The Synthesis and Anti-Cytomegalovirus Activity of Piperidine-4-Carboxamides

Round 1

Reviewer 1 Report

This manuscript by Gou et al, expands on previous manuscripts from this group that define the antiviral properties of piperidine-4-carboxamides and synthesized analogs. This is an important area of research as there is currently no vaccine against CMV and current antivirals do have toxicity or viral mutation concerns. The authors present a logical, straightforward manuscript. Enthusiasm is dampened by the lack of detail in the writing of this article. Much is this is easily addressable. The use of hepatocytes from a humanized murine model is a nice addition.      

Major Critiques

• The title states piperidine-4-carboxamides, but this naming is not used in the abstract or introduction (use NCGC3955 naming from previous manuscripts). This may lead to confusion for new readers. Additionally, in the discussion, the authors mention future analogs that would replace piperidine in hopes of better antiviral activity. Suggest amending the title or defining in the introduction.
• Please include when the compounds were added after CMV infection. Was it a single treatment? Do you know the half-life of the compounds? Mechanism (or offer insight into how the new compounds work in your discussion)?
• Infection levels are sporadically included in the text and figure legends. Please include for all experiments and normalize presentation (moi vs PFU vs MOI= 1 PFU/cell, etc.)
• Figure 2 is missing part A (structure). B,C,D are mislabeled. Figure legend is correct, but text mentions A as being chemical structure.
• Most experiments in this manuscript are an N=2, with 3 biological replicates. It would strength the rigor of the experimental approach to at least complete an N=3 for key experiments.
• This research is not being completed in a vacuum. There are numerous colleagues who have published similar findings that are not mentioned. Inclusion of other CMV antiviral approaches in your discussion strengthens your approach and results.
  • Repurposing: PMID29712656, PMID 32690644, PMID 26763147
  • Targeting host factors: PMID 34390771, PMID32759737, PMID32590041

Minor Critiques

• In abstract, “…herpesvirus 1 or 2.” should be herpes simplex virus 1 or 2 (line 32)
• “…resulted in loss anti-CMV activity…” page 2 line 88 should be corrected.
• Incomplete sentence Page 2 line 88-91
• The description of the virus being used in each figure is confusing. Naming includes: CMV Towne (page 5, line 9 & fig 1A), pp-28 luciferase CMV (page 5, line 6) or pp-28 luciferase Towne CMV (page 7, line 77-78).. This continues through the manuscript. It should be standardized to avoid reader confusion.
• Page 5 line 25 “and” is a typo
• Table 1. I really appreciated circling the modification of compound 1 and 2. I believe I would benefit the manuscript to continue this for all compounds.
• Table 1. A recognizable demarcation between cc50 values of compounds 1-4 and the name column of 8-11 would allow readers to understand the table more readily.
• Fig 1. GCV-5 reads as the name is GCV-5, not GCV at a concentration of 5uM. Perhaps layer the X-axis titles. First layer is concentration, second layer is treatment.
• All figures. Define NI, INF and GCV etc.
• Testing of MCMV and HSV1/2 could use a rationale sentence. A single sentence explaining validation of specificity to CMV family.
• Letermovir is defined as “LTV” (page 9 line 128) but LET is used in Fig 7.
• Figure 7 figure legend mislabels B and C as C and D line 141
• Section 2.6 – Are the compounds (specifically LTV) preventing egress? It seems to be what the data is suggesting, but not discussed.
• What statistical test was used in Fig 7A, 7B, 7C. Include in figure legend and methods section
• TB40 strain. Is it TB40E or TB40F? If it is TB40E, does it retain UL128-131? (is it a “clinical” or wildtype strain)?
• In the discussion, page 12, line 185-186, it is mentioned that the hepatocytes were infected with Towne and TB40. This should be clarified as Fig 4 only includes infection with Towne (not mentioned in PHH methods either).
• What murine CMV strain was used and what was the source?

Author Response

We thank the reviewers for their careful and thoughtful review. We have addressed all the comments below, which allowed us to improve the manuscript and figures.

