A fibrin targeted molecular imaging evaluation of microvascular no‐reflow in acute ischemic stroke

Abstract Objective To investigate the relationship between fibrin deposition and “no‐reflow” within microcirculation after thrombolysis in acute ischemic stroke (AIS). Materials and methods Experiments were approved by the institutional animal care and use committee. An experimental AIS model was induced in C57BL/6 mice by middle cerebral artery occlusion (MCAO) via the photothrombotic method. Mice were randomly assigned to non‐thrombolytic or thrombolytic treated groups (n = 12 per group). The modified Neurological Severity Score and Fast Beam Balance Test were performed by a researcher blinded to the treatment method. MRI was utilized to evaluate all of the mice. An FXIIIa‐targeted probe was applied to detect fibrin deposition in acute ischemic brain regions by fluorescence imaging. Necrosis and pathological changes of brain tissue were estimated via Hematoxylin and eosin staining while fibrin deposition was observed by immunohistochemistry. Results Thrombolytic therapy improved AIS clinical symptoms. The infarct area of non‐thrombolytic treated mice was significantly greater than that of the thrombolytic treated mice (p < .0001). Fluorescent imaging indicated fibrin deposition in ischemic brain tissue in both groups, with less fibrin in non‐thrombolytic treated mice than thrombolytic treated mice, though the difference was not significant. Brain cells with abnormal morphology, necrosis, and liquefication were observed in the infarcted area for both groups. Clotted red blood cells (RBCs) and fibrin build‐up in capillaries were found near the ischemic area in both non‐thrombolytic and thrombolytic treated groups of mice. Conclusion Fibrin deposition and stacked RBCs contribute to microcirculation no‐reflow in AIS after thrombolytic therapy.


INTRODUCTION
Acute ischemic stroke (AIS) is one of the most common causes of death and disability worldwide. (Virani et al., 2020) Thrombolytic therapy is widely used to recanalize occluded cerebral blood vessels to save ischemic brain tissue thereby reducing the mortality rate. (Lo, 2008;Lo et al., 2003) Alteplase, which degrades thrombus fibrin, is the only FDAapproved pharmacotherapy for AIS (Liu et al., 2017;Rha & Saver, 2007).
Recanalization of fibrin occluded vessels re-perfuses the ischemic brain region with blood and oxygen, thus promoting stroke rehabilitation (Huang et al., 2015;Panni et al., 2019). However, accumulating evidence indicates that vessel recanalization from thrombolysis does not always lead to brain reperfusion, possibly attributable to the failure of microvascular reperfusion, a phenomenon termed "no-reflow" in AIS (Dalkara & Arsava, 2012;El Amki et al., 2020;Lee et al., 2016). The underlying mechanism of AIS "no-reflow" has not yet been fully understood.
Thrombus location, size, and composition are important factors affecting thrombolytic treatment outcome (Friedrich et al., 2015;Liebeskind et al., 2011;Morales-Vidal & Biller, 2013;Riedel et al., 2011;Rohan et al., 2014). Thrombi causing AIS mainly consist of activated platelets, red blood cells (RBCs), and fibrin (Duffy et al., 2017;Niesten et al., 2014), classifiable into fibrin-rich and RBC-rich clots (Jolugbo & Ariëns, 2021). Fibrin forms the basic framework of thrombus and interacts differently with RBCs and platelets. Understanding the relationship between fibrin deposition and "no-reflow" within microcirculation after thrombolysis can inform AIS thrombolytic therapy, and also help develop new treatments targeting microvascular "no-reflow." We hypothesize that fibrin deposition in capillaries is an important mediator of "no-reflow" leading to insufficient brain recovery.

Stroke model creation and treatment
All animal studies were approved by the Institutional Animal Care and

Severity scoring and fast beam balance test
The modified Neurological Severity Score (mNSS) (0-12) was recorded at days 0, 1, 3, and 7 for sham induction, non-thrombolytic, and thrombolytic treated groups of mice by a researcher unaware of the treatment method. Higher scores indicate more severe symptoms (normal score 0, maximal deficit score 12; Table 1 in supplemental information; Chen et al., 2016). Weights were recorded daily by a blind researcher. Fast Beam Balance Test was performed on Day 1 for sham induction, non-thrombolytic, and thrombolytic treated mice groups.
Prior to the experiment, mice were trained twice a day for 1 week and then deprived of food and water for 24 hr. Mice were placed at one end of the balance beam and encouraged to cross to the other side to obtain food and water, with time and speed recorded and compared between groups.

