non-target viruses such as EAPV-AO as well as does not amplify non-specific amplicons from host plant RNA. In addition, the proposed assay employs a TMV-specific primer pair in a one-step RT-PCR rather than oligo(dT)-based reverse transcription. In silico BLAST analysis against the NCBI nucleotide database indicated that the primer pair is highly unlikely to yield non-specific cDNA from host plant RNA including tobacco, tomato and passion fruit plants under our optimized RT-PCR conditions. Consequently, only cDNA of TMV is theoritically generated during reverse transcription step, thereby reducing background noise from endogenous plant matrix components. The colorimetric results were also consistent with agarose gel electrophoresis (Fig. S4A-B). Furthermore, as seen in Supplementary Figure 4C, the colorimetric assay showed non-significant difference between RT-PCR reactions using standard plasmids as template in the absence or presence of field plant RNA background, further confirming the assay's high specificity and its robustness against interference from complex matrix components including plant endogenous substances in field samples.

8. The structural characterization of AuNPs and Au@Cys only provides TEM, zeta potential and UV–Vis data in supplementary Figure S1, without dynamic light scattering hydrodynamic particle size change data after combining with RT-PCR dsDNA amplicons. The authors should supplement DLS particle size distribution curves of Au@Cys before and after adsorbing target DNA to visually confirm the formation of DNA protective layer on nanoparticle surface.

Response to reviewer: We sincerely thank the reviewer for their valuable comments and suggestions. Following the reviewer's recommendation, we measured the hydrodynamic size of Au@Cys before and after incubation with RT-PCR dsDNA amplicons using dynamic light scattering (DLS), and the corresponding size distribution curves have been shown in figure bellow:

 

No significant increase in hydrodynamic diameter was observed. This result is likely attributable to the relatively low surface density of DNA molecules and their flexible conformation on the nanoparticle surface, which produce only a minimal change in hydrodynamic size that is below the resolution of DLS (Deng et al., 2013; Xu and Craig, 2007).

Furthermore, UV-Vis data upon salt induction in Fig. S1D demonstrated that Au@Cys incubated with RT-PCR positive sample only experiences a minor shift, indicated the good dispersion under high ionic strength conditions, whereas Au@Cys with RT-PCR negative sample (without dsDNA amplicons) rapidly aggregated and loss of the characteristic absorption peak. The enhanced salt stability is consistent with the formation of a protective DNA layer on the nanoparticle surface, which effectively suppresses salt-induced aggregation through electrostatic and steric stabilization.

In addition, representative TEM images showed that the Au@Cys incubated with dsDNA appeared to be more loosely distributed, with slightly larger separations between neighboring nanoparticles than those observed for Au@Cys alone (Fig. S1B, C). Although this observation is qualitative, it agrees well with the enhanced colloidal stability observed in UV-Vis data, suggesting that the adsorbed dsDNA molecules introduced steric stabilization to the nanoparticle surface.

Therefore, although DLS did not reveal a significant change in hydrodynamic diameter, the combined evidence from TEM results and UV-Vis consistently supports the successful adsorption of dsDNA onto the Au@Cys nanoparticles.

9. The manuscript only used EAPV and TSWV as heterologous virus controls for specificity verification, lacking common co-infecting plant viruses in local tobacco and tomato fields. The authors should add other prevalent tobamoviruses or cucumber mosaic virus as cross-reaction controls to fully prove the high specificity of the TMV coat protein targeting primer pair.

Response to reviewer: We would like to thank the reviewer for highlighting the important point regarding the specificity of the TMV coat protein-targeting primer pair. This issue was thoroughly considered during our experimental design. Experimental specificity against EAPV and TSWV has already been demonstrated in this study. However, regarding other viruses closely related to TMV, obtaining TMV-free samples concurrently infected with other tobamoviruses proved challenging due to the exceptionally wide host range and highly contagious nature of TMV. To address this limitation, we performed a sequence alignment of our primers against the genomes of several prevalent tobamoviruses using Blast function in NCBI, including Tobacco mild green mosaic virus (TMGMV), Tomato mosaic virus (ToMV), Tomato Mottle Mosaic Virus (ToMMV), Cucumber green mottle mosaic virus (CGMMV), Tomato brown rugose fruit virus (ToBRFV), and Pepper mild mottle virus (PMMoV). The alignment results demonstrate that our TMV-specific primer pair is highly unlikely to yield non-specific amplicons under our optimized RT-PCR conditions. Although experimental validation using these viruses would further strengthen the study, the corresponding viral isolates were not available during the present work. We have added this issue in the discussion section of  the revised manuscript (lines 485-496).

