Application of Combined Long Amplicon Sequencing (CoLAS) for Genetic Analysis of Neurofibromatosis Type 1: A Pilot Study

Round 1

Reviewer 1 Report

Togi and Ura et al., have used CoLAS for NF1 genetic analysis in 20 patients and found that Among 18 cases, CoLAS detected not only 10 mutations including three unreported mutations and one 19 mosaic mutation, but also various splicing abnormalities and allelic expression ratio quantitatively. Authors have shown that this NGS application can identify disease-causative mutation, but also accurately show the detailed individual genomic structure, the mRNA transcription and splicing state of NF1 gene.

1. At many places, authors write…Suspicious cases- define this term at the first mention of the term?
2. In the figure 3 b legend, describe what the colors indicate, although its very difficult to pinpoint the differences in the alignment data. Can this data be shown somehow with better clarity?

Togi and Ura et al., have used CoLAS for NF1 genetic analysis in 20 patients and found that Among 18 cases, CoLAS detected not only 10 mutations including three unreported mutations and one 19 mosaic mutation, but also various splicing abnormalities and allelic expression ratio quantitatively. Authors have shown that this NGS application can identify disease-causative mutation, but also accurately show the detailed individual genomic structure, the mRNA transcription and splicing state of NF1 gene.

1. At many places, authors write…Suspicious cases- define this term at the first mention of the term?
2. In the figure 3 b legend, describe what the colors indicate, although its very difficult to pinpoint the differences in the alignment data. Can this data be shown somehow with better clarity?

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

In this pilot study, the authors explain the application of Combined Long Amplicon Sequencing (CoLAS) 2 for Genetic Analysis of Neurofibromatosis Type 1. I advise the authors to take into consideration the following major and minor remarks to improve the quality of the manuscript:

Major comments:

1. In the materials and Methods section: Considering that CoLAS is a relatively new genetic analysis for the study of NF1, I suggest explaining in more detail the NGS data analysis pipeline. This information could be attached to supplementary material if it is the case. In addition to this text, the authors could include a complementary scheme and references that support the way to make a better data analysis process.
2. The criteria for establishing false-positive and true heterozygous variants should be clearly described in the materials and methods section.
3. Lines 136-137: Explain how it was experimentally adjusted the concentration of each primer to obtain uniform amplification products in the multiplex long PCR.
4. More methodological aspects should be detailed in terms of:
   1. MuLAS and rLAS analyses
   2. In-Silico analysis
   3. Heterozygosity mapping of NF1 region and detection of a chromosomal level large deletion
5. Coverage criteria should be included in the materials and methods section.
6. In the point mutation detection of NF1 section (lines 230 to 238), it is convenient to describe the mutations of the suspicious cases (NF_06, 09, 13, and 16) separately.
7. The results section should be redrafted due to there is information that should be in the materials and methods section (e.g. lines: 198 to 202, 238 to 239, 244 to 248, 268 to 278) and another in the discussion section (e.g. lines 298 to 310, lines 316 to 321, lines 329 to 331, lines 375 to 378 mix results and discussion).
8. Due to the small sample size and mix of confirmed and suspected cases of NF1, it is worth mentioning the limitations of the study
9. Lines 57 to 58: As a pilot study, the conclusion should be more achievable to the type of study carried out.

Minor comments:

1. Renumber the tables since there are two Table 4:

Table 4. Summary of in-silico analysis for missense variants with unknown significance.

Table 4. Heterozygosity mapping of NF1 region.

2. Improve the designs and footer descriptions of Tables 4: Summary of in-silico analysis for missense variants with unknown significance and Heterozygosity mapping of NF1 region.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

The manuscript improved a lot in its structure, therefore I accept its publication

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

see file

Comments for author File: Comments.docx

Author Response

Reviewer 1

Dear Reviewer,

Thank you for giving us the opportunity to submit a revised draft of our manuscript to the International Journal of Molecular Sciences. We appreciate the time and effort that you have dedicated to providing your valuable feedback on our manuscript. We are grateful to you for your insightful comments on our paper. We have been able to incorporate changes to reflect all of your suggestions. We highlighted the changes within the manuscript.

Here is a point-by point response to your comments and concerns.

Point 1: The major point is that this study describes a genetic technical approach but does not compare it in terms of sensitivity, specificity, cost/efficiency with the gold standard which is NGS (see Bianchessi et al 2020 or Pasmant et al. 2015).