MRI and postprocessing
Cerebral 3D cerebrovascular images from 3D FLASH-MRA were automatically generated and reconstructed on the platform. Image J software (NIH) was employed to measure the infarction size on T2W images for the different groups of mice.

A15 synthesis and NIRF imaging
The FXIIIa-targeted probe, A15, was synthesized according to the previous study with slight modification (Tung et al., 2003). Briefly, Ac-GN 13 QEQVSPLTLLK 24 WC, an FXIIIa affinity peptide based on α-2-antiplasmin, was compounded by solid-phase peptide synthesis (purchased from Fubaike Co., Beijing), and then marked with Cy7 (Molecular Probes, Fluorescence, China) via its cysteine side chain (A15-Cy7). Fluorescent images were obtained with the IVIS Spectrum In Vivo Imaging System (PerkinElmer, USA) at 24 hr post-injection of A15-Cy7 (0.5 mg/kg), which was administered immediately after MCAO. Two filter sets (excitation: 743 nm; emission: 767 nm) were used to detect probe-specific fluorescence (from A15-Cy7) and auto-fluorescence (primarily from the skin and other organs). Fluorescent images were then analyzed based on their spectral patterns using Living Imaging Software (PerkinElmer, USA).
Results were analyzed by a radiologist who was blinded to the specific treatment method. For analysis of fibrin deposition, regions of interest (ROIs) were selected manually on both hemispheres of the NIRF images. Target-to-background ratios (TBRs), defined as Intensity ROI of ipsilateral ischemic brain /Intensity ROI of contralateral normal brain , were used to quantify fibrin deposition (Klohs et al., 2008).

Histopathological analysis
After MR and NIRF imaging on Day 1, mice brains were harvested after being perfused with 20 ml ice-cold PBS transcardially, and cut into 4 μm-thick sections. The degree of necrosis and pathological changes of brain tissue was estimated via hematoxylin and eosin (H&E) staining. Fibrin deposition was observed by immunohistochemistry.

Statistical analysis
In order to avoid the influence of subjective factors, a completely randomized design was adopted, with all measurements performed blindly with regard to the treatment method. One-way ANOVA test was used to compare the results of the Fast Beam Balance Test, mNSS scoring, infarct size on T2W images, and TBRs from NIRF images between the non-thrombolytic and thrombolytic treated groups. Log-rank (Mantel-Cox) test was used to compare the survival curve differences between each group. Results were reported as mean ± standard deviation (SD).
GraphPad Prism (v.6, GraphPad Software) was used for statistical analysis, while p < .05 was considered statistically significant. The mNSS provides a reliable assessment for clinical severity, highly correlated with infarct volume in AIS (Bieber et al., 2019). AIS mice treated with thrombolytic therapy showed significantly improved clinical symptoms compared with non-thrombolytic treated mice on days 1, 3, and 7 (p < .0001 for mNSS on all test days, Figure 1c). After thrombolysis, AIS mice still showed severe clinical symptoms, which gradually recovered as represented by lower mNSS scores compared with sham induction mice (p < .0001 on day 1, p = .0002 on day 3, p = .0455 on day 7, Figure 1c). AIS mice with thrombolytic therapy showed significantly faster speeds than mice without thrombolytic therapy on the Fast Beam Balance Test on day 1 (p < .0001, Figure 1d). These findings reveal that thrombolytic therapy with rt-PA improves clinical symptoms and survival rate for photothrombotic AIS mice.

F I G U R E 1
Thrombolytic therapy improves AIS clinical symptoms. (a) Daily weight of mice in sham group (n = 6), thrombolytic treated group (n = 12), and non-thrombolytic treated group (n = 12). (b) Survival plot of sham induction mice, thrombolytic treated mice, and non-thrombolytic treated mice with survival rates of 100%, 67%, and 25%, respectively (p = .044 between thrombolytic and non-thrombolytic treated mice, p = .1207 between sham induction and thrombolytic treated mice). (c) The mNSS scores of the sham group, thrombolytic treated group, and non-thrombolytic treated group of mice on days 0, 1, 3, and 7 (thrombolytic vs non-thrombolytic treated groups: p < .0001 on all test days; sham vs. thrombolytic treated groups: p < .0001 on day 1, p = .0002 on day 3, p = .0455 on day 7). (D) Result of fast beam balance test on day 1 (p < .0001) images after 24 hr of modeling showed that the infarct area of AIS mice with thrombolytic therapy (7.40 ± 1.46 mm 2 ) was significantly smaller than that in mice without thrombolysis (23.39 ± 1.58 mm 2 ; p < .0001, Figure 2h-j). No abnormal signal was found in the brain of sham induction mice (Figure 2g). These results indicate that thrombolytic therapy can effectively recanalize the occluded MCA thus greatly reducing the infarction area. However, thrombolysis even at 1 hr after stroke occurrence cannot completely avoid infarction despite successful rapid recanalization.

