Novel Quinazoline Derivatives as Highly Effective A2A Adenosine Receptor Antagonists
, Romain Duroux, Séverine Ravez, Sergio Hernandez-Tapia, Martin Fossart, Mathilde Coevoet, Maxime Liberelle, Saïd Yous, Nicolas Lebègue and Patricia Melnyk *Round 1
Reviewer 1 Report
Comments and Suggestions for Authors
The manuscript by Laversin et al. expanded the structure-activity relationships based on knowledge gained from their previous publication describing quinazoline derivatives as potent adenosine A2A receptor antagonists. Therefore, the authors explored further substitution in positions 6 and 7 of the 2-amino-quinazoline scaffold. The most potent antagonist (Ki = 20 nM) from the previous publication was 6-bromo-4-(furan-2-yl)quinazoline-2-amine. Many of the newly synthesized derivatives retained adenosine A2A receptor affinity with Ki values smaller than 100 nM. Importantly, the affinity was now determined using a fluorescence polarization assay and not a classical radioligand binding assay. Now, additionally, potency measurements with cAMP assays were performed. The authors also explored derivatization of the 2-amino group, and expectedly, many of the derivatives were inactive or less potent. However, one derivative with 52 nM affinity was obtained carrying a 4-(piperidin-1-ylmethyl)benzene rest. The amino group is generally important for adenosine A2A receptor affinity. Thus, this definitely represents an interesting achievement. Such a 2-amino-substituted compound would also be interesting for structrual studies. In view of this referee, the manuscript raises a few questions that have to be addressed before publication:
1. The authors state that „the Kd for A2AR membrane preparation was obtained by a kinetic on/off experiment“ (line 1038). Please report the obtained KD value and preferably include the data in the SI. Please also report the displacement reference curve for ZM241385. This data is absolutely necessary to comprehend the Ki determination.
2. Related to question 1: The SI contains a few results of the fluorescence polarization assay, however, is not complete. All data (competition binding curves) that enabled the determination of Ki or IC50 values should be reported. Please also report the competition binding curves for ZM241385.
3. Additionally, the authors should provide the exact number of replicates for each data point.
4. Can the authors comment on the selectivity towards other adenosine receptor subtypes?
5. Have the authors considered to combine substitutions in 6 and 7 position?
Additional comments:
· Line 273: substituted
· Line 140: substituents
· The Supplementary Information would benefit from a table of contents.
· Line 184f: „It is known that increasing the size and volume of GPCR ligands can promote a switch to full antagonist activity:“ Please provide a reference.
Author Response
Thank you very much for taking the time to review this manuscript. Please find the detailed responses below and the corresponding revisions/corrections highlighted in the re-submitted files
1. The authors state that „the Kd for A2AR membrane preparation was obtained by a kinetic on/off experiment“ (line 1038). Please report the obtained KD value and preferably include the data in the SI. Please also report the displacement reference curve for ZM241385. This data is absolutely necessary to comprehend the Ki determination.
KD value (2 nM) was added to the experimental section of the fluorescence polarization (FP) binding assay. Displacement curve of ZM241385 was reported in the SI as required.
2. Related to question 1: The SI contains a few results of the fluorescence polarization assay, however, is not complete. All data (competition binding curves) that enabled the determination of Ki or IC50 values should be reported. Please also report the competition binding curves for ZM241385.
All missing curves were added to the SI.
3. Additionally, the authors should provide the exact number of replicates for each data point.
All compounds were tested in three independent experiments, in duplicate. This information was added to the experimental section ot the FP binding assay.
4. Can the authors comment on the selectivity towards other adenosine receptor subtypes?
From the conservation of sequence among the A2 family, A2A presents 55,6%, 43,4% and 37.4% of identity versus A2B, A1 and A3 respectively, but 72.5%, 58% and 61.1% of homology. But only considering the key residues involved in the binding of our compounds, they are totally conserved among the whole family but the glutamate at the top of the pocket for A3. That's why it is not expected to present a strong selectivity from this point of view. Nevertheless, A2B and A3 present a shift of structure, visible at the Trp279 level. The docking of compound 5m was performed on all the subtype, thanks to recent resolution of all the structure by CryoEM, and it led to a virtual binding affinity (in kcal/mol) of -8.66 for A2a vs -7.26, -7.98 and -7.55 for A1, A2b and A3, respectively. Furthermore, the hydrogen-bonding was disrupted or even lost in this model. So, we can expect at least a form of selectivity from the docking perspective.
