The mRNACalc web server accounts for the hypochromicity of modified nucleosides and enables the accurate quantification of nucleoside-modified mRNA

  1. Esteban Finol
  2. Sarah E. Krul
  3. Sean J. Hoehn
  4. Carlos E. Crespo-Hernández

Reviewed by

Read the full article

Listed in

Log in to save this article

Abstract

Nucleoside-modified mRNA technologies necessarily incorporate N 1 -methylpseudouridine into the mRNA molecules to prevent over-stimulation of cytoplasmic RNA sensors. Despite this modification, mRNA concentrations remain mostly determined through measurement of UV absorbance at 260 nm wavelength (A 260 ). Herein, we report that the N 1 -methylpseudouridine absorbs approximately 40% less UV light at 260 nm than uridine, and its incorporation into mRNAs leads to the under-estimation of nucleoside-modified mRNA concentrations, with 5-15% error, in a mRNA sequence dependent manner. We therefore examined the RNA quantification methods and developed the mRNACalc web server. It accounts for the molar absorption coefficient of modified nucleotides at 260 nm wavelength, the RNA composition of the mRNA, and the A 260 of the mRNA sample to enable accurate quantification of nucleoside-modified mRNAs. The webserver is freely available at https://www.mrnacalc.com .

Graphical Abstract

Article activity feed

  1. Version published to 10.1101/2023.07.27.550903v3 on bioRxiv
  6. Review Commons

    Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

    Learn more at Review Commons

    Reply to the reviewers

    1. General Statements

    This manuscript aimed at:

    1. a) producing the evidence that supports the need for performing RNA hydrolysis and applying the appropriate nucleoside MACs for the determination of nucleoside-modified mRNA concentrations using UV spectroscopy.
    2. b) Providing the m1Y MAC value and a new resource to the mRNA field community to perform the above-mentioned procedure. This piece is therefore a “resource” manuscript, rather than a biotechnological innovation or basic research manuscript.

    2. Point-by-point description of the revisions

    Considering that the reviewers coincided in some of their comments, we compiled them in topic and provided our response to the reviewers.

    Topics:

    • On the novelty in this manuscript
    • On the impact of nucleoside modifications on the DFBHI-Broccoli complex.
    • On the role of modified nucleosides on mRNA folding and the independent verification on a distinct mRNA.
    • On the cap used for IVT in the webserver.
    • On the average of Epsilon (e or MAC) values.
    • Other minor comments. On the novelty in this manuscript:

    Comments:

    Reviewer #1

    “Major Comments:

    • In the introduction, the authors should discuss the novelty more by describing which techniques are currently available for quantification of modified RNA and how this study is novel.” __Reviewer #2 __

    “(Significance (Required)):

    The study develops an accurate method to measure RNA concentrations which can improve dosing accuracy. The methods developed here will be beneficial for a broad range of fields employing mRNA-therapies.”

    __Reviewer #3 __

    “(Significance (Required)):

    Since 50 years, scientists that works in the field of modified nucleic acids have determined the concentration of the nucleic acids in the same way, which means by determining the epsilon values of the modified nucleosides, using the epsilon values of the natural nucleosides at same wavelength, and then calculating the concentration after measuring absorption at (for example) 260 nm (wavelength could change dependent on modified nucleoside that is incorporated). This manuscript is not really innovative.”

    Response:

    We thank the reviewers for bringing up this topic. We want to reassure to the reviewers, editors and readers that, throughout the manuscript, we have carefully selected the wording to avoid claiming any novelty on the principle of RNA hydrolysis or the use of nucleotide Molar absorption coefficients (MAC) and UV spectroscopy for the determination of RNA concentrations. We have “revised”, “assessed” and “examined” these experimental procedures, we “determined” the M1Y and we “developed” the mRNAcalc webserver.

    This “resource” manuscript therefore mainly aims at introducing the mRNAcalc webserver to the community and providing the underlying biochemical principles of the methods suggested in the webserver. These articles are often published in webserver issues or as “resource” articles in certain journals, including some of the journals in Review Commons.

    The authors understand that the data in our manuscript are not often provided for this type of “resource” manuscripts, and it might have led to a misunderstanding. For instance, the OligoCalc webserver was published in Nucleic Acid Research, it has become a valuable tool for the oligonucleotide research community (1621 citations in 15 years), and no experimental evidence supporting its underlying calculations is provided in the manuscript.

