Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This review systematically summarizes the links between neuroinflammation, apoptosis‑pyroptosis crosstalk and anxiety‑related neuropathology, and comprehensively reviews pre‑clinical evidence for paeoniflorin as a multi‑target neuroprotective agent. The overall structure is well‑organized, schematic figures are informative, and literature coverage is broad. Several issues should be addressed to improve the manuscript before acceptance.

First, most functional evidence for paeoniflorin is derived from in‑vitro cell experiments and rodent anxiety models. Human clinical data is extremely limited. The authors should more explicitly emphasize this critical translational gap in both abstract and conclusion sections, avoiding over‑optimistic therapeutic inference for clinical anxiety treatment. It is suggested to distinguish preclinical pharmacological effects from real‑world clinical applicability.

Second, the crosstalk between apoptosis and pyroptosis is one core theme of this paper. At present, the description mainly scatters in separate subsections. A short dedicated paragraph or brief summary table may be added to better clarify how caspase‑1, caspase‑8 and Bid mediate signal amplification between the two cell‑death programs under neuroinflammatory conditions, helping readers grasp this complex network rapidly.

Third, when discussing signaling pathways modulated by paeoniflorin, multiple pathways including NF‑κB, PI3K/Akt, Nrf2/HO‑1 and MAPK are listed. Some mechanistic descriptions are repetitive across different chapters. The authors need to condense duplicated content and clarify the upstream‑downstream logical relationships among these cascades, instead of enumerating each pathway independently.

Fourth, paeoniflorin shows diverse effects across different disease models. Some reported molecular actions may be context‑dependent or even conflicting. Please add brief discussion about context‑specific variability of paeoniflorin’s bioactivity, for example cell type, drug concentration and pathological background which may alter its regulatory output.

Fifth, this review focuses on neuroinflammation‑driven anxiety. However, anxiety disorders are highly heterogeneous and can originate from multiple non‑inflammatory causes. The manuscript should add a short statement to point out that the paeoniflorin‑related therapeutic concept mainly applies to the neuroinflammation‑related subtype of anxiety, and cannot generalize to all anxiety phenotypes.

Sixth, please carefully proof‑read the whole text. Several minor issues are present: inconsistent symbol formatting for cytokines and signaling molecules; occasional grammar awkward sentences; some reference entries have incomplete or malformed citations. Check all abbreviations and make sure every abbreviation is fully expanded at first appearance. In Figure 1‑3, double‑check label consistency between figure panels and main‑text descriptions.

Finally, gut‑brain axis is briefly touched in a small number of citations. Considering current advances in microbiota‑mediated neuroinflammation in psychiatric disorders, authors may give one short paragraph to comment on potential gut‑microbiota‑related indirect mechanisms of paeoniflorin, as a future research perspective.

Taken together, this is a valuable review covering an important research topic. It can be accepted for publication after minor revision upon addressing above‑mentioned concerns.

Comments on the Quality of English Language

This review systematically summarizes the links between neuroinflammation, apoptosis‑pyroptosis crosstalk and anxiety‑related neuropathology, and comprehensively reviews pre‑clinical evidence for paeoniflorin as a multi‑target neuroprotective agent. The overall structure is well‑organized, schematic figures are informative, and literature coverage is broad. Several issues should be addressed to improve the manuscript before acceptance.

First, most functional evidence for paeoniflorin is derived from in‑vitro cell experiments and rodent anxiety models. Human clinical data is extremely limited. The authors should more explicitly emphasize this critical translational gap in both abstract and conclusion sections, avoiding over‑optimistic therapeutic inference for clinical anxiety treatment. It is suggested to distinguish preclinical pharmacological effects from real‑world clinical applicability.

Second, the crosstalk between apoptosis and pyroptosis is one core theme of this paper. At present, the description mainly scatters in separate subsections. A short dedicated paragraph or brief summary table may be added to better clarify how caspase‑1, caspase‑8 and Bid mediate signal amplification between the two cell‑death programs under neuroinflammatory conditions, helping readers grasp this complex network rapidly.

Third, when discussing signaling pathways modulated by paeoniflorin, multiple pathways including NF‑κB, PI3K/Akt, Nrf2/HO‑1 and MAPK are listed. Some mechanistic descriptions are repetitive across different chapters. The authors need to condense duplicated content and clarify the upstream‑downstream logical relationships among these cascades, instead of enumerating each pathway independently.

Fourth, paeoniflorin shows diverse effects across different disease models. Some reported molecular actions may be context‑dependent or even conflicting. Please add brief discussion about context‑specific variability of paeoniflorin’s bioactivity, for example cell type, drug concentration and pathological background which may alter its regulatory output.

Fifth, this review focuses on neuroinflammation‑driven anxiety. However, anxiety disorders are highly heterogeneous and can originate from multiple non‑inflammatory causes. The manuscript should add a short statement to point out that the paeoniflorin‑related therapeutic concept mainly applies to the neuroinflammation‑related subtype of anxiety, and cannot generalize to all anxiety phenotypes.

Sixth, please carefully proof‑read the whole text. Several minor issues are present: inconsistent symbol formatting for cytokines and signaling molecules; occasional grammar awkward sentences; some reference entries have incomplete or malformed citations. Check all abbreviations and make sure every abbreviation is fully expanded at first appearance. In Figure 1‑3, double‑check label consistency between figure panels and main‑text descriptions.

Finally, gut‑brain axis is briefly touched in a small number of citations. Considering current advances in microbiota‑mediated neuroinflammation in psychiatric disorders, authors may give one short paragraph to comment on potential gut‑microbiota‑related indirect mechanisms of paeoniflorin, as a future research perspective.

Taken together, this is a valuable review covering an important research topic. It can be accepted for publication after minor revision upon addressing above‑mentioned concerns.

Author Response

Comment 1

 

Reviewer comment: The article postulates that neuroinflammation "almost always accompanies" anxiety disorder, yet it does not consider the reverse causal relationship: how an initial neurotransmitter imbalance initiates the inflammatory cascade. In particular, the tryptophan-kynurenine pathway and IDO-mediated shunting of tryptophan away from serotonin are not adequately discussed, leaving unclear what may initiate the inflammatory cascade in the absence of an overt inflammatory trigger.

 

Response: Thank you very much for identifying this important conceptual gap. We agree that the original manuscript framed neuroinflammation too broadly and did not adequately describe bidirectional interactions among stress, neurotransmission, metabolism, and immune signaling. In the revised manuscript, we therefore softened the causal language throughout the Introduction and explicitly describe neuroinflammation as one component of an interacting network rather than a universal initiating event. We added a dedicated section on the tryptophan–kynurenine pathway, including IDO1, stress-sensitive TDO2, serotonergic substrate availability, 3-hydroxykynurenine-associated oxidative injury, quinolinic-acid/NMDA-related excitotoxicity, and kynurenic-acid signaling. We also added stress-related ATP–P2X7 signaling as an example of a non-infectious neuroimmune trigger. At the same time, we deliberately avoided replacing one overly linear model with another: the revised text states that increased tryptophan catabolism does not automatically demonstrate reduced synaptic serotonin transmission, that peripheral kynurenine changes do not directly establish CNS metabolism, and that the pathway should be considered a potential mechanistic interface rather than a universal causal mechanism in anxiety disorders.