5. Have the authors considered to combine substitutions in 6 and 7 position?
Indeed, we considered to combine substitutions in 6 and 7 positions. We plan to carry out this type of modulation in the future. We can mention this idea as a prospect.
Additional comments:
Line 273: substituted. Corrected
Line 140: substituents. Corrected
The Supplementary Information would benefit from a table of contents. Table of contents added
Line 184f: „It is known that increasing the size and volume of GPCR ligands can promote a switch to full antagonist activity:“ Please provide a reference. The following reference was added: Taylor, J. B.; Triggle, D. J. Comprehensive Medicinal Chemistry II; Elsevier: Amsterdam, 2007..
Reviewer 2 Report
Comments and Suggestions for AuthorsThe manuscript “Novel quinazoline derivatives as highly effective A2A adenosine receptor antagonists” presents a synthesis and evaluation of a series of compounds based on quinazoline scaffold. The aim is to develop a SAR study for A2A adenosine receptor antagonist. The diversification concerns the type and position of substituents (C6 and C7, as well as C4) and the character of amine in position 2.
The SAR study is descriptive, whereas for activity, receptor affinity and antagonism were studied. Did the Authors consider any quantitative comparison of SAR? There are papers on SAR in A2A system, with different scaffolds, are there similarities in substituent selection and justification?
The majority of the text and accompanying supplementary data present the synthesis of the novel compounds. Unfortunately, the details of chemical procedures require some attention, especially as most of compounds seem to be characterized for the first time.
There are issues in the manuscript that need addressing:
1. description of structural elements: line 19 in Abstract: aminopentylpiperidine and 4-[(piperidin-1-yl)methyl]aniline should not be described as chains, pentyl spacer is chain-based, cyclic structures add dimension. Line 242: apolar tertiary amines – substituents at nitrogen (or the structure apart from nitrogen) may be considered apolar, however, the phrase is misleading.
2. Line 78: “the introduction of aminoalkyl chains containing tertiary amines at the C2- position” may suggest the location of tertiary amine next to C2
3. The co-crystallization of 1 bound to receptor (line 70) is not accompanied with the reference and database information, such information is also missing in Figure 2 caption. The description in 4.3 is practically at the end of the text. The docking parameters should be specified or based on reference.
4. In Scheme 1, compounds 4a-p are formed from 3a-h. the experimental part lists the range 4a-q and even 4r (synthesized according to literature). The appearance of additional substituents in position 6 (4i, 4j) suggests that the reaction was not completely selective, whereas the scheme suggests that the R1 substituent was present in series 3. Please verify the source of compounds 4i and 4j and the list of substituents, as readers may expect compounds with aromatic R1 and different Ar.
5. For lines 125-128, are the data from previous or current work? Please clarify the source.
6. Line 209: micromolar range (9r, Ki = 927 nM). The Ki value is nearly micromolar (0.927 μM), still, the use of two units in this phrase is confusing.
7. There are abbreviations that are not explained (PE, DIPEA, BINAP, RCF, IBMX, etc.)
8. In Figure 1, Compound 1 is one of the compounds reported before and analysed in the manuscript. A modification of the name (current work/reference) may clarify the confusion, as the compound is addressed as “1” in the text.
9. 4.1 Chemistry: two melting point determinations are mentioned, in line 290 and 306. Please verify this or combine in one sentence. HPLC column characteristics (dimensions, particle size) is missing. Please explain: ” The compounds were dissolved in 100 mL of buffer B and 900 mL of buffer A”. Flow rate 0.2 ml/min suggests narrow bore column or LC-MS system. Please clarify the HPLC conditions, especially solvents for LCMS analyses.
10. Melting points for compounds 2a-h are above 300 deg, which may suggest that hydrochlorides were isolated. Please comment.
11. Mass spectra for 2a-h are reported in negative mode, is there any special reason for this change? If LCMS procedure was used, was the pH of mobile phase modified?