    For our manuscript, we have cited the corresponding source of the principle of the experimental methods, and we additionally performed some experiments to reproduce the findings using nucleoside-modified mRNAs with the intention of highlighting the importance of performing RNA hydrolysis (Fig.2b) and implementing the MAC of modified nucleotides (Fig 1e and 1f) for the determination of modified-nucleoside mRNA concentration using the Beer-Lambert law. We have felt compelled to do so, despite the fact that they represent well-established science and methods, as correctly pointed out by one of the reviewers.

    We have taken into account that a few dozen of non-RNA biochemistry focused laboratories around the world are currently embracing for the first time the nucleoside-modified mRNA technologies and, to our knowledge, not a single article in the nucleoside-modified mRNA field has mentioned the need of implementing a different MAC for the determination of nucleoside-modified mRNA concentration using UV spectroscopy in either its main text or Materials & Methods section. We want to reassure the reviewers that the authors, before starting the experimental investigation, performed an extensive literature search and failed to find the m1Y MAC at 260 nm. Our search included a few hundreds of research articles, several doctoral thesis (including Sister Miriam Michael Stimson’s work), classic books such as Hall, Ross “The modified nucleosides in nucleic acids” and nucleotide manufacturers’ datasheets. However, the authors cannot rule out that other investigators in the mRNA field have previously determined the m1Y MAC at 260 nm in aqueous buffered solution and this knowledge has remained hidden under the frequently used statement of “The mRNA concentrations were determined spectroscopically” or any alike statement.

    Following the suggestion of Reviewer #1, we have also included a brief comment in the introduction on the fluorescence-based techniques for the determination of nucleic acid concentration (lines 87-91), as follows:

    “Other non-UV-spectroscopic methods relying on the unspecific RNA binding of certain fluorophores (such as RiboGreen, Thermo Fisher Scientific) for the determination of RNA concentration may help to overcome any change in the MAC of modified nucleoside mRNA. However, the impact of RNA modifications on the binding affinity of these fluorophores also remains unknown.”

    On the impact of nucleoside modifications on the DFBHI-Broccoli complex:

    Comments:

    Reviewer #1

    “3) The broccoli aptamer has U in it which when mutated to pseudouridine (Ψ) or N1-methylpseudouridine may change the structure minutely affecting the cis-trans transition in aptamer- DFHBI-1 complex and hence in fluorophore properties. A control which shows the effect (or lack thereof) of aptamer modifications on fluorophore properties should be carried out. The ratio of A260/F507 can get affected by the denominator although it may/may not be insignificant.”

    Reviewer #2

    “Specific comments:

    …

    * In Fig. 1D, the authors normalize the absorbance on mRNA to fluorescence of DFHBI-1T when bound to dBroccoli aptamer. The aptamer will contain uridines and therefore modified uridines. Will modified uridines affect binding affinity of the substrate to the aptamer? Could the differences in fluorescence be because of stronger/weaker binding of the substrate with modified uridines?”*

    Response:

    We thank reviewers for enquiring about the effect of U-to-Y and U-to-M1Y substitutions on the DFHBI-1T-dBroccoli interaction, RNA folding or fluorophore properties. We have indeed investigated thoroughly and observed that there was no significant difference in the binding affinity, melting point, or relative brightness across the three DFHBI-1T-Broccoli complexes. These results go in line with the previously published photophysical and biochemical properties of the Broccoli−DFHBI-1T (reference 15 in manuscript). These data are provided as supplementary Table 1 in the revised manuscript.

    Supplementary Table 1: photophysical and biochemical properties of mutated Broccoli−DFHBI-1T complexes.

    Complex

    Max em (nm)

    Relative brightness*

    KD (nM)+

    Tm (°C)+

    U-Broc−DFHBI-1T

    (ref. 15)

    507

    360

    48

    U-Broc−DFHBI-1T

    507

    1.000 ± 0.002

    379.6 ± 13.89

    49.13 ± 0.13

    Y-Broc-DFHBI-1T

    507

    1.005 ± 0.004

    378.7 ± 8.11

    49.46 ± 0.09

    m1Y-Broc-DFHBI-1T

    507

    1.004 ± 0.003

    375.6 ± 8.17

    49.23 ± 0.07

    *Relative to the U-Broc-DFHBI-1T complex. Data are shown as mean ± SD.

    • Data are shown as KD ± Error of the fit or Tm ± Error of the fit.

    On the role of modified nucleosides on mRNA folding and the independent verification on a distinct mRNA:

    Comments:

    Reviewer #1

    “2) Fig 1d- In this experiment, the RNA is not hydrolyzed prior to concentration measurement. The authors should discuss how nucleoside modifications in the RNA may affect structure of the RNA, how significant that effect is on the ____e ____(MAC) and how justified it is to attribute the reduction in ____e ____(MAC) entirely to the mutations.