 

Revised locations in the manuscript:

• Introduction: Lines 51–81 (multifactorial framing, bidirectional stress–neurotransmission–inflammation interactions, and endogenous danger signaling).
• Section 1.3: Lines 247–275 (oxidative stress and kynurenine-associated cellular vulnerability).
• New Section 1.4: Lines 277–316 (IDO1/TDO2, tryptophan–kynurenine metabolism, 3-HK, quinolinic acid, kynurenic acid, serotonergic/glutamatergic implications, and clinical heterogeneity).
• Conclusion: Lines 1138–1148 (neuroinflammation placed within a broader metabolic, neuroendocrine, and circuit-level framework).

 

Comment 2

 

Reviewer comment: The manuscript discusses neuroinflammation without sufficient regional specificity. The amygdala, hippocampus, and PFC should be treated as functionally distinct but interconnected compartments, and PF-related regional findings should be analyzed without assuming regional selectivity.

 

Response: We sincerely appreciate this comment and agree that the original manuscript treated the hippocampus, PFC, and amygdala too often as sites of inflammatory measurement rather than as functionally distinct components of anxiety-related networks. We therefore rewrote Section 1.2 around the basolateral amygdala (BLA), hippocampus, and prefrontal cortex, followed by an explicit network-integration paragraph. In the BLA discussion, we now distinguish increased excitability and excitation/inhibition imbalance from demonstrated excitotoxic neuronal death. In the hippocampal discussion, we separate evidence for inflammation-related changes in neurogenesis and plasticity from evidence on contextual processing and HPA feedback. In the PFC discussion, we avoid a universal descending-inhibition model and instead emphasize projection- and subregion-dependent regulation of amygdala-centered threat circuits. We also added an explicit statement that the behavioral consequences of inflammatory signaling depend on the affected cell population, anatomical region, projection, timing, and disease context. For PF, we re-audited regional evidence and now clearly distinguish a measured regional effect from proof of pharmacological regional selectivity. Non-PF BLA electrophysiological evidence is now identified as general circuit-level support rather than as a PF-specific finding.

 

Revised locations in the manuscript:

• Rewritten Section 1.2: Lines 159–229 (BLA/amygdala, hippocampus, PFC, and network integration).
• Section 7.2: Lines 880–907 (PVN direct PF evidence, hippocampal cross-model support, and explicit limits on PFC/BLA selectivity).
• Section 7.3: Lines 908–925 (mechanistic interpretation and validation priorities, including regional neurotransmitter sampling and electrophysiology).
• Table 2: Lines 826–844 (direct PF evidence organized by model, region/cell type, measured endpoints, behavioral outcome, and evidence interpretation).

 

Comment 3

 

Reviewer comment: The manuscript lacks an integrated discussion of the HPA axis, BBB-related peripheral-to-central immune communication, glucocorticoid receptor dysfunction/resistance, and their possible contribution to chronic neuroinflammation and anxiety.

 

Response: Thank you for emphasizing this important omission. We agree that the original manuscript did not adequately integrate the neuroendocrine component of stress-related pathology. We therefore added a dedicated HPA-axis subsection. The revised text now explains CRH–ACTH–glucocorticoid signaling and GR-dependent feedback, while clarifying that glucocorticoids normally exert anti-inflammatory actions and that HPA activation should not be equated with inevitable cytokine elevation. We discuss glucocorticoid-related microglial priming, chronic stress-associated peripheral glucocorticoid resistance, neurovascular/endothelial communication, and selected examples of BBB/tight-junction changes. We also revised the overly linear concepts that cytokines must first cross a damaged BBB or directly 'block cortisol receptors'. Instead, the manuscript now distinguishes endothelial signaling from barrier leakage and describes inflammatory impairment of GR signaling/function. Human peripheral glucocorticoid resistance, regional GR expression, and central HPA feedback are explicitly treated as related but non-equivalent measures. Finally, we emphasize heterogeneity in cortisol stress reactivity and avoid assuming persistent hypercortisolemia across all anxiety disorders. PF-related HPA findings are discussed separately and cautiously, with attention to time-, region-, and model-dependent effects.

 

Revised locations in the manuscript:

• New Section 1.5: Lines 318–378 (HPA physiology, glucocorticoid feedback, microglial priming, endothelial/BBB communication, GR dysfunction, cortisol heterogeneity, and dopamine-related limits).
• Section 7.1: Lines 863–879 (PF-associated serotonergic and HPA findings in SPS models, with association distinguished from mediation).
• Section 7.2: Lines 882–888 (PVN findings and the explicit statement that PVN localization does not establish normalization of the entire HPA axis).
• Section 9.4: Lines 1123–1130 (PF-associated HPA feedback and corticosteroid-receptor findings, with the full PF-mediated endocrine–immune–neurotransmitter loop identified as unproven).
• Section 8.4: Lines 988–1020 (candidate HPA/endocrine biomarkers and their limitations).

 

Comment 4

 

Reviewer comment: The PF section describes molecular suppression of inflammatory and cell-death pathways but does not sufficiently explain how these effects may relate to serotonergic transmission, dopaminergic function, GABA/glutamate balance, HPA-axis regulation, and ultimately anxiety-related behavior.

 

Response: We are very grateful for this comment because it identified the principal integrative weakness of the original PF discussion. We agree that anti-inflammatory or cytoprotective molecular changes should not be assumed to produce anxiolysis. To address this concern, we reorganized the direct PF evidence in Section 7 and added a new Section 9.4 devoted specifically to functional interpretation. We now discuss purified-PF evidence for regional 5-HT/5-HIAA changes and HPA-related measures in stress models; PF-related TDO/tryptophan-metabolism evidence from a supportive depression model; glutamate-related cellular effects together with the lack of evidence for a generalized GABAergic action; limited dopamine-related findings; and PF-associated HPA feedback/GR/MR changes. Throughout these revisions, we distinguish concurrent association, mechanistic support, and causal mediation. We explicitly state that regional tissue 5-HT is not equivalent to restored synaptic serotonergic transmission, that PF-related glutamate effects do not establish normalization of BLA excitation/inhibition balance, and that the currently available dopamine and HPA data do not prove a fixed sequence of HPA restoration followed by dopaminergic normalization and anxiolysis. We also revised Figure 3 and Table 2 to present these findings as model-dependent evidence rather than as a single fully demonstrated causal cascade.