Reviewer #1

This manuscript by Guo et al, expands on previous manuscripts from this group that define the antiviral properties of piperidine-4-carboxamides and synthesized analogs. This is an important area of research as there is currently no vaccine against CMV and current antivirals do have toxicity or viral mutation concerns. The authors present a logical, straightforward manuscript. Enthusiasm is dampened by the lack of detail in the writing of this article. Much is this is easily addressable. The use of hepatocytes from a humanized murine model is a nice addition.      

 Major Critiques

• The title states piperidine-4-carboxamides, but this naming is not used in the abstract or introduction (use NCGC3955 naming from previous manuscripts). This may lead to confusion for new readers. Additionally, in the discussion, the authors mention future analogs that would replace piperidine in hopes of better antiviral activity. Suggest amending the title or defining in the introduction. Response: We agree. A statement was added in the introduction (page 2, line 79): “Structurally related compounds to NCGC2955, which all contain the piperidine-4-carboxamide motif (see purple rectangle for Compound 1, Table 1)…”
• Please include when the compounds were added after CMV infection. Was it a single treatment? Do you know the half-life of the compounds? Mechanism (or offer insight into how the new compounds work in your discussion)? Response: Compounds were added in a single treatment following infection. Information was added (lines 269-271): “Following 90-minute adsorption, media containing virus was removed and replaced by DMEM with 4% FBS (Gibco) in the presence or absence of compounds. Infected or infected-treated HFFs were collected at specific time points depending on the assay used”. We have not yet tested the half-life of the compounds.

The following was added to the discussion (lines 222-226): “The terminase complex, targeted by LTV,  cleaves DNA to package the viral genome into the capsid (Ligat et al., 2018). Compared to NCGC2955 analogs LTV showed a stronger reduction in viral progeny and viral loads in supernatants. These data may suggest the production of CMV DNA containing non-infectious virus that lacks some viral proteins”. 

• Infection levels are sporadically included in the text and figure legends. Please include for all experiments and normalize presentation (moi vs PFU vs MOI= 1 PFU/cell, etc.). Response: All figure legends now specify the virus titer used.
• Figure 2 is missing part A (structure). B,C,D are mislabeled. Figure legend is correct, but text mentions A as being chemical structure. Response: The figure and text have been corrected (page 6, lines 58-61).
• Most experiments in this manuscript are an N=2, with 3 biological replicates. It would strength the rigor of the experimental approach to at least complete an N=3 for key experiments. Response: Most experiments were executed at least twice and confirmed by several methods. Each analog was tested by the pp28-luciferase assay, Western blot, and plaque assay and there was a high correlation between all assays. In addition, for the combination studies, each analog was tested independently again before the drug combination experiment was performed.
• This research is not being completed in a vacuum. There are numerous colleagues who have published similar findings that are not mentioned. Inclusion of other CMV antiviral approaches in your discussion strengthens your approach and results.
• Repurposing: PMID29712656, PMID 32690644, PMID 26763147
• Targeting host factors: PMID 34390771, PMID32759737, PMID32590041

Response: We certainly agree. Until now we have not selected for resistant mutants, thus it  is possible that host factors are involved in NCGC2955 activity. The discussion now includes the following section (lines 194-201: “Despite broad-spectrum antiviral activity, our tested compounds did not inhibit HSV1 or HSV2, suggesting a specific mechanism of action that may involve host factors required for efficient replication of several groups of viruses. Interest in host factors that restrict virus replication, such as the electron transport system, cytochrome P450 51, mitochondrial regulatory proteins, and autophagy led to the identification of agents that could be further developed or repurposed for CMV therapeutics as monotherapy or combination therapy with direct-acting FDA approved antiviral agents (Clark et al., 2018; Combs et al., 2021; Hahn et al., 2020; Mercorelli et al., 2020; Mukhopadhyay et al., 2018):.

