Involvement of Secondary Induced Thrombus on Hemorrhage Induced by Both Delayed Recanalization and Delayed t-PA Treatment in Murine Ischemic Stroke Models

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This study presents a well-designed and mechanistically insightful investigation into delayed reperfusion-associated hemorrhagic complications. The experimental strategy is comprehensive, the data are robust, and the findings are clearly presented. I particularly appreciated the way the authors integrated vascular pathology, secondary induced thrombus formation, and proteolytic activity to address a clinically relevant problem. Overall, the manuscript is scientifically sound, well written, and has clear publication potential; however, the following points should be addressed before the manuscript can proceed to the publication stage.

• The Introduction section would benefit from modest expansion to better frame the clinical and epidemiological relevance of the study.
• The statement “It accounts for about 60% of cerebrovascular disorders and affects approximately 12 million people yearly” should be supported by an appropriate and up-to-date reference.
• The manuscript does not specify how the number of animals used in the study was determined. The authors should clarify whether an a priori power analysis was performed.
• Given that animal loss is relatively common in MCAO models, the authors should transparently report whether any animal mortality or exclusion occurred during the experiments.
• Considering the large number of experimental groups and animals used, inclusion of an experimental setup schematic is strongly recommended to improve clarity and reproducibility.
• All brain micrographs should include scale bars.
• For TTC staining, the anatomical positions of the coronal brain sections should be clearly indicated.
• The Discussion section is well written and effective; however, adding a brief limitations subsection at the end of this section would further strengthen the manuscript.

Author Response

This study presents a well-designed and mechanistically insightful investigation into delayed reperfusion-associated hemorrhagic complications. The experimental strategy is comprehensive, the data are robust, and the findings are clearly presented. I particularly appreciated the way the authors integrated vascular pathology, secondary induced thrombus formation, and proteolytic activity to address a clinically relevant problem. Overall, the manuscript is scientifically sound, well written, and has clear publication potential; however, the following points should be addressed before the manuscript can proceed to the publication stage.

Comment 1: The Introduction section would benefit from modest expansion to better frame the clinical and epidemiological relevance of the study.

Response 1: According to this comment, we added clinical and epidemiological relevance in Introduction, line 49-50.

 

Comment 2: The statement “It accounts for about 60% of cerebrovascular disorders and affects approximately 12 million people yearly” should be supported by an appropriate and up-to-date reference.

Response 2: According to this comment, an adequate reference is now added in Introduction, in line 39 (b).

 

Comment 3: The manuscript does not specify how the number of animals used in the study was determined. The authors should clarify whether a priori power analysis was performed.

Response 2: The numbers of animals were decided from previous experiments performed prior power analysis. It was now described in Materials and Methods, line 72-77 (c).

 

Comment 4: Given that animal loss is relatively common in MCAO models, the authors should transparently report whether any animal mortality or exclusion occurred during the experiments.

Response 4 In MCA ligation and ischemia/reperfusion models, animal loss is rare because of smaller ischemic damage area compared to the intracranial suture model. Practically no animal was lost in the current study. It is now described in Materials and Methods, line 78-79.

 

Comment 5: Considering the large number of experimental groups and animals used, inclusion of an experimental setup schematic is strongly recommended to improve clarity and reproducibility.

Response 5: We are now including the experimental setup in Supplemental Figure 1, which is described in line 81.

 

Comment 6: All brain micrographs should include scale bars.

Response 6: It is now added in all micrographs.

 

Comment 7: For TTC staining, the anatomical positions of the coronal brain sections should be clearly indicated.

Response 7: The method of TTC staining including anatomical position of each section is now described in Materials and Methods, line 128-130. According to the revision, a reference [15] was added.

 

 

Commnet 8: The Discussion section is well written and effective; however, adding a brief limitations subsection at the end of this section would further strengthen the manuscript.

Response 8: The major limitations in this manuscript are that involvement of GCX on SIT formation is unclear and that only ischemic core event is observed in BALB/c strain. These points were individually described in Discussion. Thus, we do not add the subsection.