10. The digital color extraction code is ueploaded to GitHub, but the manuscript does not describe the repeatability of color value extraction under different light intensities and shooting angles. The authors should perform repeated color value extraction under three different lighting environments, calculate intra-group and inter-group variation coefficients, and verify the stability of digital quantitative color analysis.

Response to reviewer: We sincerely thank the reviewer for their valuable comments and suggestions. In the previous manuscript, the exact light intensity and shooting angle for the colorimetric assay were in advertently omitted. We have now included these conditions in the revised manuscript (lines 525-527).

Furthermore, in accordance with the reviewer’s suggestion, we performed the colorimetric assay on 10 field samples, together with 10⁷ copies of the TMV plasmid as a positive control and nuclease-free water as a negative control. These assays were performed under four representative illumination levels: a baseline of 600 lux (representing typical indoor laboratory lighting) and three alternative intensities of 260, 420, and 690 lux corresponding to lower and higher light levels relative to the baseline. The camera-to-sample distance (40 cm) and shooting angle (approximately 90°) were kept constant throughout image acquisition to eliminate geometric variation. Consequently, only illumination intensity was evaluated as the experimental variable affecting digital color extraction. The experiments were performed in two independent RT-PCR trials, and the resulting RT-PCR products from each run were analyzed in triplicate (n=6). The data are presented in the table below:

 

The intra-group coefficient of variation (CV_intra) remained consistently below 2% across all tested illumination levels, indicating excellent repeatability of the colorimetric assay. Furthermore, the inter-group coefficient of variation (CV_inter) was below 3%, demonstrating that the digital quantitative color analysis implemented using our open-source GitHub code is robust under different lighting conditions and supports its practical applicability for on-site detection.

The data further showed that the color thresholds established in Fig. 6 for distinguishing positive and negative samples remained valid across all tested illumination levels. However, under low-light conditions (260 lux), samples with low TMV infection levels, such as Tom#4 and Tom#5, may require more cautious interpretation.

Accordingly, a new subsection entitled "3.6. Stability of digital colorimetric analysis under varying lighting conditions" has been added to the revised manuscript.

References

Deng, H., Zhang, X., Kumar, A., Zou, G., Zhang, X., and Liang, X.-J. (2013). Long genomic DNA amplicons adsorption onto unmodified gold nanoparticles for colorimetric detection of Bacillus anthracis. Chemical Communications 49, 51-53.

Wang, Y., Wang, X., Yan, Y., Wang, J., Lu, Y., Abd El-Aty, A., and Wang, X. (2023). A visual colorimetric assay based on phage T156 and gold nanoparticles for the sensitive detection of Salmonella in lettuce. Analytica Chimica Acta 1272, 341501.

Xu, J., and Craig, S.L. (2007). Influence of the extent of hybridization on the hydrodynamic radius of DNA-functionalized gold nanoparticles. Langmuir 23, 2015-2020.

 

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript titled: One-step RT-PCR-Coupled Cysteamine-Functionalized Gold Nanoparticle Assay for Colorimetric Detection of Tobacco Mosaic Virus by Thuy-Duong Thi Tran et al. developed and optimized a one‑step RT‑PCR–coupled, cysteamine‑functionalized gold nanoparticle assay for the colorimetric detection of the virus. However, two major concerns need to be addressed in this study.

Lack of appropriate healthy controls: The assay currently does not include essential healthy controls, which are necessary to accurately evaluate background signal and assay reliability.

Specificity assessment: To demonstrate assay specificity, the authors should test non‑target but closely related species, including  ToMV, ToMMV, and ToBRFV. This is critical to confirm that the assay does not produce false‑positive reactions with other tobamoviruses.

Additional Comments

Sensitivity and repeatability: For the sensitivity evaluation, the authors should provide data demonstrating repeatability, ideally from two independent experiments conducted side‑by‑side by the same operator.