 Minor Critiques

• In abstract, “…herpesvirus 1 or 2.” Should be herpes simplex virus 1 or 2 (line 32). Response: The change was made.
• “…resulted in loss anti-CMV activity…” page 2 line 88 should be corrected. Response: The sentence has been corrected to “in loss of anti-CMV activity”.
• Incomplete sentence Page 2 line 88-91. Response: The sentence has been corrected to “Since the synthesis of the thieno[3,2-b]pyrrole heterocycle found in 1 required the synthesis of potentially explosive azido precursors on gram scale, we opted to test the more readily available truncated pyrrole 4 or indole 3 replacements”.
• The description of the virus being used in each figure is confusing. Naming includes: CMV Towne (page 5, line 9 & fig 1A), pp-28 luciferase CMV (page 5, line 6) or pp-28 luciferase Towne CMV (page 7, line 77-78). This continues through the manuscript. It should be standardized to avoid reader confusion. Response: We carefully revised the Materials and Methods and text. The pp28-luciferase was constructed in the background of Towne CMV and noted as “pp28-luciferase CMV”. Western blots used the original Towne CMV strain.
• Page 5 line 25 “and” is a typo. Response: This has been corrected. The sentence now reads “The CC₅₀ of Compounds 7 and 9 was > 500 µM, yielding a selectivity index of >1500”.
• Table 1. I really appreciated circling the modification of compound 1 and 2. I believe I would benefit the manuscript to continue this for all compounds. Response: The table has been revised, and specific notes were added to the text. For example, line 88: ”Compound 2 (isopropyl carboxamide in red)”, page 5 line 4: “the pyrrole 4-(2-aminoethyl)pyridine 7 (Sindac et al., 2013) was resynthesized (4-(2-aminoethyl)pyridine colored blue)”.  
• Table 1. A recognizable demarcation between cc50 values of compounds 1-4 and the name column of 8-11 would allow readers to understand the table more readily. Response: This has now been provided.
• Fig 1. GCV-5 reads as the name is GCV-5, not GCV at a concentration of 5uM. Perhaps layer the X-axis titles. First layer is concentration, second layer is treatment. Response: The figure has been revised; drug concentrations appear in parentheses.
• All figures. Define NI, INF and GCV etc. Response: This is now included in the specific legends.
• Testing of MCMV and HSV1/2 could use a rationale sentence. A single sentence explaining validation of specificity to CMV family. Response: We agree and have added the following statement (line 95-96): “To evaluate the antiviral activity of these compounds for future in vivo studies, and determine whether other herpesviruses were inhibited, compounds 7 and 8 were tested against MCMV and HSV1/2, respectively”.
• Letermovir is defined as “LTV” (page 9 line 128) but LET is used in Fig 7. Response: Fig. 7 has been revised, letermovir is now defined as LTV.
• Figure 7 figure legend mislabels B and C as C and D line 141. Response: Figure 7 legend  has been corrected.
• Section 2.6 – Are the compounds (specifically LTV) preventing egress? It seems to be what the data is suggesting, but not discussed. Response: The data may represent a difference in the mechanism of action between LTV and NCGC2955 analogs. LTV showed a stronger reduction in intracellular viral progeny and viral loads in supernatants. These data may also suggest production of CMV DNA containing non-infectious virus that lacks some viral proteins. A statement was added (line 224-228).
• What statistical test was used in Fig 7A, 7B, 7C. Include in figure legend and methods section. Response: A statement was added: “Statistical significance between two groups were analyzed by two-tailed Student’s t-test, and asterisks indicate the statistical significance: *, p < 0.05; **, P <0.01; *** p < 0.001”.
• TB40 strain. Is it TB40E or TB40F? If it is TB40E, does it retain UL128-131? (is it a “clinical” or wildtype strain)? Response: The UL32-EGFP-TB40 (Sampaio et al., 2005) (VR-1578) was used, page 13, lines 256-257. Although we don’t have information on UL128-UL131, based on recent PCR studies in the laboratory, this strain contains genes in the UL/b’ boundary, such as UL144, indicating it has properties of a clinical isolate.
• In the discussion, page 12, line 185-186, it is mentioned that the hepatocytes were infected with Towne and TB40. This should be clarified as Fig 4 only includes infection with Towne (not mentioned in PHH methods either). Response: That is correct, the hepatocytes can be infected with both Towne and TB40 strains, however the data in Fig. 4 is showing only results with the pp28-luciferase CMV and Towne strain. The text was modified accordingly.  
• What murine CMV strain was used and what was the source? Response: The MCMV Smith strain (ATCC VR-1399) was used. This information has been added, line 261.