 

Revised locations in the manuscript:

• Section 7.1: Lines 845–879 (stress-related models; direct PF behavioral, serotonergic, and HPA findings).
• Section 7.2–7.3: Lines 880–925 (regional PF findings, separation of non-PF evidence, and mechanistic validation priorities).
• New Section 9.4: Lines 1081–1137 (serotonergic, TDO/tryptophan, glutamate/GABA, dopamine, HPA, and causal-mediation boundaries).
• Figure 3 caption: Lines 813–825 (proposed integrative framework rather than a demonstrated causal mechanism).
• Table 2: Lines 826–844 (direct PF evidence and explicit interpretation limits).

 

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This review examines the roles of neuroinflammation, apoptosis, and pyroptosis in anxiety disorders, with a focus on the potential mechanisms of action of paeoniflorin and its prospects for clinical translation. Bringing these two forms of cell death and their crosstalk into a single framework is valuable. However, some sections stray from the central focus on paeoniflorin and anxiety disorders, the discussion is often repetitive, and the limitations of the evidence and conflicting findings receive insufficient attention. The authors should first systematically check whether the cited studies support the corresponding claims, and then revise the manuscript’s structure, figures and tables, and language. Specific comments are as follows:

1. The phrase “Neuroinflammation-Induced Anxiety Disorders” suggests that neuroinflammation is a demonstrated cause of anxiety disorders, which is stronger than the evidence reviewed here supports. The title should be revised to avoid implying an established causal relationship between neuroinflammation and anxiety disorders. For example: “Neuroinflammation, Apoptosis, and Pyroptosis in Anxiety Disorders: The Potential Role of Paeoniflorin.”
2. The manuscript moves between microglia, neurons, astrocytes, and other cell types, but often attributes the resulting effects to neuronal death. For each key finding, specify the cell type studied and distinguish microglial pyroptosis, neuronal injury caused indirectly by microglial inflammatory mediators, and apoptosis or pyroptosis occurring within neurons.
3. In Section 4.2 (lines 750–761, lines 784–803, lines 826–865 and 868–880), these studies provide useful mechanistic clues, but their relevance to anxiety remains indirect. The authors should identify these findings as extrapolations, present evidence from anxiety-related models separately, and qualify the conclusions of both sections accordingly.
4. In Sections 4.2 and 4.3, findings from studies of depression, Alzheimer’s disease, neuropathic pain, and liver or cardiac injury are used to infer paeoniflorin’s effects in anxiety disorders. These studies may offer mechanistic clues, but their indirect nature should be clear. The authors should distinguish evidence from patients with anxiety disorders, animal studies assessing anxiety-like behavior, other neurological disease models, and non-neural tissues or general cell models.
5. Section 7.1 mainly discusses spinosin, other compounds, herbal formulations, and exercise, although the section is titled as a review of paeoniflorin studies. These examples should be reduced to brief model background. Instead, summarize paeoniflorin studies by species, sex, anxiety model, dose, route, treatment duration, behavioral outcomes, and mechanistic findings.
6. Sections 4-6 and 9 repeatedly discuss NF-κB, NLRP3, PI3K/Akt, mitochondrial protection, and antioxidant effects. Sections 9.1 and 9.2 also report both activation and inhibition of STAT3 and PI3K/Akt by paeoniflorin, but offer little explanation beyond "context dependence." The authors should examine whether cell type, disease model, dose, or treatment duration accounts for these differences, and identify which findings come directly from anxiety studies.
7. Section 8 presents paeoniflorin’s potential advantages as a multitarget agent, along with its tolerability, personalized dosing, genetic stratification, and brain-targeted delivery, in overly optimistic terms. The authors should clarify which of these claims are supported by existing paeoniflorin studies and which remain proposals for future research.
8. The authors should systematically verify that each cited study examines the stated compound, model, cell type, and outcome. Several references appear mismatched: reference [5] concerns a TSPO ligand rather than paeoniflorin, while references [142] and [146] examine harmine and quercetin but are cited for paeoniflorin-specific claims in Section 7.2.

Author Response

Response to Reviewer

Dear Reviewer,

 

We sincerely thank you for your careful reading of our manuscript and for the thoughtful, constructive, and scientifically important comments. We fully appreciate your central concern that the original version, although detailed at the molecular level, did not sufficiently place neuroinflammation, apoptosis, pyroptosis, and paeoniflorin (PF) within the broader neurochemical, neuroendocrine, metabolic, regional, and circuit-level context of anxiety-related pathology. Your comments prompted a substantial revision of the manuscript. We revised the title, Abstract, Introduction, and Conclusion; rewrote the regional neuroinflammation section; added dedicated sections on the tryptophan–kynurenine pathway and HPA-axis/neuroimmune communication; re-audited the PF-specific evidence; reorganized the preclinical PF section; added a functional integration section linking molecular findings to neurotransmitter and neuroendocrine outcomes; expanded the translational limitations section; rebuilt Table 2; and revised Figures 1–3 to avoid unsupported or overly causal interpretations. We have also carefully distinguished direct PF evidence from cross-model evidence, findings involving other compounds or multi-component preparations, and hypothesis-generating observations.

 

Line-number note: We rechecked all locations against the latest revised manuscript (200.docx). All line numbers cited below refer to that version with continuous line numbering. Minor shifts may occur only if the journal system reformats the file.

 

Comment 1

 

Reviewer comment: The article postulates that neuroinflammation "almost always accompanies" anxiety disorder, yet it does not consider the reverse causal relationship: how an initial neurotransmitter imbalance initiates the inflammatory cascade. In particular, the tryptophan-kynurenine pathway and IDO-mediated shunting of tryptophan away from serotonin are not adequately discussed, leaving unclear what may initiate the inflammatory cascade in the absence of an overt inflammatory trigger.

 

Response: Thank you very much for identifying this important conceptual gap. We agree that the original manuscript framed neuroinflammation too broadly and did not adequately describe bidirectional interactions among stress, neurotransmission, metabolism, and immune signaling. In the revised manuscript, we therefore softened the causal language throughout the Introduction and explicitly describe neuroinflammation as one component of an interacting network rather than a universal initiating event. We added a dedicated section on the tryptophan–kynurenine pathway, including IDO1, stress-sensitive TDO2, serotonergic substrate availability, 3-hydroxykynurenine-associated oxidative injury, quinolinic-acid/NMDA-related excitotoxicity, and kynurenic-acid signaling. We also added stress-related ATP–P2X7 signaling as an example of a non-infectious neuroimmune trigger. At the same time, we deliberately avoided replacing one overly linear model with another: the revised text states that increased tryptophan catabolism does not automatically demonstrate reduced synaptic serotonin transmission, that peripheral kynurenine changes do not directly establish CNS metabolism, and that the pathway should be considered a potential mechanistic interface rather than a universal causal mechanism in anxiety disorders.