Reviewer 2 Report

Comments and Suggestions for Authors

This article explores the mechanisms by which delayed recanalization and delayed t-PA treatment induce hemorrhage in a mouse model of ischemic stroke, with a focus on the role of secondary thrombosis induction (SIT). By establishing permanent middle cerebral artery occlusion (MCA-O) and modified ischemia/reperfusion (I/R) models (with 2-hour and 4.5-hour ischemia followed by recanalization), the effects of SIT formation, hemorrhage conditions, MMP activity, and related intervention measures (t-PA, heparin) were evaluated to reveal the mechanism by which SIT leads to hemorrhage through the degradation of glycocalyx and activation of fibrinolysis/MMP-9, providing a basis for the prevention and control of bleeding risks in delayed recanalization treatment.

The shortcomings of the study: The BALB/c mice were used, which have limited collateral circulation and sparse ischemic penumbra. The results may only reflect the mechanism of hemorrhage in the ischemic core, and there are differences from the SIT conditions in the penumbra in humans or other mouse models; The specific molecular mechanism of glycolipid degradation and the cell source of SIT formation were not deeply explored.

To investigate the effects of different ischemic durations (such as 3 hours, 5 hours) on the formation of SIT and bleeding, and to determine the time threshold; to study the key enzymes for glycoconjugate degradation (such as hyaluronidase, heparinase) and their regulatory mechanisms; to verify the relationship between SIT and subendplate injury in different mouse strains (such as C57BL/6); to develop new anti-thrombotic drugs targeting SIT formation, and to balance the risks of anticoagulation and bleeding.

Author Response

This article explores the mechanisms by which delayed recanalization and delayed t-PA treatment induce hemorrhage in a mouse model of ischemic stroke, with a focus on the role of secondary thrombosis induction (SIT). By establishing permanent middle cerebral artery occlusion (MCA-O) and modified ischemia/reperfusion (I/R) models (with 2-hour and 4.5-hour ischemia followed by recanalization), the effects of SIT formation, hemorrhage conditions, MMP activity, and related intervention measures (t-PA, heparin) were evaluated to reveal the mechanism by which SIT leads to hemorrhage through the degradation of glycocalyx and activation of fibrinolysis/MMP-9, providing a basis for the prevention and control of bleeding risks in delayed recanalization treatment.

Comment 1: The shortcomings of the study: The BALB/c mice were used, which have limited collateral circulation and sparse ischemic penumbra. The results may only reflect the mechanism of hemorrhage in the ischemic core, and there are differences from the SIT conditions in the penumbra in humans or other mouse models;

Response 1: As is pointed out, SIT observed in this study is thought to be a reaction associated with delayed reperfusion in the ischemic core. However, since an ischemic core also forms in human patients, it is possible that a similar reaction is induced and contributes to bleeding. However, it is possible that a different type of SIT may be formed. This possibility is already described in the Discussion, line 377-380.

 

Comment 2: The specific molecular mechanism of glycolipid degradation and the cell source of SIT formation were not deeply explored.

Response 2: In this manuscript, we presented a decrease in glycocalyx as one possible cause of SIT formation.  Another possibility is now described in Discussion in line 367-368. Further investigation is needed to proof the relationship between decrease in glycocalyx and SIT formation as already mentioned Discussion in line 367-368.

 

Comment 3: To investigate the effects of different ischemic durations (such as 3 hours, 5 hours) on the formation of SIT and bleeding, and to determine the time threshold;

Response 3: This study revealed that delayed recanalization leads to the formation of SIT, which contributes to bleeding. Therefore, we chose 4.5h as a sufficient delay based on the clinical time window for t-PA.

 

Comment 4: To study the key enzymes for glycoconjugate degradation (such as hyaluronidase, heparinase) and their regulatory mechanisms;

Response 4: As mentioned above, glycocalyx degradation is one of the possible mechanisms for SIT formation. Therefore, further investigation is required, including the mechanism of GCX degradation, as already mentioned in the Discussion in line 367-368.

 

Comment 5: To verify the relationship between SIT and subendplate injury in different mouse strains (such as C57BL/6);

Response 5: We do not understand the reason to verify the relationship between SIT in ischemic stroke and subendplate injury. There is no histological or fundamental similarity between ischemic stroke and subendoplate injury.