Line 46: Clarify the mention of 130K and 180K proteins (replicase and RdRP). Provide concise descriptions or relevant references for clarity.

Comparison to existing assays: The manuscript should include a discussion comparing this assay to existing diagnostic assays, particularly with regard to sensitivity.

Line 139: The description is unclear; please provide the correct parameters and ensure the conditions are explicitly stated.

Line 233: A minimum of three replicates and two independent experiments is required to validate the results.

Figure 4: There appears to be minimal visible color change across the five‑fold concentrations of cysteamine‑functionalized gold nanoparticles when comparing positive RT‑PCR amplicons and negative controls.

Line 351: Specify whether 10^-3 refers to dilution or copy number. If it refers to copy number, provide the calculation method used to derive these values.

Figure 6: Please clarify whether the figure represents copy number or dilution series.

Author Response

Reviewer 2:

1. Lack of appropriate healthy controls: The assay currently does not include essential healthy controls, which are necessary to accurately evaluate background signal and assay reliability.

Response to reviewer: We sincerely thank the reviewer for the constructive comments. True TMV-negative tobacco and tomato field samples were unavailable during this study, as all collected field samples, including commercially sourced asymptomatic seedlings, tested positive for TMV. Therefore, a field sample from passion fruit, a non-host of TMV infected with East Asian Passiflora virus AO strain (EAPV-AO), was used as an alternative plant control. As shown in Supplementary Fig. S4, RT-PCR products from this sample showed no detectable amplification by either agarose gel electrophoresis or the colorimetric assay, demonstrating that the TMV-specific primer pair does not cross-react with EAPV-AO or produce detectable non-specific amplification from host plant RNA.

In addition, the assay employs a TMV-specific primer pair in a one-step RT-PCR. In silico BLAST analysis against the NCBI nucleotide database further indicated high primer specificity toward TMV, making non-specific amplification from tobacco, tomato or passion fruit RNA highly unlikely under the optimized RT-PCR conditions.

Furthermore, the newly added data (Fig. S4C) showed no significant difference between RT-PCR reactions using TMV plasmids in the absence or presence of field plant RNA, providing additional evidence that endogenous plant matrix components do not interfere with the proposed colorimetric assay. Accordingly, Section 3.5 of the revised manuscript has been substantially revised.

2. Specificity assessment: To demonstrate assay specificity, the authors should test non‑target but closely related species, including  ToMV, ToMMV, and ToBRFV. This is critical to confirm that the assay does not produce false‑positive reactions with other tobamoviruses.

Response to reviewer: We would like to thank the reviewer for highlighting the important point regarding the assay specificity. This issue was thoroughly considered during our experimental design. However, obtaining TMV-free samples concurrently infected with other tobamoviruses proved challenging due to the exceptionally wide host range and highly contagious nature of TMV. To address this limitation, we performed in silico BLAST analysis against the NCBI nucleotide database using representative genome sequences of several prevalent tobamoviruses, including Tobacco mild green mosaic virus (TMGMV), Tomato mosaic virus (ToMV), Tomato Mottle Mosaic Virus (ToMMV), Cucumber green mottle mosaic virus (CGMMV), Tomato brown rugose fruit virus (ToBRFV), and Pepper mild mottle virus (PMMoV). The alignment results demonstrate that our TMV-specific primer pair is highly unlikely to yield non-specific amplicons under our optimized RT-PCR conditions. Although experimental validation using these viruses would further strengthen the study, the corresponding viral isolates were not available during the present work. We have added this issue in the discussion section of  the revised manuscript (lines 485-496).

Additional Comments

3. Sensitivity and repeatability: For the sensitivity evaluation, the authors should provide data demonstrating repeatability, ideally from two independent experiments conducted side‑by‑side by the same operator.

Response to reviewer: We sincerely thank the reviewer for the constructive comments. We would like to clarify that in the sensitivity evaluation (Fig. 6), the RT-PCR reactions were conducted in three independent trials. Subsequently, the RT-PCR products from each independent run were analyzed in duplicate using the colorimetric assay, resulting in a total of six replicates (n = 6) as demonstrated in figure legend.