Response 1: First of all, this is a pilot study to optimize and confirm the performance of CoLAS for genetic analysis of NF1. There is no intention to statistically discuss the superiority or inferiority of this method compared to past large scaled studies with NGS analysis. As reviewer 1 pointed out, the number of patients is small and statistical analysis is meaningless. The study also includes patients who do not meet clinical diagnostic criteria. To clarify this point, the title has been changed as follows. “Application of Combined Long Amplicon Sequencing (CoLAS) for genetic analysis of Neurofibromatosis type 1: a Pilot Study”

Point 2: In the introduction authors claim “In recent years, with the advent of next-generation sequencing (NGS) the mass sequencing has become easier, but the currently widely used capture sequencing method has not been able to diagnose the entire mutation spectrum of NF1.” This statement should be modulated, and references provided. Current panel NGS approaches (which provide qualitative SNV and quantitative CNV analyses) can identify an NF1 pathogenic variant in >95% of patients with an NF1 phenotype. The authors should explain how their approach is relevant in this context and what it adds to the field.

Response 2: Certainly, the NGS diagnosis of NF1 has reached a sufficient level only from the viewpoint of mutation detection rate. However, there is still room for improvement in the NGS method currently in use. A significant proportion of NF1 missense mutations (30%) were deleterious by affecting pre-mRNA splicing. Therefore, the true significance of NF1 missense mutations cannot be determined without simultaneous testing of DNA and RNA. In addition, the MLPA and the NGS diagnosis of large deletion/duplication mutation due to alternation of the copy number of exons. That do not determine the mutation itself, i.e. breakpoint sequence on DNA. There is a need for a method that can detect mutations in a wide range and determine the elaborate nature of mutations in more detail in a single experimental system using NGS. This is the purpose of testing CoLAS in this study. To clarify this point, we have added this content in abstract lines 12 to14 and in introduction lines 66-86 with new citations.

Point 3: In the discussion, the authors state: line 300: “In this pilot study, CoLAS was applied to the genetic diagnosis of NF1 and also demonstrated its high utility" but this is not demonstrated in their study which analyses too few patients. Presenting a study of a new diagnostic approach with only few patients studied is a major limitation of this publication.

Response 3: I agree with the reviewers in this regard, but as mentioned earlier, this study is a pilot study. “High utility” does not mean that the mutation detection rate is high, but that this method provides a new perspective that was not available in previous methods. As shown in Figure 4, CoLAS can accurately show the effect of mutations at the DNA level on mRNA splicing. For example, you can see in one shot that what looks like a missense mutation, such as NF_03, is actually a splicing mutation. In addition, it has become clear that protein-truncating mutations do not necessarily produce nonsense-mediated decay (NMD), and CoLAS has become a tool for predicting the degree of NMD by comparing DNA and RNA and correcting base-specific PCR bias. To clarify this meaning, we added lines 320 to 323.

Point 4: A cost study should be provided.

Response 4: CoLAS is an inexpensive method. Since it is based on Long-range PCR, it is cheaper to set up than creating a capture probe newly. The running cost is also low, and since the library is created directly from the long PCR product with Nextera (illumine), it takes less time and effort than the capture probe method. We evaluated the CoLAS running costs at approximately 10,000 Japanese Yen per sample, including DNA and RNA extraction, cDNA synthesis, library preparation, NGS sequencing and Sanger sequencing. This is described in lines 132-134.

Point 5: Moreover, the question of large deletions (>50 nt) of one/a few exons is not addressed because it is not shown that the method is effective in detecting them because it relies on the need for heterozygous variants in the regions tested in patients.

Response 5: That is why we used long-range PCR and showed efficiency of heterozygous mapping by that (Table 4). Since one PCR amplifying region is long, 10~20Kb, and contains intron sequences, it is highly likely that it will contain heterozygous polymorphisms. In fact, heterozygous SNP were detected at most sites in the majority of cases. Since the Long PCR product is fragmented and read by NGS, if one or more heterozygous SNPs are detected, it is indicated that the entire PCR-amplified region is heterozygous (2 copies). Also, each Long PCR region overlaps, the heterozygosity of the entire NF1 gene region can be monitored. Conversely, deletions of the entire NF1 gene by chromosomal micro-deletions can be easily identified by the absence of heterozygous SNPs in all long PCR products.

In the PCR region where heterozygotes have been proven, large deletions/duplication mutations within it are reliably detected. This is the fact we have shown in our TSC study (reference 16) and the break points were accurately analysed by the Pindel software (reference 18). Although this approach is not perfect because some unproven regions of heterozygosity remain, but it has the advantage of being able to directly determine the DNA breakpoint sequence of the deletion. The capture-probe method and MLPA can detect only excess or deficiency in the number of gene copies, but cannot detect the DNA sequence of the mutation itself.

Point 6: There are frequent deletions of the whole NF1 locus (of different types, see below) but also deletions of one or more exons: Some of these types of deletions (that a "standard" NGS approach can identify; see Imbard et al. 2015) should have been tested to demonstrate that the presented approach was able to call them. It seems difficult to deduce sensitivity and specificity values for this new approach from experiments performed on a small number of control samples. There are 3 types of recurrent deletions of the NF1 locus, as well as atypical deletions. Only a type 1 has been tested. How does this methodology analyze the different deletion types of the NF1 locus?