    …

    4) The reduction in A260 in modified nucleosides should be accurately measured and independent of the RNA. Hence, the values determined here should be shown to be independent of at least another RNA sequence.”

    Response:

    We want to express our gratitude to Reviewer #1 for enquiring about the potential impact of the modified nucleosides on the mRNA folding. We have further discussed this aspect on our interpretation of the data in Fig 1d. No doubt, this reviewer’s comment has substantially enriched the discussion in our manuscript.

    For the revised version of the manuscript, we have also performed the same measurements using a different mRNA. We have used an mRNA with a higher m1Y composition. We have observed a stronger reduction in mRNA UV absorption (A260) in the m1Y-modified mRNA, confirming that the MAC of the nucleobase composition is the main determinant of mRNA UV absorption. We have appended these data to the manuscript as supplementary Figure 2 and the associated text can be found in lines 141-155 of the manuscript and in the following lines:

    “By normalizing the UV absorbance (A260) of each mRNA by its corresponding fluorescence (F507), it was observed that in practice the relative UV absorbance of the nucleoside-modified mRNA was significantly reduced as compared to the standard mRNA (DA260 = -10.6%, Fig. 1d and 1e). The hypochromicity was more pronounced in a second m1Y-mRNA with higher m1Y* composition (DA260 = -11.8%, Supplementary Figure 1). In principle, the modified nucleosides can also promote mRNA folding and reduce its UV absorption. This is particularly relevant for the pseudouridine modification. Its N1-hydrogen can engage in additional hydrogen bonds, promoting and stabilizing RNA folding. For instance, the U-to-Y substitution in tRNA stabilizes the folded structure that is essential for translation (reviewed in ref. 16). However, the m1Y nucleobase lacks this additional hydrogen bonding capability, and it is expected to have little or no effect on the RNA folding of low CG-content (1Y-mRNAs followed the anticipated hypochromicity associated to the nucleobase hypochromicity at 260 nm wavelength and not their expected contribution to RNA folding, these data suggest that the observed reduction in nucleoside-modified mRNA UV absorption is mainly determined by the nucleobase composition and the intrinsic MAC of the nucleosides in these mRNA.”*

    On the cap used for IVT in the webserver:

    Comment:

    __Reviewer #2 __

    “(Evidence, reproducibility and clarity (Required)):

    Specific comments:

    1. * The authors should expand discussion on the effect of different caps used for IVT as the choice of the cap appears to be an important selection on the server.”* Response:

    We thank Reviewer #2 for spotting that we had forgotten to comment on this feature of mRNAcalc web server in the manuscript. We have included a short paragraph on the 5’ mRNA cap nucleotides and their implementation in the mRNACalc webserver (line 160-164), as follow:

    *“In the mRNACalc webserver, the MACs of distinct modified nucleosides that form the capping nucleotide in the 5’ mRNA cap were also implemented for the sake of completeness. The capping nucleotide only represents one nucleotide out of thousands of nucleotides in a mRNA molecule and its contribution to the mRNA molar absorption is rather negligible.” *

    On the averaged Epsilon (____e ____or____* MAC) values:*

    Comment:

    __Reviewer #3 __

    (Evidence, reproducibility and clarity (Required)):

    “I find the way the authors use the epsilon value of pseudoU (and analogues), as a mean value of literature data to be incorrect. The epsilon value is absolute and can not vary from one measurement to another. In fact it is a good parameter to define concentration. When different values are obtained, it means that compound is not pure , or measured at different pH or solvent, or the compound is not weighted exactly. When publishing a methodology to determine concentration of nucleic acids, it might be good to determine the exact epsilon value you want to use, yourself.”

    Response:

    We thank Reviewer #3 for bringing up this topic. We had experimentally determined the MAC values when we observed that it was completely absent in the literature and it was essential to provide a valuable tool to the nucleoside-modified mRNA community, as it is the case of the m1Y nucleoside. For the Y and m5C nucleosides, we have now determined their MAC for this revised version of the manuscript and recalculated the average, this time including our own determination and the previously determined values (in aqueous buffered solution) from multiple sources in the literature or manufacturer’s product datasheets and we implemented it in the webserver. We agree that the different values may relate to distinct amount and nature of the impurities in the manufacturer’s preparation. We believe that the average of these values may reflect a better approximation to their absolute epsilon value. Detailed information on these calculations and methods are now provided in the supplementary notes (Lines 109 to 142).

    Other minor comments:

    Reviewer #1

    “5) Fig 2c- The figure should be remade using larger symbols as it is difficult to see how different the concentration measurements are depending on the method of hydrolysis.”