 In addition, in this research we routinely performed parallel analyses using both agarose gel electrophoresis and the colorimetric assay on the exact same amplification products (lines 343-350 in revised manuscript). Specifically, immediately following the RT-PCR reaction, a 1µL aliquot of the RT-PCR product was loaded for agarose gel electrophoresis, while another 1µL aliquot of the identical sample was subjected to the colorimetric assay. This parallel approach eliminates variations in template concentration or amplification efficiency, rendering the comparative sensitivity data highly robust when comparing conventional agarose gel electrophoresis method and our colorimetric assay.

To further address the reviewer's concern, we have added the standard deviation and coefficient of variation (CV) values for the colorimetric measurements of the standard samples in the revised Fig. 6 (summarized in the table below). The CV values were all below 5%, demonstrating excellent repeatability of the proposed colorimetric assay. The corresponding text has also been revised (lines 365-368).

                   

We also supplement another gel image of Fig. S2 as follow:

                

4. Line 46: Clarify the mention of 130K and 180K proteins (replicase and RdRP). Provide concise descriptions or relevant references for clarity.

Response to reviewer: We agree and revised in the text (lines 45-50 in revised manuscript). The detail as follow:

“The TMV genome encodes four proteins: an approximately 130K protein that functions as an RNA helicase and methyltransferase; an approximately 180K protein, generated via translational read-through of the 130K protein termination codon, which contains an RNA-dependent RNA polymerase domain and forms a large subunit of the replicase complex; a 30K protein involved in cell-to-cell movement and a 17K coat protein [10, 11].”

5. Comparison to existing assays: The manuscript should include a discussion comparing this assay to existing diagnostic assays, particularly with regard to sensitivity.

Response to reviewer: We thank the reviewer for these constructive comments. We agree with the suggestion and have added more comparison explanation between our colorimetric assay and the standard agarose gel electrophoresis method to the Discussion section (lines 478–485 in the revised manuscript) as follows:

“The superior sensitivity of the colorimetric assay stems from the unique optical properties of gold nanoparticles (AuNPs). AuNPs possess exceptionally high extinction coefficients, up to a million times greater than those of organic dyes used in agarose gel electrophoresis, allowing them to intensely absorb and scatter light even at nanomolar concentrations [22, 23]. Furthermore, while agarose gel electrophoresis is a separation-based method that can cause diffusion and smearing of DNA molecules (resulting in signal loss), the colorimetric assay is performed in-tube. This in-tube approach allows the reaction to maintain equilibrium without diluting the signal.”

Regarding other published methods for TMV detection, such as multiplex RT-PCR (Kumar et al., 2011; Liu et al., 2019) or portable paper-based molecularly imprinted sensors (Gong et al., 2025), we evaluated these for potential sensitivity comparisons. However, these studies evaluated sensitivity using parameters like total extracted RNA dilutions or viral concentrations, which are not directly comparable to our method of quantifying TMV cDNA copies. Therefore, to ensure an accurate and fair comparison, we chose not to include these specific publications for sensitivity benchmarking in the manuscript.

6. Line 139: The description is unclear; please provide the correct parameters and ensure the conditions are explicitly stated.

Response to the reviewer: We sincerely thank the reviewer for this valuable comment. Since our experimental design focuses on systematically determining the optimized conditions for TMV detection, presenting the final parameters prematurely in the Materials and Methods section would not follow the logical flow of the study.

Therefore, in the revised manuscript, we have restructured the text to better reflect this chronological development. Section 2.6 (Colorimetric assay and digital color analysis) under Materials and Methods has been completely removed. Instead, we have introduced a new section, "5. Conclusion" to summarize all optimized conditions established throughout the entire TMV detection procedure. The detailed parameters have now been incorporated into this new section.

7. Line 233: A minimum of three replicates and two independent experiments is required to validate the results.

Response to Reviewer: As illustrated in Fig. 1 (line 233 in previous manuscript and line 230 in revised manuscript), the execution of "at least two independent experiments" signifies that the RT-PCR assays were conducted at least in two separate trials. Subsequently, the RT-PCR products from each independent run were analyzed in duplicate using the colorimetric assay, resulting in a total of four replicates (n = 4). This rigorous experimental design was consistently applied across all other experiments, ensuring the reliability, reproducibility, and statistical validity of the reported data.  