Reviewer 2 Report

Comments to the author:

HCMV can cause serious diseases in immunocompromised patients. Current antiviral inhibitors (ganciclovir, cidofovir and foscarnet) all target the viral DNA polymerase. They have adverse effects and prolonged treatment can select for drug resistance mutations. Thus, we need new drugs. In this context, Guo and co- performed a high throughput screen using a CMV-luciferase reporter Towne. They identified inhibitors of CMV replication. Then, they synthesized and tested in vitro 13 analogs of the original NCGC2955 hit.

The manuscript is well written. The review is highly relevant to the field since we need new drugs targeting other stages of the viral replication.

Comments:

1. Page 2, line 54. What about letermovir mechanism of action? The authors should extend the DNA packaging steps and cite Heming et al., 2017 and Ligat et al., 2018.
2. The authors could also test a combination of newly identified HCMV inhibitors with letermovir.
3. The authors need to clearly demonstrate the clinical relevance and impact of their study.
4. Did you check other strains? clinical strains?
5. More generally, what about the toxicity of the drugs?

Author Response

We thank the reviewers for their careful and thoughtful review. We have addressed all the comments below, which allowed us to improve the manuscript and figures.

Reviewer #2

HCMV can cause serious diseases in immunocompromised patients. Current antiviral inhibitors (ganciclovir, cidofovir and foscarnet) all target the viral DNA polymerase. They have adverse effects and prolonged treatment can select for drug resistance mutations. Thus, we need new drugs. In this context, Guo and co- performed a high throughput screen using a CMV-luciferase reporter Towne. They identified inhibitors of CMV replication. Then, they synthesized and tested in vitro 13 analogs of the original NCGC2955 hit.

The manuscript is well written. The review is highly relevant to the field since we need new drugs targeting other stages of the viral replication.

 Comments:

1. Page 2, line 54. What about letermovir mechanism of action? The authors should extend the DNA packaging steps and cite Heming et al., 2017 and Ligat et al., 2018. Response: A statement was added (Line 222-225): The terminase complex, targeted by LTV,   cleaves DNA to package the viral genome into the capsid (Ligat et al., 2018). Compared to NCGC2955 analogs, LTV showed a stronger reduction in viral progeny and viral loads in supernatants. These data may suggest production of CMV DNA containing non-infectious virus that lacks some viral proteins. 
2. The authors could also test a combination of newly identified HCMV inhibitors with letermovir. Response: In our original report, NCGC2955 showed additive activity with letermovir (Kapoor et al, J Med Chem 2020). Here we aimed to report on one of the analogs (Compound 8) with the other compounds of which their mechanism of action is undefined.
3. The authors need to clearly demonstrate the clinical relevance and impact of their study. Response: Our study is the beginning of structural activity relationship following a large, high throughput screening for CMV inhibitors. Further studies including resistance selection and target identification along with structure-based design may lead to the identification of novel CMV inhibitor for in vivo studies. A statement was added, lines 227-230.
4. Did you check other strains? clinical strains? Response: Our studies included TB40E and Towne. The TB40 strain includes genes in the UL/b’ boundary that are detected in clinical strains, based on our PCR of UL144.

     5. More generally, what about the toxicity of the drugs? Response: Toxicity experiments were performed in vitro. The CC₅₀ is reported in Table 1. Line 28-29: “The CC₅₀ of Compounds 7 and 9 was > 500 µM, yielding a selectivity index  (CC₅₀ /EC₅₀ ) of >1500. The table now includes a demarcation (Reviewer 1), which shows the information more clearly

Round 2

Reviewer 2 Report

The authors have addressed the points that I brought up in the previous round of review.