 

Revised locations in the manuscript:

• Introduction: Lines 51–81 (multifactorial framing, bidirectional stress–neurotransmission–inflammation interactions, and endogenous danger signaling).
• Section 1.3: Lines 247–275 (oxidative stress and kynurenine-associated cellular vulnerability).
• New Section 1.4: Lines 277–316 (IDO1/TDO2, tryptophan–kynurenine metabolism, 3-HK, quinolinic acid, kynurenic acid, serotonergic/glutamatergic implications, and clinical heterogeneity).
• Conclusion: Lines 1138–1148 (neuroinflammation placed within a broader metabolic, neuroendocrine, and circuit-level framework).

 

Comment 2

 

Reviewer comment: The manuscript discusses neuroinflammation without sufficient regional specificity. The amygdala, hippocampus, and PFC should be treated as functionally distinct but interconnected compartments, and PF-related regional findings should be analyzed without assuming regional selectivity.

 

Response: We sincerely appreciate this comment and agree that the original manuscript treated the hippocampus, PFC, and amygdala too often as sites of inflammatory measurement rather than as functionally distinct components of anxiety-related networks. We therefore rewrote Section 1.2 around the basolateral amygdala (BLA), hippocampus, and prefrontal cortex, followed by an explicit network-integration paragraph. In the BLA discussion, we now distinguish increased excitability and excitation/inhibition imbalance from demonstrated excitotoxic neuronal death. In the hippocampal discussion, we separate evidence for inflammation-related changes in neurogenesis and plasticity from evidence on contextual processing and HPA feedback. In the PFC discussion, we avoid a universal descending-inhibition model and instead emphasize projection- and subregion-dependent regulation of amygdala-centered threat circuits. We also added an explicit statement that the behavioral consequences of inflammatory signaling depend on the affected cell population, anatomical region, projection, timing, and disease context. For PF, we re-audited regional evidence and now clearly distinguish a measured regional effect from proof of pharmacological regional selectivity. Non-PF BLA electrophysiological evidence is now identified as general circuit-level support rather than as a PF-specific finding.

 

Revised locations in the manuscript:

• Rewritten Section 1.2: Lines 159–229 (BLA/amygdala, hippocampus, PFC, and network integration).
• Section 7.2: Lines 880–907 (PVN direct PF evidence, hippocampal cross-model support, and explicit limits on PFC/BLA selectivity).
• Section 7.3: Lines 908–925 (mechanistic interpretation and validation priorities, including regional neurotransmitter sampling and electrophysiology).
• Table 2: Lines 826–844 (direct PF evidence organized by model, region/cell type, measured endpoints, behavioral outcome, and evidence interpretation).

 

Comment 3

 

Reviewer comment: The manuscript lacks an integrated discussion of the HPA axis, BBB-related peripheral-to-central immune communication, glucocorticoid receptor dysfunction/resistance, and their possible contribution to chronic neuroinflammation and anxiety.

 

Response: Thank you for emphasizing this important omission. We agree that the original manuscript did not adequately integrate the neuroendocrine component of stress-related pathology. We therefore added a dedicated HPA-axis subsection. The revised text now explains CRH–ACTH–glucocorticoid signaling and GR-dependent feedback, while clarifying that glucocorticoids normally exert anti-inflammatory actions and that HPA activation should not be equated with inevitable cytokine elevation. We discuss glucocorticoid-related microglial priming, chronic stress-associated peripheral glucocorticoid resistance, neurovascular/endothelial communication, and selected examples of BBB/tight-junction changes. We also revised the overly linear concepts that cytokines must first cross a damaged BBB or directly 'block cortisol receptors'. Instead, the manuscript now distinguishes endothelial signaling from barrier leakage and describes inflammatory impairment of GR signaling/function. Human peripheral glucocorticoid resistance, regional GR expression, and central HPA feedback are explicitly treated as related but non-equivalent measures. Finally, we emphasize heterogeneity in cortisol stress reactivity and avoid assuming persistent hypercortisolemia across all anxiety disorders. PF-related HPA findings are discussed separately and cautiously, with attention to time-, region-, and model-dependent effects.

 

Revised locations in the manuscript:

• New Section 1.5: Lines 318–378 (HPA physiology, glucocorticoid feedback, microglial priming, endothelial/BBB communication, GR dysfunction, cortisol heterogeneity, and dopamine-related limits).
• Section 7.1: Lines 863–879 (PF-associated serotonergic and HPA findings in SPS models, with association distinguished from mediation).
• Section 7.2: Lines 882–888 (PVN findings and the explicit statement that PVN localization does not establish normalization of the entire HPA axis).
• Section 9.4: Lines 1123–1130 (PF-associated HPA feedback and corticosteroid-receptor findings, with the full PF-mediated endocrine–immune–neurotransmitter loop identified as unproven).
• Section 8.4: Lines 988–1020 (candidate HPA/endocrine biomarkers and their limitations).

 

Comment 4

 

Reviewer comment: The PF section describes molecular suppression of inflammatory and cell-death pathways but does not sufficiently explain how these effects may relate to serotonergic transmission, dopaminergic function, GABA/glutamate balance, HPA-axis regulation, and ultimately anxiety-related behavior.

 

Response: We are very grateful for this comment because it identified the principal integrative weakness of the original PF discussion. We agree that anti-inflammatory or cytoprotective molecular changes should not be assumed to produce anxiolysis. To address this concern, we reorganized the direct PF evidence in Section 7 and added a new Section 9.4 devoted specifically to functional interpretation. We now discuss purified-PF evidence for regional 5-HT/5-HIAA changes and HPA-related measures in stress models; PF-related TDO/tryptophan-metabolism evidence from a supportive depression model; glutamate-related cellular effects together with the lack of evidence for a generalized GABAergic action; limited dopamine-related findings; and PF-associated HPA feedback/GR/MR changes. Throughout these revisions, we distinguish concurrent association, mechanistic support, and causal mediation. We explicitly state that regional tissue 5-HT is not equivalent to restored synaptic serotonergic transmission, that PF-related glutamate effects do not establish normalization of BLA excitation/inhibition balance, and that the currently available dopamine and HPA data do not prove a fixed sequence of HPA restoration followed by dopaminergic normalization and anxiolysis. We also revised Figure 3 and Table 2 to present these findings as model-dependent evidence rather than as a single fully demonstrated causal cascade.