8. Figure 4: There appears to be minimal visible color change across the five‑fold concentrations of cysteamine‑functionalized gold nanoparticles when comparing positive RT‑PCR amplicons and negative controls.

Response to reviewer: We mentioned this in our text (lines 284-288 in previous manuscript and lines 281-285 in revised manuscript). Furthermore, we would like to clarify that the RT-PCR products (both positive and negative samples) were routinely confirmed in parallel by agarose gel electrophoresis before the colorimetric assay. At a low dilution (5-fold), both positive and negative samples exhibit a similar red coloration with almost equal values (~172 units), yielding a contrast delta near zero (Fig. 4). This may be due to the high concentration of RT-PCR matrix components at low dilutions, especially thiolated-primers, which can shield the AuNP surface and prevent salt-induced aggregation, thereby leading to false positive results.

9. Line 351: Specify whether 10^-3refers to dilution or copy number. If it refers to copy number, provide the calculation method used to derive these values.

Response to reviewer:  10³ is copy number of plasmid we used as template in one reaction of RT-PCR. We supplemented the calculation method of plasmid copy number in “2.5. Recombinant plasmid standard preparation of Materials and Methods part”. The detail as follow:

“Plasmid copy number per microliter was calculated according to following equation:

 

where C is the plasmid copy number per microliter, [DNA] represents the DNA concentration (g×µL⁻¹), P is the total plasmid length in base pairs (bp), 660 g×mol⁻¹ is the average molecular weight of a double-stranded DNA base pair, and 6.022 × 10²³ molecules×mol⁻¹ is the Avogadro’s constant. The total plasmid length was determined by the sum of the pCR 2.1-TOPO TA vector (3,931 bp) and the TMV insert (430 bp). Then, a 10-fold serial dilution (10⁸ to 10¹ copies/µL) was prepared in nuclease-free water and one microliter of each dilution served as the RT-PCR template for further experiments.”

10. Figure 6: Please clarify whether the figure represents copy number or dilution series.

Response to reviewer: Fig. 6 represents the plasmid copy number used as template in a single RT-PCR reaction. We would like to clarify that the template copy numbers match the dilution series exactly because we prepared a 10-fold serial dilution of the plasmid (from 10⁸ to 10¹ copies/µL) and then utilized exactly 1µL of each dilution as the template for the subsequent RT-PCR. Therefore, the dilution series directly translate to the absolute template copy numbers per RT-PCR reaction. We revised in “2.5. Recombinant plasmid standard preparation of Materials and Methods part” for clearer method.

 

 

References

Gong, H., Pang, T., Yang, X., Chen, F., Jiang, N., Li, Y., Chen, C., and Cai, C. (2025). Rapid visual detection of tobacco mosaic virus using a portable paper-based molecularly imprinted sensor. Sensors and Actuators B: Chemical 424, 136909.

Kumar, S., Udaya Shankar, A., Nayaka, S., Lund, O., and Prakash, H. (2011). Detection of Tobacco mosaic virus and Tomato mosaic virus in pepper and tomato by multiplex RT–PCR. Letters in Applied Microbiology 53, 359-363.

Liu, H., Wu, K., Wu, W., Mi, W., Hao, X., and Wu, Y. (2019). A multiplex reverse transcription PCR assay for simultaneous detection of six main RNA viruses in tomato plants. Journal of virological methods 265, 53-58.

 

Reviewer 3 Report

Comments and Suggestions for Authors

In the reviewed manuscript (MS), a new method for virus diagnosis was developed using tobacco mosaic virus as an example. The method is based on the conventional one-step RT-PCR, in which colloidal gold particles are used to visualize PCR products obtained. These particles aggregate when interacting with negatively charged dsDNA fragments generated by amplification and change color from bright red to purple. The differences between positive and negative samples are estimated by naked eye and can be enhanced by using colloidal gold particles coated with cysteamine, and thiolated primers.

I think it would be premature to accept this manuscript for publication for the next reasons.

First, Material and Methods, Results and Discussion sections differ little from one another. I think that the MS should be structured properly, and, most importantly, a clear protocol should be presented that describes the analysis process from beginning to end. Any improvements should be described separately, step-by-step and in other paragraphs.