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript addresses an important and clinically relevant question: why hemorrhage risk rises when reperfusion (and/or t-PA) is delayed after ischemic stroke, and whether a “secondary induced thrombus” (SIT) contributes. The overall concept is interesting, and the attempt to compare permanent occlusion vs. reperfusion at 2 h vs. 4.5 h in a distal MCA ligation model could be useful for dissecting time-dependent vascular pathology. However, the current version overstates novelty (“novel MCA-I/R model”) and mechanistic certainty relative to what is actually demonstrated. The work reads largely as descriptive imaging with pharmacologic perturbations, but key experimental rigor elements are missing or unclear (randomization, blinding, inclusion/exclusion criteria, a priori sample size justification, and how the very large total N reported relates to each endpoint/figure). Because your main conclusions hinge on subtle differences in microvascular thrombosis and petechial hemorrhage, these omissions materially limit confidence and must be fixed to support publication.

A major methodological issue is the definition/verification of “SIT.” The study uses IV fibrinogen-FITC and interprets “vessel-like” FITC structures as thrombi, yet the paper itself acknowledges increased vascular permeability with dextran leakage in multiple models; this creates a real risk that at least part of the signal reflects extravasated tracer or non-occlusive deposition rather than bona fide intraluminal thrombus. You need stronger evidence that these structures are intravascular and thrombotic (e.g., endothelial marker co-localization, z-stack/confocal confirmation of luminal localization, staining for fibrin/platelets, and ideally an orthogonal thrombus measure such as histologic fibrin/platelet quantification). The glycocalyx interpretation has a similar inferential gap: tomato lectin negativity is treated as “glycocalyx loss,” but reduced perfusion/microvascular no-reflow in the 4.5 h group could also simply prevent lectin delivery/labeling. Without controls separating delivery failure from true endothelial surface loss (additional glycocalyx markers, perfusion normalization controls, or intravascular tracer distribution checks), the proposed sequence “glycocalyx loss → SIT formation” remains speculative. The statistical approach also needs upgrading: repeated blood-flow measures are analyzed as if independent rather than longitudinal, multiple group comparisons/timepoints are extensive without clear multiplicity control, and results are mostly presented as mean±SD without effect sizes or confidence intervals—this makes it hard to judge robustness.

Finally, the pharmacology-based causal claims require tightening and, in places, are internally inconsistent. t-PA is said to accelerate SIT degradation, yet delayed t-PA is also argued to drive hemorrhage via a t-PA/plasmin/MMP pathway “initiated on SIT”; if SIT is reduced by t-PA, the manuscript needs a clearer reconciliation of why hemorrhage still increases (and whether this depends on timing, dose, or residual microthrombi). The heparin conclusion—suggesting pharmacologic inhibition of SIT formation as a preventive strategy for hemorrhage in delayed EVT/t-PA—also needs much more careful framing because systemic anticoagulation has obvious translational constraints; at minimum, readers need coagulation context, dosing justification, and discussion of how this maps (or does not map) to clinical practice. The MMP story is currently based largely on co-localization images (MMPsense signal; MMP-9 immunoreactivity) without quantitative linkage to hemorrhage magnitude across conditions or direct pathway validation (e.g., MMP inhibition/knockout, or quantitative MMP activity changes with heparin/t-PA/aprotinin). The manuscript would benefit from a major rewrite to reduce overreach, add the missing rigor details, strengthen SIT/glycocalyx validation, and present analyses in a way that supports (rather than asserts) causality.

 

Author Response

The manuscript addresses an important and clinically relevant question: why hemorrhage risk rises when reperfusion (and/or t-PA) is delayed after ischemic stroke, and whether a “secondary induced thrombus” (SIT) contributes. The overall concept is interesting, and the attempt to compare permanent occlusion vs. reperfusion at 2 h vs. 4.5 h in a distal MCA ligation model could be useful for dissecting time-dependent vascular pathology. Comment 1: However, the current version overstates novelty (“novel MCA-I/R model”) and mechanistic certainty relative to what is actually demonstrated.

Reponse 1: This manuscript is submitted to Special issue: Animal Models for the Study of Cardiovascular
> Physiology—Second Edition. According to the subjection of this issue, we explain the characteristics of the novel model.

 

Comment 2: The work reads largely as descriptive imaging with pharmacologic perturbations, but key experimental rigor elements are missing or unclear (randomization, blinding, inclusion/exclusion criteria, a priori sample size justification, and how the very large total N reported relates to each endpoint/figure). Because your main conclusions hinge on subtle differences in microvascular thrombosis and petechial hemorrhage, these omissions materially limit confidence and must be fixed to support publication.

Respone 2: The key experimental rigor elements were now described in Materials and Methods. line 72-77 with a additional reference [14]. The total N given in Materials and Methods is the total for all experiments, and the animals used in each experiment are described in figure legends.