 

Revised locations in the manuscript:

• Section 7.1: Lines 845–879 (stress-related models; direct PF behavioral, serotonergic, and HPA findings).
• Section 7.2–7.3: Lines 880–925 (regional PF findings, separation of non-PF evidence, and mechanistic validation priorities).
• New Section 9.4: Lines 1081–1137 (serotonergic, TDO/tryptophan, glutamate/GABA, dopamine, HPA, and causal-mediation boundaries).
• Figure 3 caption: Lines 813–825 (proposed integrative framework rather than a demonstrated causal mechanism).
• Table 2: Lines 826–844 (direct PF evidence and explicit interpretation limits).

 

Response to the Reviewer’s Recommendations

 

Recommendation 1

 

Reviewer recommendation: Add to the introduction and/or a separate section a description of the tryptophan–kynurenine pathway and the role of IDO as a link between neurotransmitter imbalance and neuroinflammation in anxiety disorders.

 

Response: Addressed. We added the bidirectional framework in the Introduction and a dedicated Section 1.4 on tryptophan–kynurenine metabolism, including IDO1, TDO2, serotonergic substrate availability, 3-HK, quinolinic acid, kynurenic acid, oxidative/excitotoxic stress, and evidence limitations.

 

Location: Lines 68–81 and 277–316.

 

Recommendation 2

 

Reviewer recommendation: Introduce regional specificity: describe the inflammatory dysfunction of the amygdala, hippocampus, and PFC as interconnected compartments.

 

Response: Addressed. Section 1.2 was rewritten around the BLA/amygdala, hippocampus, and PFC, followed by an explicit network-integration discussion. PF regional findings were also reclassified so that regional measurement is not equated with regional selectivity.

 

Location: Lines 159–229 and 880–907.

 

Recommendation 3

 

Reviewer recommendation: Include a section on the HPA axis, mechanisms of BBB damage by cytokines, and the role of cortisol resistance in the chronification of neuroinflammation and anxiety.

 

Response: Addressed with a dedicated Section 1.5. We discuss normal HPA physiology, glucocorticoid feedback, microglial priming, peripheral glucocorticoid resistance, endothelial/neurovascular communication, selected BBB permeability changes, inflammatory impairment of GR signaling, and heterogeneity of cortisol responses. We intentionally avoid presenting BBB disruption or hypercortisolemia as universal features.

 

Location: Lines 318–378.

 

Recommendation 4

 

Reviewer recommendation: Revise the PF therapeutic section by adding discussion of tryptophan metabolism/serotonergic transmission, dopaminergic function, HPA feedback, and GABA/glutamate balance.

 

Response: Addressed by reorganizing Section 7 and adding Section 9.4. Direct PF findings and cross-model support are now separated. We discuss PF-related 5-HT/5-HIAA findings, TDO/tryptophan metabolism, glutamate-related cellular responses, the lack of generalized GABAergic evidence, limited dopamine data, and HPA feedback/GR/MR findings. Proposed links are explicitly distinguished from causal mediation.

 

Location: Lines 845–925 and 1081–1137.

 

Recommendation 5

 

Reviewer recommendation: Add a section with the limitations of preclinical data and prospects for translation, including doses, bioavailability, and possible biomarkers of response.

 

Response: Addressed by restructuring Section 8 into model validity/causal inference, dose-route-bioavailability/exposure–response, safety/clinical development, and candidate biomarkers/response stratification. We also added regimen and evidence-boundary information to Table 2 and clarified that animal regimens do not define clinical doses.

 

Location: Lines 926–1020; Table 2 at Lines 826–844.

 

Additional Evidence Corrections Made During Revision

 

In addition to the requested conceptual revisions, we performed a detailed evidence audit throughout the PF sections. We corrected or removed statements that could imply a direct PF effect when the underlying study involved another compound, a multi-component preparation, a PF derivative, a peripheral disease model, or an indirect computational result. For example, we now explicitly separate findings involving harmine, fustin, hydrogen sulfide, quercetin, puerarin, and magnoflorine from PF-specific evidence (Lines 626–634). We also treat death receptor-related PF involvement as unresolved where direct evidence is lacking (Lines 721–733); distinguish reduced NLRP3-pathway markers from direct PF–NLRP3 binding (Lines 736–769); distinguish pathway modulation from direct caspase-1–GSDMD blockade and separately identify peripheral GSDMD target-engagement evidence (Lines 770–791); and explicitly state that parallel apoptotic and pyroptotic changes do not establish a single causal PF-regulated cell-death cascade (Lines 792–811).

 

Revisions to Figures and Tables

 

We revised the visual material so that the figures follow the same evidence hierarchy as the text. Figure 1 was corrected to distinguish NLRP3 priming from activation, remove fixed M1 terminology, clarify the context-dependent caspase-1/BID relationship, and identify the primarily rodent basis of caspase-12-related evidence (Lines 231–246). Figure 2 was revised to avoid implying unverified direct PF binding or a universal direction of STAT3/PI3K/Akt signaling; its caption now explicitly states that the depicted pathways derive from different experimental systems and require model-specific validation (Lines 691–705). Figure 3 was revised as a proposed integrative framework rather than a demonstrated causal chain, and its caption now states that regional observations and concurrent molecular, neurochemical, neuroendocrine, and behavioral changes do not by themselves establish PF regional selectivity, causal anxiolytic mediation, or clinical efficacy (Lines 813–825). Table 2 was rebuilt to summarize direct PF evidence across cellular, regional, neurochemical, and neuroendocrine domains, with model, regimen, region/cell type, measured outcomes, behavioral outcomes, and evidence interpretation presented explicitly (Lines 826–844). Table 1 was also revised to separate established cellular mechanisms from their model-dependent relevance to anxiety-related pathology (Lines 553–566).

 

Closing Response

 

We again sincerely thank the reviewer for these thoughtful and constructive comments. The revision has substantially improved the conceptual coherence, evidentiary precision, and translational balance of the manuscript. We have tried to address each concern carefully while avoiding conclusions that exceed the available evidence. In particular, the revised manuscript now distinguishes established cellular mechanisms from anxiety-specific evidence, direct PF findings from cross-model support, regional observations from circuit restoration, and concurrent molecular/behavioral changes from causal mediation. We hope that the revised manuscript adequately addresses the reviewer’s concerns and is now clearer, more rigorous, and more useful to readers.

 

With sincere thanks and respect,

 

The Authors

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

The article is a review work dedicated to the therapeutic potential of paeoniflorin (PF) in anxiety disorders through the modulation of neuroinflammation, apoptosis, and pyroptosis. The authors describe in detail and in a structured manner the molecular mechanisms of programmed cell death, microglial activation, oxidative stress, and signaling pathways, as well as the pharmacokinetics and pleiotropic effects of PF.

An extremely detailed analysis of intracellular cascades, good visualization, coverage of experimental models, and genetic knockouts are provided.

However, despite the abundance of molecular details, the review demonstrates a fragmentary pathophysiological picture: neuroinflammation is considered in isolation, outside the systemic context of anxiety disorder. The main remarks are formulated below.