Second, the analysis of PCR products using agarose gel electrophoresis has an undoubted advantage over the new method. It makes it possible to see any PCR products generated and immediately assess whether the expected target product was obtained.In contrast, the new method lacks resolution.That is, it remains unknown which PCR products exactly caused the color change. In addition, assessing the shade change from red to purple with the naked eye can be very subjective. Furthermore, agarose gel electrophoresis takes one hour maximum, so, the time gain is also questionable. Perhaps the results obtained could find some other application in the future.

Third, the authors should explain: i) what cysteamine is; ii) how were the thiolated primers obtained; iii) how was the quantitative assessment of the staining degree carried out? Also, virus names are not italicized (lines 11, 34,377, 378).

I think, the MS needs the major revision before acceptance.

 

Author Response

Reviewer 3:

1. First, Material and Methods, Results and Discussion sections differ little from one another. I think that the MS should be structured properly, and, most importantly, a clear protocol should be presented that describes the analysis process from beginning to end. Any improvements should be described separately, step-by-step and in other paragraphs.

Response to the reviewer: We sincerely thank the reviewer for the valuable comments and constructive suggestions. We recognized that since our experimental design focuses on systematically determining the optimized conditions for TMV detection, presenting the final parameters prematurely in the Materials and Methods section would not follow the logical flow of the study.

Therefore, we have carefully restructured the manuscript accordingly. The specific modifications are as follows:

• Section 2.2 (Synthesis and surface functionalization of gold nanoparticles (AuNPs)) under Materials and Methods: We have removed several sentences that overlapped with the Results section.
• Section 2.6 (Colorimetric assay and digital color analysis) under Materials and Methods: This section has been completely removed. Instead, we have added a new section, "5. Conclusion" to summarize all optimized conditions throughout the entire TMV detection procedure.
• Results section: This section has been substantially revised and updated with new experimental data.
• Discussion section: We have expanded this section to provide additional explanations regarding the thiolated primers, compare our colorimetric assay with conventional gel electrophoresis, and discuss the specificity and repeatability of the assay in greater detail.

 

2. Second, the analysis of PCR products using agarose gel electrophoresis has an undoubted advantage over the new method. It makes it possible to see any PCR products generated and immediately assess whether the expected target product was obtained. In contrast, the new method lacks resolution. That is, it remains unknown which PCR products exactly caused the color change. In addition, assessing the shade change from red to purple with the naked eye can be very subjective. Furthermore, agarose gel electrophoresis takes one hour maximum, so, the time gain is also questionable. Perhaps the results obtained could find some other application in the future.

Response to reviewer: We would like to thank the reviewer for the interesting points.

• Regarding the first issue concerning the resolution of agarose gel electrophoresis and our colorimetric assay: we agree that colorimetric assays generally lack the resolution to confirm the presence of PCR products. However, the colorimetric assay in this study was developed specifically to detect Tobacco Mosaic Virus (TMV) and was routinely validated in parallel with agarose gel electrophoresis throughout all experiments. By utilizing a TMV-specific primer pair and optimizing various conditions, we ensured the reliability and reproducibility of the method.

The results obtained from our colorimetric assay were highly consistent with those from agarose gel electrophoresis, confirming that our method can be reliably used as an alternative despite its lower resolution. Remarkably, our assay achieves higher sensitivity (as shown in Fig. 6), is more time-efficient, and eliminates the need for expensive optical equipment. In conclusion, this assay can be applied directly for TMV detection, bypassing conventional gel electrophoresis. Furthermore, for other viruses, the assay can be readily adapted by optimizing their respective parameters.