Response 6: Due to the small sample size, there were no cases of intragenic deletion, and detection of this type of mutation could not be confirmed. However, heterozygous SNPs were detected in most long PCR amplicons, confirming that both alleles were amplified. Therefore, it is presumed that if an intragenic deletion is present, it can be reliably detected.

The presence or absence of a deletion of the entire NF1 locus can be easily inferred from the heterozygous mapping of long PCR products as shown in Table 4. The probability of the absence of heterozygous SNPs in all 290 kb DNA covered by long PCR amplicon is extremely low, suggesting a large chromosomal level deletion or isodisomy. Although this study did not contain micro-deletions other than Type 1, it is easy to design long PCR primers with breakpoints for Type 2 and 3 as well, if there are cases.

Point 7: Major technical parameters to evaluate the results are missing such as average sequencing depth, 100x coverage percentages of all regions and coding/non-coding regions. Is the minimum promoter covered by the DNA approach?

Response 7: Average sequencing depth is 200X (at least 100X) through out all coding/non-coding regions in DNA level, and more than 600X (at least 400X) in RNA level. We added this to lines 129-132.

The most 5’ side DNA long-range PCR primer set NF1-A1 cover chr17: 31,089,750_31,096,525 and NF1 exon 1 located at 31,094,927_31,095,369. Long PCR amplicon start at 5177 bp upstream of NF1 first exon and of course including core promoter region.

Point 8: The need to have a double DNA and RNA approach is a limitation of this methodology, in particular because the RNA sampling is not always simple, nor the conservation of the samples. This could be discussed (see Evans et al. EBioMedicine. 2016).

Response 8: Compared to the method of Evans et al., Our method is easy because it is only separate PBMC from blood for RNA extraction without culturing. Evans et al. have been culturing for a short period with inhibitors to eliminate the effects of NMD. Our method does not culture blood and also measures the effect of NMD by comparing it with DNA data. Double measurement of DNA and RNA is rather an advantage of CoLAS, and the experimental procedure is easy and hassle-free.

Minor remarks:

Point 9: Introduction line 63: “highly homologous DNA sequences…”; homologous sequences are evolutionarily derived from the same ancestral sequence. Please replace with “highly similar”.

Rephrase the sentence: lines 82-83 “However, these patients are young and therefore it is possible to not been appeared due to the late onset or infrequent onset nature of symptoms other than café au lait macules.”

Line 153. “2.4. Heterozygosity mapping of NF1 legion and detection of chromosomal level large deletion”

Line 251. “However, the difference founded…”

Response 9: We rephrased the pointed-out sentences.

Line 66                similar

Lines 102-103     However, these patients may not be accompanied by cutaneous and/or plexiform neurofibroma because they are young.

Line 180             2.4. Heterozygosity mapping of NF1 region and detection of chromosomal level large deletion

Lines 276-277    However, this difference of mRNA expression

Point 10: The following references of the field could be cited.

Response 10: We cited some of these papers and added them to the reference (reference number 6, 15 and 19).

Sincerely,

Yo Niida, M.D., Ph.D.

Center for Clinical Genomics

Kanazawa Medical University Hospital

• Daigaku, Uchinada, Kahoku, Ishikawa, 920-0293, JAPAN

Phone No: +81 076-286-2211

Email Address: niida@kanazawa-med.ac.jp

Author Response File: Author Response.pdf

Reviewer 2 Report

The authors aimed to assess the role of the "combined long amplicon sequencing" (CoLAS), a method for simultaneously analysing the whole genomic DNA region and also the full-length cDNA of the disease-causative genes, for NF1 evaluation. Due to the extreme variation among NF1 patients e the still unclear genotype-fenotype relationship, the topic is absolutely intriguing. Howere, I have some issues:

-the first one is related to the simple size: does it was pre-planned and evaluated?15 new patiens for a total of 20 could be a little sample, for a so variable syndrome.

The second one is a consequence of the first one, and related to the general approach while performing such analyses: any overall evaluation according to the STREGA statement, to evaluate the strenght of the experimental design? [see Little J et al., STrengthening the REporting of Genetic Association Studies (STREGA): an extension of the STROBE statement. PLoS Med. 2009 Feb 3;6(2):e22. doi: 10.1371/journal.pmed.1000022.]

Author Response

Dear Reviewer,

Thank you for giving us the opportunity to submit a revised draft of our manuscript titled “Dual deep sequencing improves the accuracy of low-frequency somatic mutation detection in cancer gene panel testing” to the International Journal of Molecular Sciences. We appreciate the time and effort that you have dedicated to providing your valuable feedback on our manuscript. We are grateful to you for your insightful comments on our paper. We have been able to incorporate changes to reflect all of your suggestions. We highlighted the changes within the manuscript.