12. In description “General procedure for synthesis of compounds 4a-q” and other places, reagents are added to a sealed tube. Please clarify the construction of the vessel, and the moment of sealing. In line 458, reaction mixture in such tube was refluxed, please clarify.
13. Synthesis of compounds 6a-c is presented in different style, is there a reason?
14. Line 1033: the A2AR membranes mentioned here were obtained according to procedure 4.2.2? Please clarify or provide reference. The protein concentration (line 1028) was adjusted to 9 mg/mL. Is this in any way related to final concentration of A2AR membranes (0.8 μg/μL) in FP binding assay?
15. In supplementary data, “LC-MS chromatograms” (actually, MS spectra) are presented. There is no information on MS system used and LCMS conditions. The mass analyser should be mentioned, as the reported precision of m/z values differs in compound descriptions and does not seem sufficient to provide confirmations. No calculated m/z values are provided, and in case of Br- and Cl-containing compounds more than one value is listed, not only monoisotopic, but including abundant isotopic peaks. The range of m/z values ( X-axis scale) differs in the supplementary data. For example, “LC-MS chromatogram of compound 5b” is shown in 100-1000 range, with no clear isotopic profile, and the current picture does not support the presence of bromine atom (isotopic peaks should be of similar abundance, like in 5l (range 270-370))?
16. Why is compound 9h marked as HCl salt in supplementary data as the only one in the series?
Minor issues:
in Abstract, indices seem lost
Reagents and conditions in figure comments are in different font sizes.
line 169: please correct: These result was consistent
line 304: a gradient form 100% of buffer A
line 964: In a sealed tube was added 2-chloroquinazoline
line 1052: concentration of 5.105 cells
Comments on the Quality of English Language
There are only few problems, mentioned in the main review.
Author Response
Thank you very much for taking the time to review this manuscript. Please find the detailed responses below and the corresponding revisions/corrections highlighted in the re-submitted files
1. Description of structural elements: line 19 in Abstract: aminopentylpiperidine and 4-[(piperidin-1-yl)methyl]aniline should not be described as chains, pentyl spacer is chain-based, cyclic structures add dimension. Line 242: apolar tertiary amines – substituents at nitrogen (or the structure apart from nitrogen) may be considered apolar, however, the phrase is misleading.
The term “chains” was replaced by “substituents” in line 19
The term “apolar tertiary amines” was replaced by “hindered tertiary amines” in line 249
2. Line 78: “the introduction of aminoalkyl chains containing tertiary amines at the C2- position” may suggest the location of tertiary amine next to C2
The sentence “the introduction of aminoalkyl chains containing tertiary amines at the C2-position” was replaced by “the introduction of substituents containing hindered tertiary amines on the nitrogen at the C2-position” in line 85
3. The co-crystallization of 1 bound to receptor (line 70) is not accompanied with the reference and database information, such information is also missing in Figure 2 caption. The description in 4.3 is practically at the end of the text. The docking parameters should be specified or based on reference.
PDB code: 8DU3 was added to line 78 and in Figure 2, 3 and 5. The PDB code 3EML of ZM-241385 co-crystallized with A2A receptor was also added. The reference corresponds to reference 29 for both co-crystallization and docking data and was added to paragraph 4.3. in line 1078.
4. In Scheme 1, compounds 4a-p are formed from 3a-h. the experimental part lists the range 4a-q and even 4r (synthesized according to literature). The appearance of additional substituents in position 6 (4i, 4j) suggests that the reaction was not completely selective, whereas the scheme suggests that the R1 substituent was present in series 3. Please verify the source of compounds 4i and 4j and the list of substituents, as readers may expect compounds with aromatic R1 and different Ar.
The following sentence was added in line 99: “Even though chlorine in position 4 is more reactive than bromine in position 6, the suzuki reaction on compound 3a is not completely selective and yields di-arylated compounds 4i and 4j”
5. For lines 125-128, are the data from previous or current work? Please clarify the source.
Data correspond from current work.
6. Line 209: micromolar range (9r, Ki = 927 nM). The Ki value is nearly micromolar (0.927 μM), still, the use of two units in this phrase is confusing.
Units have been homogenized in line 215
7. There are abbreviations that are not explained (PE, DIPEA, BINAP, RCF, IBMX, etc.)
Missing abbreviations were added.