• Regarding the second issue concerning virus detection via visual inspection: We agree that evaluating color shifts from red to purple by the naked eye can be highly subjective. To address this limitation, we developed and provided an open-source code on GitHub (available at: https://github.com/Thaonguyennnee/ColorChooser) to objectively extract color values based on the CIELAB color space. While visual perception may vary among individuals, the color coordinates extracted via the CIELAB color space remain consistent and quantifiable. Furthermore, by utilizing the regression equation derived from the standard curve, extracted color values enable us to not only determine positive or negative results but also quantify the severity level of the viral infection.
• Regarding the third issue concerning the time efficiency of our method compared to agarose gel electrophoresis: We evaluated both methods using a throughput of 20 samples, a standard capacity facilitated by our laboratory's 20-well agarose gel casting systems. The proposed colorimetric assay requires a maximum of 35 minutes to achieve complete analysis. This duration encompasses: mixing PCR products with gold nanoparticles (AuNPs) (10 min), incubation (5 min), incubation post-NaCl supplementation (10 min), and colorimetric readout based on the CIELAB color space (10 min). While both methods exhibit comparable turnaround times for small sample sizes, agarose gel electrophoresis becomes logistically challenging when scaling up to high-throughput analysis due to labor-intensive gel preparation, sample loading, and equipment constraints. Furthermore, although the current iteration of our assay is not yet optimized for immediate on-site deployment, this AuNPs-based platform inherently possesses a significant potential for field-based viral detection—an application that remains unfeasible for conventional electrophoresis. Our ongoing studies are focused on optimizing this method to deliver a rapid, cost-effective, and point-of-care diagnostic

3. Third, the authors should explain: i) what cysteamine is; ii) how were the thiolated primers obtained; iii) how was the quantitative assessment of the staining degree carried out? Also, virus names are not italicized (lines 11, 34,377, 378).

Comment 1: The authors should explain what cysteamine is.

Response to reviewer: We thank the reviewer for this valuable suggestion. Cysteamine (2-aminoethanethiol) is a small  molecule containing both a thiol group (-SH) and an amine group (-NH₂). The thiol group forms a strong covalent gold-sulfur (Au-S) bond with the AuNP surface, while the positively charged amine group remains exposed to the aqueous environment, facilitating the binding efficiency of the negatively charged dsDNA amplicons. Following the reviewer's recommendation, we have added an explanation of cysteamine's structure and role to Section 3.1 (lines 148-152 in the revised manuscript) as follows:

“ To achieve this, cysteamine (2-aminoethanethiol), a small molecule containing a thiol group (-SH) and an amine group (-NH₂), was employed for surface functionalization. The thiol group binds covalently to the AuNP surface through strong gold-sulfur (Au-S) interactions, while the exposed, positively charged amine groups facilitate the binding efficiency of the negatively charged dsDNA amplicons.”

Comment 2: The authors should explain how the thiolated primers were obtained.

Response to reviewer: We appreciate the reviewer's comment. The TMV-specific thiolated primers were chemically synthesized by a commercial vendor (IDT) with a 5' thiol modification (5' Thiol Modifier C6 S-S). These modified primers were used in the one-step RT-PCR so that the resulting amplicons contain thiol groups, allowing them to interact effectively with the Au@Cys system to induce the observed color change. We have updated Section 2.4 of the manuscript to explicitly clarify the source and the modification details of these thiolated primers (lines 110-112 in the revised manuscript) as follows:

“To enable subsequent interaction with the gold nanoparticles, these primers were chemically synthesized with a 5' thiol modification (5' Thiol Modifier C6 S-S) by Integrated DNA Technologies (IDT).”

Comment 3: The authors should explain how the quantitative assessment of the staining degree was carried out.

Response to reviewer: We sincerely thank the reviewer for the constructive suggestions. In this research, the quantitative assessment was performed using digital image analysis based on the CIELAB color space, focusing on the green-red axis. The procedure is carried out as follows:

Image Capture: Images of the reaction tubes are recorded 10 minutes post-NaCl addition under controlled lighting approximately 600lux, with the camera positioned 40 cm away at a 90-degree angle to the surface.

Image Processing: Images are uploaded to the open-source ColorChooser software (https://github.com/Thaonguyennnee/ColorChooser). Identical Regions of Interest (ROIs) are selected from the center of each liquid solution.

Viral Load Calculation: Color values of samples were extracted as the mean A value of selected liquid region based on CIELAB color space (green-red axis). The color values (y) then were used to calculate the log10 of viral load (x) in samples using the linear equation of y=3.5078x + 137.42.

We haved added a new section "5. Conclusion"  including these details.

Comment 4: Virus names are not italicized (lines 11, 34,377, 378).

Response to reviewer: We sincerely apologize for this oversight. We have corrected all the virus names (lines 11, 34, 379, and 381 in the revised manuscript) to normal font in the revised manuscript.

 

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript can be accepted in the present form.

Reviewer 3 Report

Comments and Suggestions for Authors

Thank you very much for revising of the manuscript.