Here is a point-by point response to your comments and concerns.

Point 1: The first one is related to the simple size: does it was pre-planned and evaluated?15 new patiens for a total of 20 could be a little sample, for a so variable syndrome.

Response 1: First of all, this is a pilot study to confirm the performance and features of CoLAS for molecular analysis of NF1. There is no intention to statistically discuss the superiority or inferiority of this method compared to past large scaled studies with NGS analysis. As reviewer 2 pointed out, the number of patients is small and statistical analysis is meaningless. The study also includes patients who do not meet clinical diagnostic criteria. To clarify this point, the title has been changed as follows. “Application of Combined Long Amplicon Sequencing (CoLAS) for genetic analysis of Neurofibromatosis type 1: a Pilot Study”

              Previously reported NGS diagnosis of NF1 has reached a sufficient level, more than 90%, from the viewpoint of mutation detection rate. However, there is still room for improvement in the NGS method currently in use. A significant proportion of NF1 missense mutations (30%) were deleterious by affecting pre-mRNA splicing. Therefore, the true significance of NF1 missense mutations cannot be determined without simultaneous testing of DNA and RNA. In addition, the MLPA and the NGS diagnosis of large deletion/duplication mutation due to alternation of the copy number of exons. That do not determine the mutation itself, i.e. breakpoint sequence on DNA. There is a need for a method that can detect mutations in a wide range and determine the elaborate nature of mutations in more detail in a single experimental system using NGS. This is the purpose of testing CoLAS in this study. To clarify this point, we have added this content in abstract lines 12 to14 and in introduction lines 66-86 with new citations.

Point 2: The second one is a consequence of the first one, and related to the general approach while performing such analyses: any overall evaluation according to the STREGA statement, to evaluate the strength of the experimental design? [see Little J et al., STrengthening the REporting of Genetic Association Studies (STREGA): an extension of the STROBE statement. PLoS Med. 2009 Feb 3;6(2):e22. doi: 10.1371/journal.pmed.1000022.]

Response 2: As mentioned above, this study is a pilot study to clarify the advantages and disadvantages of using CoLAS for the diagnosis of NF1. Through this research, we were convinced of its convenience, economy and usefulness. On the other hand, it was shown that there are some patients whose copy count of the entire region cannot be grasped only by mapping heterozygous SNPs, and it was found that addition of MLPA etc. may be required as secondary screening. We are currently planning a genetic analysis cohort study of Japanese NF1 patients using CoLAS. For this next research, we would like to refer to the evaluation method of the paper you pointed out.

We rewrote the parts pointed out by other reviewers. All rewrites are highlighted in yellow.

Sincerely,

Yo Niida, M.D., Ph.D.

Center for Clinical Genomics

Kanazawa Medical University Hospital

• Daigaku, Uchinada, Kahoku, Ishikawa, 920-0293, JAPAN

Phone No: +81 076-286-2211

Email Address: niida@kanazawa-med.ac.jp

Author Response File: Author Response.pdf

Reviewer 3 Report

Very interesting & well designed study.

Some changes to phrasing required to make the text more readable, e.g. lines 52-56 & elsewhere.

Author Response

Dear Reviewer,

Thank you for giving us the opportunity to submit a revised draft of our manuscript to the International Journal of Molecular Sciences. We appreciate the time and effort that you have dedicated to providing your valuable feedback on our manuscript. We are grateful to you for your insightful comments on our paper. We have been able to incorporate changes to reflect all of your suggestions. We highlighted the changes within the manuscript.

Here is a response to your comments and concerns.

Point: Some changes to phrasing required to make the text more readable, e.g. lines 52-56 & elsewhere.

Response: We have rewritten the part you pointed out in line 52-57. In addition, we rewrote the parts pointed out by other reviewers. All rewrites are highlighted in yellow.

Sincerely,

Yo Niida, M.D., Ph.D.

Center for Clinical Genomics

Kanazawa Medical University Hospital

• Daigaku, Uchinada, Kahoku, Ishikawa, 920-0293, JAPAN

Phone No: +81 076-286-2211

Email Address: niida@kanazawa-med.ac.jp

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

I thank the authors who made a rapid and argued response to the various points I had raised in my review.

Nevertheless, they did not respond to the substantive criticisms: this article is a technological development which lacks essential elements (including a comparison with reference methods, and it has an insufficient numbers of tested samples with low diversity of the types of NF1 genetic variants) to evaluate the interest of the approach proposed by the authors.

For example, detection of CNVs based on heterozygous SNVs alone and not on comparison of sequencing depths does not seem relevant to me and the authors do not test the samples needed to show this.

Reviewer 2 Report

Despite the authors more clearly stated that this could represent a pilot study, I have always some issues about the sample size and the study strength. I think that the manuscript may greatly be improved with a bigger sample size