NIRF imaging uncovers fibrin deposition after thrombolysis
In order to find out what hindered brain recovery after thrombolysis, an FXIIIa-targeted probe (A15-Cy7) to track fibrin deposition in AIS via NIRF imaging, as previously reported (Jaffer et al., 2004;Kim et al., 2016;Tung et al., 2003) The sham induction mice showed low fluorescent intensity in bilateral brain regions both in vivo (Figure 3a) and ex vivo (Figure 3d). For mice in the nonthrombolytic treated group (Figure 3b,e) and thrombolytic treated group (Figure 3c,f), high fluorescence intensity was displayed in the affected side with relatively low intensity on the contralateral side. Significant fibrin deposition was found in AIS mice both with and without rt-PA thrombolysis compared with sham induction mice (p = .013 for thrombolytic treated mice, p = .0029 for nonthrombolytic treated mice). No significant difference in TBRs was found between the non-thrombolytic treated group and the thrombolytic treated group of mice (p = .644) though TBR in the thrombolytic treated group was slightly lower (Figure 3g), indicating fibrin F I G U R E 2 Thrombolytic therapy recanalizes occluded vessels and reduces the infarct area. 3D FLASH-MRA images of sham induction mice (a), AIS mice in non-thrombolytic treated (b), and thrombolytic treated (c) groups, with red arrows indicating contralateral MCA and blue arrows indicating ipsilateral MCA. Patency rate of the right MCA in sham induction group (d), non-thrombolytic treated group (e), and thrombolytic treated group (f). Brain T2W MR images of sham induction mice (g), AIS mice in non-thrombolytic treated (h), and thrombolytic treated (i) groups. (j) Infarct lesion area was measured (n = 12) revealing that thrombolytic treated mice experienced significantly smaller lesion areas than non-thrombolytic treated mice (p < .0001) deposition in ipsilateral brain tissue even after successful thrombolytic therapy.

Histopathological analysis reveals RBC aggregation and fibrin deposition in capillaries
After imaging, brains were harvested for histopathological analysis.
Contralateral unaffected brain regions maintained normal morphology (Figure 4a,d). The border between the ischemic and non-ischemic areas was clearly delineated with H&E staining and immunohistochemical staining in both non-thrombolytic (Figures 4b and 5b) and thrombolytic treated mice (Figures 4c and 5c). Non-thrombolytic treated mice manifested an obviously larger infarction than thrombolytic treated mice.
Brain cell liquefactive necrosis was found in the infarct area of nonthrombolytic treated mice (Figure 4b,e). An accumulation of RBCs stacked in micrangiums and capillaries was detected at the border of the ischemic area (Figure 4e,f, as indicated by arrows). Immunohistochemistry staining of fibrin was undertaken to detect fibrin deposition in the ischemic brain area. Unimpacted contralateral brain F I G U R E 3 NIRF imaging uncovers fibrin deposition after thrombolysis. In vivo (a-c) and in vitro (d-f) NIRF images of sham induction mice (a, d), AIS mice in non-thrombolytic treated (b, e), and thrombolytic treated (c, f) groups. (g) Significant differences in fibrin deposition were found in thrombolytic treated (p = .013) and non-thrombolytic treated (p = .0029) mice compared with sham induction mice. No significant difference in TBRs was found between non-thrombolytic treated group and thrombolytic treated group of mice (p = .644) tissue showed sporadic fibrin staining (Figure 5a,d). Patchy fibrin was found in the infarction area in both groups (Figure 5b,c,e,f), which accords with NIRF imaging results. Fibrin deposits occupied a significantly larger area in non-thrombolytic treated mice (Figure 5b) compared with thrombolytic treated mice (Figure 5c). Interestingly, when comparing the H&E staining ( Figure 4c) and fibrin staining of adjacent sections (Figure 5c) in thrombolytic treated mice, the area of fibrin deposition was larger than the area of necrosis, indicating that fibrin deposited not only within the infarction area but also in the penumbral zone. Clotty fibrin deposition was found in micrangium and capillary walls in both thrombolytic and non-thrombolytic mice (Figure 5e,f, indicated by arrows). Therefore, we conclude that both RBC accumulation and fibrin deposition lead to "no-reflow" of microcirculation after thrombolytic therapy.