1. The article postulates that neuroinflammation "almost always accompanies" anxiety disorder, yet it does not consider the reverse causal relationship: how an initial neurotransmitter imbalance initiates the inflammatory cascade. In particular, the tryptophan-kynurenine pathway, which is central to the pathogenesis of anxiety disorder, is not described: activation of indoleamine-2,3-dioxygenase (IDO) by pro-inflammatory cytokines (IFN-γ, TNF-α, IL-6) shifts tryptophan metabolism away from the serotonergic pathway toward kynurenine. This leads to the accumulation of neurotoxic metabolites (3-hydroxykynurenine, quinolinic acid), enhances excitotoxicity via NMDA receptors, and stimulates microglial activation. At the same time, reduced tryptophan availability for serotonin synthesis exacerbates the serotonergic deficiency underlying anxiety. This creates a vicious circle "neurotransmitter imbalance → neuroinflammation → further imbalance," which is not even mentioned in the introduction of the review. Without this context, it is unclear to the reader what exactly triggers the cascades described by the authors in the absence of an external inflammatory stimulus (LPS, trauma), and why anxiety disorders often manifest without an obvious pro-inflammatory trigger.

 

2. The authors operate with the concept of "neuroinflammation in the brain," mentioning the hippocampus, PFC, and amygdala predominantly as sites of cytokine measurement rather than as functionally distinct structural elements with their own inflammatory logic.For example, inflammatory activation of the amygdala enhances excitotoxicity of glutamatergic circuits (basolateral nucleus), which forms a self-sustaining loop of hyperexcitability and anxiety. Neuroinflammation in the hippocampus disrupts neurogenesis and contextual threat processing, which contributes to the generalization of anxiety. Finally, inflammatory dysfunction of the prefrontal cortex reduces descending inhibitory control over the amygdala, disinhibiting the fear response. Without such regional specificity, it is impossible to understand why the same cytokines produce anxious symptomatology rather than, for instance, depressive or cognitive symptoms. This is especially important for the discussion of PF, which, judging by the presented data, acts in different structures (hippocampus, PFC, BLA), but this selectivity is not analyzed at all.

 

3. There is a complete absence of discussion of the hypothalamic-pituitary-adrenal (HPA) axis, a key link connecting stress, neuroinflammation, and anxiety. Stress-induced activation of the HPA axis → increased systemic pro-inflammatory cytokines (IL-1β, IL-6, TNF-α), cytokines penetrate the brain through the damaged BBB and activate microglia toward a pro-inflammatory phenotype. In turn, pro-inflammatory cytokines released by microglia block cortisol receptors (GR) in the hippocampus and hypothalamus, disrupting negative feedback, which ultimately leads to hypercortisolemia, which, in turn, further damages the hippocampus and disrupts dopaminergic transmission (mesolimbic pathway), exacerbating anxiety. This is a central vicious circle, without which it is impossible to explain either the chronification of anxiety disorder or the systemic nature of neuroinflammation. Its absence is a serious conceptual gap, especially since many of the models cited by the authors (CRS, LPS) specifically involve the HPA axis.

 

4. The authors describe in detail how PF suppresses TLR4/NF-κB, NLRP3, caspase-1/3, Bax/Bcl-2, etc. However, nowhere is it discussed how these effects may, for example, restore serotonergic transmission; normalize dopaminergic function; influence GABAergic/glutamatergic balance in the amygdala (the authors only indirectly touch upon this issue in section 7.2, but without connection to neurotransmitters); or modulate the HPA axis (there is not a word about cortisol, GR, CRH). Since anxiety disorders are primarily neurotransmitter disorders, and neuroinflammation is merely one of the pathophysiological mechanisms, the review loses its integrity: it describes how PF suppresses inflammation and cell death but does not answer the question of why this should lead to a clinically significant anxiolytic effect.

Recommendations to the authors

1. Add to the introduction and/or a separate section a description of the tryptophan-kynurenine pathway and the role of IDO as a link between neurotransmitter imbalance and neuroinflammation in anxiety disorders.
2. Introduce regional specificity: describe the inflammatory dysfunction of the amygdala (excitotoxicity), hippocampus (neurogenesis, HPA feedback), and PFC (descending control) as interconnected compartments.
3. Include a section on the HPA axis, mechanisms of BBB damage by cytokines, and the role of cortisol resistance in the chronification of neuroinflammation and anxiety.
4. Revise the section on the therapeutic effects of PF by adding a discussion (at least hypothetical) of its influence on tryptophan metabolism and serotonergic transmission; dopaminergic function through restoration of HPA feedback; and GABA/glutamate balance in the amygdala.
5. Add a section with the limitations of preclinical data and prospects for translation, including a discussion of doses, bioavailability, and possible biomarkers of response.

 

Author Response

Response to Reviewer

Dear Reviewer,

 

We sincerely thank you for your careful reading of our manuscript and for the thoughtful, constructive, and scientifically important comments. We fully appreciate your central concern that the original version, although detailed at the molecular level, did not sufficiently place neuroinflammation, apoptosis, pyroptosis, and paeoniflorin (PF) within the broader neurochemical, neuroendocrine, metabolic, regional, and circuit-level context of anxiety-related pathology. Your comments prompted a substantial revision of the manuscript. We revised the title, Abstract, Introduction, and Conclusion; rewrote the regional neuroinflammation section; added dedicated sections on the tryptophan–kynurenine pathway and HPA-axis/neuroimmune communication; re-audited the PF-specific evidence; reorganized the preclinical PF section; added a functional integration section linking molecular findings to neurotransmitter and neuroendocrine outcomes; expanded the translational limitations section; rebuilt Table 2; and revised Figures 1–3 to avoid unsupported or overly causal interpretations. We have also carefully distinguished direct PF evidence from cross-model evidence, findings involving other compounds or multi-component preparations, and hypothesis-generating observations.

 

Line-number note: We rechecked all locations against the latest revised manuscript (200.docx). All line numbers cited below refer to that version with continuous line numbering. Minor shifts may occur only if the journal system reformats the file.

 

Comment 1

 

Reviewer comment: The article postulates that neuroinflammation "almost always accompanies" anxiety disorder, yet it does not consider the reverse causal relationship: how an initial neurotransmitter imbalance initiates the inflammatory cascade. In particular, the tryptophan-kynurenine pathway and IDO-mediated shunting of tryptophan away from serotonin are not adequately discussed, leaving unclear what may initiate the inflammatory cascade in the absence of an overt inflammatory trigger.