8. In Figure 1, Compound 1 is one of the compounds reported before and analysed in the manuscript. A modification of the name (current work/reference) may clarify the confusion, as the compound is addressed as “1” in the text.
For a better understanding, 1 was replaced by compound 1 in the text.
9. 4.1 Chemistry: two melting point determinations are mentioned, in line 290 and 306. Please verify this or combine in one sentence. HPLC column characteristics (dimensions, particle size) is missing. Please explain: ” The compounds were dissolved in 100 mL of buffer B and 900 mL of buffer A”. Flow rate 0.2 ml/min suggests narrow bore column or LC-MS system. Please clarify the HPLC conditions, especially solvents for LCMS analyses.
Indeed, there was a mistake for the determination of melting point. The first method is the method used for all the compounds. The second is an old laboratory method that we forgot to remove. HPLCs have not been integrated into this manuscript, so we have removed the characteristics related to the device and the method. The LC-MS protocol has been rewritten and made more detailed.
10. Melting points for compounds 2a-h are above 300 deg, which may suggest that hydrochlorides were isolated. Please comment.
They did not correspond to hydrochloride salts. Quinazolinediones are known for their high melting point.
11. Mass spectra for 2a-h are reported in negative mode, is there any special reason for this change? If LCMS procedure was used, was the pH of mobile phase modified?
Mass spectra are always recorded in both positive and negative modes. Some compounds respond only in the negative mode, as in the case of quinazolinediones 2a-h.
12. In description “General procedure for synthesis of compounds 4a-q” and other places, reagents are added to a sealed tube. Please clarify the construction of the vessel, and the moment of sealing. In line 458, reaction mixture in such tube was refluxed, please clarify.
Protocols were clarified in line 393, 464 and 895
13. Synthesis of compounds 6a-c is presented in different style, is there a reason?
Synthesis of compounds 6a-c is now presented in the same style as the other compounds
14. Line 1033: the A2AR membranes mentioned here were obtained according to procedure 4.2.2? Please clarify or provide reference. The protein concentration (line 1028) was adjusted to 9 mg/mL. Is this in any way related to final concentration of A2AR membranes (0.8 μg/μL) in FP binding assay?
For cell culture, we were inspired by the protocol described in the article of Kecskes et al. It was specified in the experimental section. Membrane preparations are at a concentration of 9mg/mL after the cell culture. For FP tests, we diluted membrane preparations in the buffer solution to reach a concentration at 0.8 mg/mL. 9 mg/mL is far too concentrated for FP testing.
15. In supplementary data, “LC-MS chromatograms” (actually, MS spectra) are presented. There is no information on MS system used and LCMS conditions. The mass analyser should be mentioned, as the reported precision of m/z values differs in compound descriptions and does not seem sufficient to provide confirmations. No calculated m/z values are provided, and in case of Br- and Cl-containing compounds more than one value is listed, not only monoisotopic, but including abundant isotopic peaks. The range of m/z values ( X-axis scale) differs in the supplementary data. For example, “LC-MS chromatogram of compound 5b” is shown in 100-1000 range, with no clear isotopic profile, and the current picture does not support the presence of bromine atom (isotopic peaks should be of similar abundance, like in 5l (range 270-370))?
LCMS conditions were added to experimental section. The protocol described at first, was wrong.
We don’t see the isotopic profile for bromine-containing compounds like 5b because the zooms are not adjusted on the spectra, but we described them in the experimental section.
16. Why is compound 9h marked as HCl salt in supplementary data as the only one in the series?
Purification of this compound was too complicated to be achieved by flash chromatography, so the hydrochloride salt was formed and purified by recrystallization.
All the minor issues were also corrected
Reviewer 3 Report
Comments and Suggestions for AuthorsThe publication presented here focuses on the search for new A2a adenosine receptor ligands designed based on a previously obtained ligand co-crystallized with a target protein. The authors systematically investigate the effect of different substituents at various, specified positions of the core of the molecule, basing the SAR analysis on biological results and molecular docking results. Here, out of obligation as a reviewer, I would like to ask why molecular dynamics simulations were not performed? In my opinion, understanding the dynamic changes of the complex over time would have provided a more complete picture of the SAR under study.