DISCUSSION
Thrombolytic treatment of AIS with rt-PA within 4.5 hr after symptom onset has shown beneficial effects for patients, the efficacy of which is particularly time-dependent, with earlier treatment generating better stroke outcome (Emberson et al., 2014). However, F I G U R E 4 H&E staining of contralateral normal brain (a and d), non-thrombolytic treated mice (b and e) and thrombolytic treated mice (c and f) with different magnification (a-c 10×, scale bar = 200 μm; d-f 40×, scale bar = 50 μm). Brain cells maintained normal contralateral morphology in normal unimpacted brain regions (a and d). Necrotic tissue area in non-thrombolytic treated mice (b) was substantially larger than in thrombolytic treated mice (c). Demarcation between necrotic tissue and normal tissue can be seen in both groups (b and c). (f) Accumulated RBCs in micrangiums and capillaries were found in affected brain regions in thrombolytic treated mice F I G U R E 5 Immunohistochemistry staining of contralateral normal brain (a and d), non-thrombolytic treated mice (b and e) and thrombolytic treated mice (c and f) with different magnification (a-c 10×, scale bar = 200 μm; d-f 40×, scale bar = 50 μm). Limited positive staining of fibrin was found in contralateral normal brain (a and d). Fibrin deposited area in non-thrombolytic treated mice (B) was obviously larger than that in thrombolytic treated mice (c). Demarcation between necrotic tissue and normal tissue can be seen in both groups (b and c). (f) Clotty fibrin deposition was observed in micrangiums and capillaries of mice with thrombolytic therapy no-reflow of microcirculation hinders the full recovery of affected brain tissue even with prompt and successful recanalization of occluded vessels (Dalkara & Arsava, 2012). Our study ascertained that despite reopening an obstructed artery as early as 1 hr after inducing a stroke in mice, significant infarction still occurred in the affected brain region. Treatments targeting microvascular "no-reflow" may therefore result in increased recovery of at-risk tissue. Duffy et al. reported that recanalization of microcirculation after thrombolytic therapy is one of the key factors determining the functional and morphological recovery of ischemic brain tissue (Duffy et al., 2017).
Recovery of microcirculation blood flow should achieve full perfusion of blood in the ischemic hypoxia zone, and restore neurocyte functionality, while improving the effect and prognosis of thrombolytic treatment (Lee et al., 2016 Kim et al., 2015;Spuentrup et al., 2005;Tung et al., 2003).
This study utilized a fibrin-targeted NIRF probe to reveal fibrin accumulation after reopening the occluded artery via in vivo NIRF imaging.
A histopathological analysis confirmed that deposited fibrin, as well as accumulated RBCs entangled by fibrin in micrangiums and capillaries around the infarct area, leads to insufficient reperfusion of microcirculation, a condition known as "no-reflow." Therefore, the in vivo noninvasive visualization of deposited fibrin after thrombolysis in AIS with rt-PA can reflect the degree of reperfusion within the microcirculation and the severity of no-reflow, providing critical guidance for further clinical treatment and assessment.

LIMITATIONS
Histopathological findings in this study determined that the no-reflow phenomenon is caused by fibrin deposits and RBC accumulation in microvessels. Fibrin-targeted molecular imaging should be capable of evaluating microvascular no-reflow in acute ischemic stroke; however, there is no direct in vivo evidence supporting this. Future studies can investigate microcirculation blood flow and fibrin deposit quantification.

CONCLUSION
Molecular imaging of fibrin deposition after recanalization of occluded vessels in AIS can reflect the degree of microcirculation reperfusion while indicating the no-reflow severity. Post-thrombolysis in vivo visualization of deposited fibrin offers a potential non-invasive method to evaluate microcirculation patency thereby providing prompt guidance for next stage clinical treatment.

ACKNOWLEDGMENTS
This work was supported by the Shanghai Pujiang Program (Grant No. 20PJ1402200) and the National Natural Science Foundation of China (Grant No. 81771242).

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.

PEER REVIEW
The peer review history for this article is available at https://publons. com/publon/10.1002/brb3.2474