 

Comment 3: A major methodological issue is the definition/verification of “SIT.” The study uses IV fibrinogen-FITC and interprets “vessel-like” FITC structures as thrombi, yet the paper itself acknowledges increased vascular permeability with dextran leakage in multiple models; this creates a real risk that at least part of the signal reflects extravasated tracer or non-occlusive deposition rather than bona fide intraluminal thrombus. You need stronger evidence that these structures are intravascular and thrombotic (e.g., endothelial marker co-localization, z-stack/confocal confirmation of luminal localization, staining for fibrin/platelets, and ideally an orthogonal thrombus measure such as histologic fibrin/platelet quantification).

Response 3: According to the comments, we add additional data of immunohistochemistry of CD31, an endothelial marker with Fbg-FITC in Figure 2O-R. As is shown in this figure, SIT was distributed in vessels and more bright than leaked Fbg-FITC. Therefore, it is easily identified from leaked Fbg-FITC. Relating, it was now described in Materials and Methods in line140-147, Results inline 176-177 and figure legend in line 190-194.

 

Comment 4: The glycocalyx interpretation has a similar inferential gap: tomato lectin negativity is treated as “glycocalyx loss,” but reduced perfusion/microvascular no-reflow in the 4.5 h group could also simply prevent lectin delivery/labeling. Without controls separating delivery failure from true endothelial surface loss (additional glycocalyx markers, perfusion normalization controls, or intravascular tracer distribution checks), the proposed sequence “glycocalyx loss → SIT formation” remains speculative.

Response 4:  We agree that TL signal was lost by microvascular no-reflow. This possibility is now described in the Discussion in line 366-367. According to the revision, a reference was added [33].

 

Comment 5: The statistical approach also needs upgrading: repeated blood-flow measures are analyzed as if independent rather than longitudinal, multiple group comparisons/timepoints are extensive without clear multiplicity control, and results are mostly presented as mean±SD without effect sizes or confidence intervals—this makes it hard to judge robustness.

Response 5: In the current study, blood flow was indicated as percentage of flow before ligation. So, we think it is possible to compare between groups without multiplicity control.

 

Comment 6: Finally, the pharmacology-based causal claims require tightening and, in places, are internally inconsistent. t-PA is said to accelerate SIT degradation, yet delayed t-PA is also argued to drive hemorrhage via a t-PA/plasmin/MMP pathway “initiated on SIT”; if SIT is reduced by t-PA, the manuscript needs a clearer reconciliation of why hemorrhage still increases (and whether this depends on timing, dose, or residual microthrombi).

Response: We think that in delayed t-PA administration, t-PA dissolves the thrombus of primary cause, and this induces recanalization, which is thought to lead to the formation of SIT. Then, hemorrhage is induced by activation of the t-PA/plasmin/MMP pathway on the SIT. This discussion is added to the Discussion in line 321-324.

 

Comment 7: The heparin conclusion—suggesting pharmacologic inhibition of SIT formation as a preventive strategy for hemorrhage in delayed EVT/t-PA—also needs much more careful framing because systemic anticoagulation has obvious translational constraints; at minimum, readers need coagulation context, dosing justification, and discussion of how this maps (or does not map) to clinical practice.

Response 7: In this study, heparin is used as one of the anticoagulant agents for preventing SIT formation. Indeed other anticoagulants may be beneficial for suppression of SIT formation and subsequent hemorrhage. This consideration has already described in the Discussion in line 327-330,

 

Comment 8: The MMP story is currently based largely on co-localization images (MMPsense signal; MMP-9 immunoreactivity) without quantitative linkage to hemorrhage magnitude across conditions or direct pathway validation (e.g., MMP inhibition/knockout, or quantitative MMP activity changes with heparin/t-PA/aprotinin).

Response 8: We agree that it is better to describe the effect of MMP inhibition on hemorrhage. However, contribution of MMP on hemorrhage is widely observed in various animal models including both MCA thrombotic occlusion model and intraluminal thread model. Therefore, it is not essential to describe the effect in this manuscript.

 

Comment 9: The manuscript would benefit from a major rewrite to reduce overreach, add the missing rigor details, strengthen SIT/glycocalyx validation, and present analyses in a way that supports (rather than asserts) causality.

Response 9: This manuscript has been revised as stated above.

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

the authors improved the paper.