A few minor editing issues:
Line 32 (and the whole manuscript) - References in square brackets are somehow misleading, especially that they’re put after the dot.
Lines 70-73- It’s unclear whether this structure has been deposited in the PDB. This information appears in the materials & methods section at the end of manuscript, yet I think, it should be stated at the very beginning.
Lines 96-115 please check the font/size. I’d suggest giving the reaction condition in the Figure caption, not as a separate paragraph
Line 101 – please use capital leters for Scheme, Figure etc..
Author Response
Thank you very much for taking the time to review this manuscript. Please find the detailed responses below and the corresponding revisions/corrections highlighted in the re-submitted files
1. Here, out of obligation as a reviewer, I would like to ask why molecular dynamics simulations were not performed? In my opinion, understanding the dynamic changes of the complex over time would have provided a more complete picture of the SAR under study.
Molecular dynamics (MD) were not envisaged here as it is not reported that flexibility of the receptor at the orthosteric site may significatively influence the binding mode. Even more, in Hui Wen Ng et al (2013) or Dr. Giuseppe Deganutti et al (2017), we can see that the region involved in the inner orthosteric site is remarkable stable during of MD and that ligand energy interaction does not vary. As docking study for compound 1 had predicted the actual resolved structure, we solely relied on molecular docking for evaluating the SAR.
2. Line 32 (and the whole manuscript) - References in square brackets are somehow misleading, especially that they’re put after the dot.
Reference rating has changed
3. Lines 70-73- It’s unclear whether this structure has been deposited in the PDB. This information appears in the materials & methods section at the end of manuscript, yet I think, it should be stated at the very beginning.
PDB numbers were added to the text and in the figures
4. Lines 96-115 please check the font/size. I’d suggest giving the reaction condition in the Figure caption, not as a separate paragraph
Corrected
5. Line 101 – please use capital leters for Scheme, Figure etc..
Corrected
Round 2
Reviewer 2 Report
Comments and Suggestions for AuthorsThe Authors of the manuscript “Novel quinazoline derivatives as highly effective A2A adenosine receptor antagonists” provided the modified version in impressively short time, addressing most of the comments.
Unfortunately, there are still points that should be discussed or modified:
1. The SAR studies described in the manuscript show limited relation to known A2A ligands, mainly ZM-241385. Some additional comments may be beneficial (similar substituents or original structures), as the modifications appear without deeper background.
2. Line 99: Suzuki reaction (named rection) – capital letter is usually expected
3. Line 312: the purity of evaluated compounds was judged to be > 95% as determined by UPLC – LC-UV or LC-MS? In case of UV, full scan or specific wavelength?
4. The problem of sealed tube seems to be solved, although in line 464 reagents are added to sealed tube, please confirm.
5. The information on LC-MS system (mass analyzer/detector ACQUITY QDa (Performance) detector (scan 50–1250)), single quadrulople, explains limited resolution of mass spectra. However, some uniform presentation of data should be followed: in the description of compounds, up to group 5, it is usually full unit precision, from 5a, the m/z values show better precision. If the same system was used, the same presentation is expected, especially as the mass spectra in supplementary data offer at least the decimal resolution. Considerable work is required, but the data may be used as reference.
In case of compounds containing Br or Cl, the Authors list more than one m/z value, corresponding to major isotopic peaks, although not in all cases (for example, 6a).
6. The provided answer: “We don’t see the isotopic profile for bromine-containing compounds like 5b because the zooms are not adjusted on the spectra, but we described them in the experimental section.” needs comment – was the profile not visible in zoomed spectrum (then, there is no confirmation of Br), or is zoomed spectrum not available (then, where are the data for experimental section coming from?)? In both cases, explanation is required.
7. In supplementary data, all figures with titles like “LC-MS chromatograms of compound ….” should be renamed, as they present mass spectra (=mass spectrum of compound ….). The spectra were probably obtained for the centers of chromatographic peaks in LC-MS chromatograms at specific retention times, but the difference between chromatogram and spectrum is fundamental.
8. The range of m/z values in supplementary figures differs (no comment from the Authors was offered). If the digital LC-MS data are not available (as no zoom was presented in several cases) and unified presentation of mass spectra is not possible, verification of values in experimental section is the minimum.
Author Response