 

Response: Thank you very much for identifying this important conceptual gap. We agree that the original manuscript framed neuroinflammation too broadly and did not adequately describe bidirectional interactions among stress, neurotransmission, metabolism, and immune signaling. In the revised manuscript, we therefore softened the causal language throughout the Introduction and explicitly describe neuroinflammation as one component of an interacting network rather than a universal initiating event. We added a dedicated section on the tryptophan–kynurenine pathway, including IDO1, stress-sensitive TDO2, serotonergic substrate availability, 3-hydroxykynurenine-associated oxidative injury, quinolinic-acid/NMDA-related excitotoxicity, and kynurenic-acid signaling. We also added stress-related ATP–P2X7 signaling as an example of a non-infectious neuroimmune trigger. At the same time, we deliberately avoided replacing one overly linear model with another: the revised text states that increased tryptophan catabolism does not automatically demonstrate reduced synaptic serotonin transmission, that peripheral kynurenine changes do not directly establish CNS metabolism, and that the pathway should be considered a potential mechanistic interface rather than a universal causal mechanism in anxiety disorders.

 

Revised locations in the manuscript:

• Introduction: Lines 51–81 (multifactorial framing, bidirectional stress–neurotransmission–inflammation interactions, and endogenous danger signaling).
• Section 1.3: Lines 247–275 (oxidative stress and kynurenine-associated cellular vulnerability).
• New Section 1.4: Lines 277–316 (IDO1/TDO2, tryptophan–kynurenine metabolism, 3-HK, quinolinic acid, kynurenic acid, serotonergic/glutamatergic implications, and clinical heterogeneity).
• Conclusion: Lines 1138–1148 (neuroinflammation placed within a broader metabolic, neuroendocrine, and circuit-level framework).

 

Comment 2

 

Reviewer comment: The manuscript discusses neuroinflammation without sufficient regional specificity. The amygdala, hippocampus, and PFC should be treated as functionally distinct but interconnected compartments, and PF-related regional findings should be analyzed without assuming regional selectivity.

 

Response: We sincerely appreciate this comment and agree that the original manuscript treated the hippocampus, PFC, and amygdala too often as sites of inflammatory measurement rather than as functionally distinct components of anxiety-related networks. We therefore rewrote Section 1.2 around the basolateral amygdala (BLA), hippocampus, and prefrontal cortex, followed by an explicit network-integration paragraph. In the BLA discussion, we now distinguish increased excitability and excitation/inhibition imbalance from demonstrated excitotoxic neuronal death. In the hippocampal discussion, we separate evidence for inflammation-related changes in neurogenesis and plasticity from evidence on contextual processing and HPA feedback. In the PFC discussion, we avoid a universal descending-inhibition model and instead emphasize projection- and subregion-dependent regulation of amygdala-centered threat circuits. We also added an explicit statement that the behavioral consequences of inflammatory signaling depend on the affected cell population, anatomical region, projection, timing, and disease context. For PF, we re-audited regional evidence and now clearly distinguish a measured regional effect from proof of pharmacological regional selectivity. Non-PF BLA electrophysiological evidence is now identified as general circuit-level support rather than as a PF-specific finding.

 

Revised locations in the manuscript:

• Rewritten Section 1.2: Lines 159–229 (BLA/amygdala, hippocampus, PFC, and network integration).
• Section 7.2: Lines 880–907 (PVN direct PF evidence, hippocampal cross-model support, and explicit limits on PFC/BLA selectivity).
• Section 7.3: Lines 908–925 (mechanistic interpretation and validation priorities, including regional neurotransmitter sampling and electrophysiology).
• Table 2: Lines 826–844 (direct PF evidence organized by model, region/cell type, measured endpoints, behavioral outcome, and evidence interpretation).

 

Comment 3

 

Reviewer comment: The manuscript lacks an integrated discussion of the HPA axis, BBB-related peripheral-to-central immune communication, glucocorticoid receptor dysfunction/resistance, and their possible contribution to chronic neuroinflammation and anxiety.

 

Response: Thank you for emphasizing this important omission. We agree that the original manuscript did not adequately integrate the neuroendocrine component of stress-related pathology. We therefore added a dedicated HPA-axis subsection. The revised text now explains CRH–ACTH–glucocorticoid signaling and GR-dependent feedback, while clarifying that glucocorticoids normally exert anti-inflammatory actions and that HPA activation should not be equated with inevitable cytokine elevation. We discuss glucocorticoid-related microglial priming, chronic stress-associated peripheral glucocorticoid resistance, neurovascular/endothelial communication, and selected examples of BBB/tight-junction changes. We also revised the overly linear concepts that cytokines must first cross a damaged BBB or directly 'block cortisol receptors'. Instead, the manuscript now distinguishes endothelial signaling from barrier leakage and describes inflammatory impairment of GR signaling/function. Human peripheral glucocorticoid resistance, regional GR expression, and central HPA feedback are explicitly treated as related but non-equivalent measures. Finally, we emphasize heterogeneity in cortisol stress reactivity and avoid assuming persistent hypercortisolemia across all anxiety disorders. PF-related HPA findings are discussed separately and cautiously, with attention to time-, region-, and model-dependent effects.

 

Revised locations in the manuscript:

• New Section 1.5: Lines 318–378 (HPA physiology, glucocorticoid feedback, microglial priming, endothelial/BBB communication, GR dysfunction, cortisol heterogeneity, and dopamine-related limits).
• Section 7.1: Lines 863–879 (PF-associated serotonergic and HPA findings in SPS models, with association distinguished from mediation).
• Section 7.2: Lines 882–888 (PVN findings and the explicit statement that PVN localization does not establish normalization of the entire HPA axis).
• Section 9.4: Lines 1123–1130 (PF-associated HPA feedback and corticosteroid-receptor findings, with the full PF-mediated endocrine–immune–neurotransmitter loop identified as unproven).
• Section 8.4: Lines 988–1020 (candidate HPA/endocrine biomarkers and their limitations).

 

Comment 4

 

Reviewer comment: The PF section describes molecular suppression of inflammatory and cell-death pathways but does not sufficiently explain how these effects may relate to serotonergic transmission, dopaminergic function, GABA/glutamate balance, HPA-axis regulation, and ultimately anxiety-related behavior.

 

Response: We are very grateful for this comment because it identified the principal integrative weakness of the original PF discussion. We agree that anti-inflammatory or cytoprotective molecular changes should not be assumed to produce anxiolysis. To address this concern, we reorganized the direct PF evidence in Section 7 and added a new Section 9.4 devoted specifically to functional interpretation. We now discuss purified-PF evidence for regional 5-HT/5-HIAA changes and HPA-related measures in stress models; PF-related TDO/tryptophan-metabolism evidence from a supportive depression model; glutamate-related cellular effects together with the lack of evidence for a generalized GABAergic action; limited dopamine-related findings; and PF-associated HPA feedback/GR/MR changes. Throughout these revisions, we distinguish concurrent association, mechanistic support, and causal mediation. We explicitly state that regional tissue 5-HT is not equivalent to restored synaptic serotonergic transmission, that PF-related glutamate effects do not establish normalization of BLA excitation/inhibition balance, and that the currently available dopamine and HPA data do not prove a fixed sequence of HPA restoration followed by dopaminergic normalization and anxiolysis. We also revised Figure 3 and Table 2 to present these findings as model-dependent evidence rather than as a single fully demonstrated causal cascade.