    Response:

    Figure 2c is intended to show a schematic representation of the experimental workflow and use of the mRNAcalc webserver. We assume that Reviewer #1 referred to Fig 2b. We have enlarged the symbols to ease the visibility of the data.

    “6) Lines 27, 45 and 33 have an incorrect symbol for ____e____(MAC) and for the word 'coefficient”

    Response:

    We thank Reviewer #1 for the interest of improving our manuscript in detail. The typos have been corrected in the revised manuscript.

    We thank Reviewers for all the values comments to our manuscript; they have enriched and substantially improved its quality and readability.

  7. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #3

    Evidence, reproducibility and clarity

    I find the way the authors use the epsilon value of pseudoU (and analogues), as a mean value of literature data to be incorrect. The epsilon value is absolute and can not vary from one measurement to another. In fact it is a good parameter to define concentration. When different values are obtained, it means that compound is not pure , or measured at different pH or solvent, or the compound is not weighted exactly. When publishing a methodology to determine concentration of nucleic acids, it might be good to determine the exact epsilon value you want to use, yourself.

    Significance

    Since 50 years, scientists that works in the field of modified nucleic acids have determined the concentration of the nucleic acids in the same way, which means by determining the epsilon values of the modified nucleosides, using the epsilon values of the natural nucleosides at same wavelength, and then calculating the concentration after measuring absorption at (for example) 260 nm (wavelength could change dependent on modified nucleoside that is incorporated). This manuscript is not really innovative.

  8. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #2

    Evidence, reproducibility and clarity

    General Comments:

    Finol and colleagues argue that uridine modifications absorb less UV light than uridine, leading to underestimation of RNA concentration for modified RNAs. Based on this observation, they created a web server for accurate calculation of RNA concentrations. This is an interesting manuscript and would benefit the field. Specific comments are listed below:

    Specific comments:

    1. The authors should expand discussion on the effect of different caps used for IVT as the choice of the cap appears to be an important selection on the server.
    2. In Fig. 1D, the authors normalize the absorbance on mRNA to fluorescence of DFHBI-1T when bound to dBroccoli aptamer. The aptamer will contain uridines and therefore modified uridines. Will modified uridines affect binding affinity of the substrate to the aptamer? Could the differences in fluorescence be because of stronger/weaker binding of the substrate with modified uridines?

    Significance

    The study develops an accurate method to measure RNA concentrations which can improve dosing accuracy. The methods developed here will be beneficial for a broad range of fields employing mRNA-therapies. Reviewer expertise: Lipid nanoparticles, mRNA, self-amplifying RNA, immunoengineering

  9. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #1

    Evidence, reproducibility and clarity

    Summary: The study aims at developing a method of accurate quantification of concentration of self- amplifying RNA and modified RNA by determining molar absorption coefficient of modified nucleosides with higher accuracy.

    Major Comments:

    1. In the introduction, the authors should discuss the novelty more by describing which techniques are currently available for quantification of modified RNA and how this study is novel.
    2. Fig 1d- In this experiment, the RNA is not hydrolyzed prior to concentration measurement. The authors should discuss how nucleoside modifications in the RNA may affect structure of the RNA, how significant that effect is on the (MAC) and how justified it is to attribute the reduction in (MAC) entirely to the mutations.
    3. The broccoli aptamer has U in it which when mutated to pseudouridine (Ψ) or N1-methylpseudouridine may change the structure minutely affecting the cis-trans transition in aptamer- DFHBI-1 complex and hence in fluorophore properties. A control which shows the effect (or lack thereof) of aptamer modifications on fluorophore properties should be carried out. The ratio of A260/F507 can get affected by the denominator although it may/may not be insignificant.
    4. The reduction in A260 in modified nucleosides should be accurately measured and independent of the RNA. Hence, the values determined here should be shown to be independent of at least another RNA sequence.

    Minor comments:

    1. Fig 2c- The figure should be remade using larger symbols as it is difficult to see how different the concentration measurements are depending on the method of hydrolysis.
    2. Lines 27, 45 and 33 have an incorrect symbol for (MAC) and for the word 'coefficient'

    Significance

    Accurate measurement of RNA concentrations can be key where precise quantification of RNA is required and any error gets amplified such as RNA-based therapeutics dependent on RNA amplification or RNA modification. This study is also important for research on the effect of RNA modifications on RNA structure in vitro or in vivo.

  10. Version published to 10.1101/2023.07.27.550903v2 on bioRxiv
  11. Version published to 10.1101/2023.07.27.550903v1 on bioRxiv