 

Revised locations in the manuscript:

• Section 7.1: Lines 845–879 (stress-related models; direct PF behavioral, serotonergic, and HPA findings).
• Section 7.2–7.3: Lines 880–925 (regional PF findings, separation of non-PF evidence, and mechanistic validation priorities).
• New Section 9.4: Lines 1081–1137 (serotonergic, TDO/tryptophan, glutamate/GABA, dopamine, HPA, and causal-mediation boundaries).
• Figure 3 caption: Lines 813–825 (proposed integrative framework rather than a demonstrated causal mechanism).
• Table 2: Lines 826–844 (direct PF evidence and explicit interpretation limits).

 

Response to the Reviewer’s Recommendations

 

Recommendation 1

 

Reviewer recommendation: Add to the introduction and/or a separate section a description of the tryptophan–kynurenine pathway and the role of IDO as a link between neurotransmitter imbalance and neuroinflammation in anxiety disorders.

 

Response: Addressed. We added the bidirectional framework in the Introduction and a dedicated Section 1.4 on tryptophan–kynurenine metabolism, including IDO1, TDO2, serotonergic substrate availability, 3-HK, quinolinic acid, kynurenic acid, oxidative/excitotoxic stress, and evidence limitations.

 

Location: Lines 68–81 and 277–316.

 

Recommendation 2

 

Reviewer recommendation: Introduce regional specificity: describe the inflammatory dysfunction of the amygdala, hippocampus, and PFC as interconnected compartments.

 

Response: Addressed. Section 1.2 was rewritten around the BLA/amygdala, hippocampus, and PFC, followed by an explicit network-integration discussion. PF regional findings were also reclassified so that regional measurement is not equated with regional selectivity.

 

Location: Lines 159–229 and 880–907.

 

Recommendation 3

 

Reviewer recommendation: Include a section on the HPA axis, mechanisms of BBB damage by cytokines, and the role of cortisol resistance in the chronification of neuroinflammation and anxiety.

 

Response: Addressed with a dedicated Section 1.5. We discuss normal HPA physiology, glucocorticoid feedback, microglial priming, peripheral glucocorticoid resistance, endothelial/neurovascular communication, selected BBB permeability changes, inflammatory impairment of GR signaling, and heterogeneity of cortisol responses. We intentionally avoid presenting BBB disruption or hypercortisolemia as universal features.

 

Location: Lines 318–378.

 

Recommendation 4

 

Reviewer recommendation: Revise the PF therapeutic section by adding discussion of tryptophan metabolism/serotonergic transmission, dopaminergic function, HPA feedback, and GABA/glutamate balance.

 

Response: Addressed by reorganizing Section 7 and adding Section 9.4. Direct PF findings and cross-model support are now separated. We discuss PF-related 5-HT/5-HIAA findings, TDO/tryptophan metabolism, glutamate-related cellular responses, the lack of generalized GABAergic evidence, limited dopamine data, and HPA feedback/GR/MR findings. Proposed links are explicitly distinguished from causal mediation.

 

Location: Lines 845–925 and 1081–1137.

 

Recommendation 5

 

Reviewer recommendation: Add a section with the limitations of preclinical data and prospects for translation, including doses, bioavailability, and possible biomarkers of response.

 

Response: Addressed by restructuring Section 8 into model validity/causal inference, dose-route-bioavailability/exposure–response, safety/clinical development, and candidate biomarkers/response stratification. We also added regimen and evidence-boundary information to Table 2 and clarified that animal regimens do not define clinical doses.

 

Location: Lines 926–1020; Table 2 at Lines 826–844.

 

Additional Evidence Corrections Made During Revision

 

In addition to the requested conceptual revisions, we performed a detailed evidence audit throughout the PF sections. We corrected or removed statements that could imply a direct PF effect when the underlying study involved another compound, a multi-component preparation, a PF derivative, a peripheral disease model, or an indirect computational result. For example, we now explicitly separate findings involving harmine, fustin, hydrogen sulfide, quercetin, puerarin, and magnoflorine from PF-specific evidence (Lines 626–634). We also treat death receptor-related PF involvement as unresolved where direct evidence is lacking (Lines 721–733); distinguish reduced NLRP3-pathway markers from direct PF–NLRP3 binding (Lines 736–769); distinguish pathway modulation from direct caspase-1–GSDMD blockade and separately identify peripheral GSDMD target-engagement evidence (Lines 770–791); and explicitly state that parallel apoptotic and pyroptotic changes do not establish a single causal PF-regulated cell-death cascade (Lines 792–811).

 

Revisions to Figures and Tables

 

We revised the visual material so that the figures follow the same evidence hierarchy as the text. Figure 1 was corrected to distinguish NLRP3 priming from activation, remove fixed M1 terminology, clarify the context-dependent caspase-1/BID relationship, and identify the primarily rodent basis of caspase-12-related evidence (Lines 231–246). Figure 2 was revised to avoid implying unverified direct PF binding or a universal direction of STAT3/PI3K/Akt signaling; its caption now explicitly states that the depicted pathways derive from different experimental systems and require model-specific validation (Lines 691–705). Figure 3 was revised as a proposed integrative framework rather than a demonstrated causal chain, and its caption now states that regional observations and concurrent molecular, neurochemical, neuroendocrine, and behavioral changes do not by themselves establish PF regional selectivity, causal anxiolytic mediation, or clinical efficacy (Lines 813–825). Table 2 was rebuilt to summarize direct PF evidence across cellular, regional, neurochemical, and neuroendocrine domains, with model, regimen, region/cell type, measured outcomes, behavioral outcomes, and evidence interpretation presented explicitly (Lines 826–844). Table 1 was also revised to separate established cellular mechanisms from their model-dependent relevance to anxiety-related pathology (Lines 553–566).

 

Closing Response

 

We again sincerely thank the reviewer for these thoughtful and constructive comments. The revision has substantially improved the conceptual coherence, evidentiary precision, and translational balance of the manuscript. We have tried to address each concern carefully while avoiding conclusions that exceed the available evidence. In particular, the revised manuscript now distinguishes established cellular mechanisms from anxiety-specific evidence, direct PF findings from cross-model support, regional observations from circuit restoration, and concurrent molecular/behavioral changes from causal mediation. We hope that the revised manuscript adequately addresses the reviewer’s concerns and is now clearer, more rigorous, and more useful to readers.

 

With sincere thanks and respect,

 

The Authors

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

ok