PCSK9 Inhibitor Attenuates Platelet TLR4 in ACS and TLR4 Blockers Prevent PCSK9-Mediated Platelet Activation

Abstract

Background

Proprotein convertase subtilisin/kexin type 9 (PCSK9) regulates platelet activation through multiple signaling pathways. This study aimed to determine whether PCSK9 modulates platelet activation via Toll-like receptor 4 (TLR4).

Methods

A total of 120 patients with acute coronary syndrome (ACS) were divided into a control group and an experimental group receiving subcutaneous evolocumab. Platelet surface TLR4 was quantified by flow cytometry. Platelet-rich plasma from healthy volunteers (n = 5) was incubated with normal saline, PCSK9, PCSK9 + evolocumab, or PCSK9 + the TLR4 inhibitor resatorvid (TAK-242). Platelets were then stimulated with adenosine diphosphate, and aggregation was measured. Plasma thromboxane A2 (TXA2) was measured by ELISA. Megakaryoblastic leukemia cells (MEG-01) were co-incubated with PCSK9 and stained for TLR4 and PCSK9 by immunofluorescence.

Results

PCSK9 significantly enhanced platelet aggregation and TXA2 release in vitro; both effects were abolished by evolocumab and TAK-242. In ACS patients, evolocumab treatment reduced platelet surface TLR4 expression and decreased plasma TXA2 levels. Dual immunofluorescence staining demonstrated co-localization of TLR4 and PCSK9 in MEG-01 cells.

Conclusions

This study provides the first evidence that PCSK9 inhibitors downregulate platelet TLR4 expression in ACS patients. The pro-aggregatory effects of PCSK9 on platelets were counteracted by TLR4 inhibition, indicating a mechanistic link between PCSK9 and TLR4 signaling in platelet activation. Co-localization of TLR4 and PCSK9 in MEG-01 cells supports a potential interaction between the two proteins.

Abbreviations

ACS, Acute coronary syndrome

ADP, Adenosine diphosphate

CD36, CD36 molecule/scavenger receptor class B member 3/thrombospondin receptor

CD41, Cluster of differentiation antigen 41

DAPI, 4′,6-diamidino-2-phenylindole

ELISA, Enzyme-linked immunosorbent assay

HS, Healthy subjects

LDL-C, Low-density lipoprotein cholesterol

LOX-1, Lectin-like oxidized low-density lipoprotein receptor-1

MEG-01, Megakaryoblastic leukemia cell line

MyD88, Myeloid differentiation primary response 88

NF-κB, Nuclear factor kappa-B

NLRP3, NACHT, LRR, and PYD domains-containing protein 3

OD, Optical density

PCSK9, Proprotein convertase subtilisin/kexin type 9

PRP, Platelet-rich plasma

RT, Room temperature

SEM, Standard error of the mean

TAK-242, Resatorvid

TLR4, Toll-like receptor 4

TXA₂, Thromboxane A₂

INTRODUCTION

Atherosclerotic cardiovascular disease remains one of the most prevalent health issues worldwide, characterized by high morbidity and mortality. Despite substantial advances in mechanical and pharmacological reperfusion strategies to alleviate acute myocardial infarction, the incidence of acute coronary syndrome (ACS) resulting from coronary artery stenosis or complete occlusion continues to pose a major global health challenge. The formation, rupture, and erosion of atherosclerotic plaques are the fundamental pathological basis of myocardial infarction, with thrombosis playing a pivotal role in the progression of ACS. Platelet activation is a key trigger in these processes. However, the factors influencing platelet activation in ACS patients are not yet fully understood. Therefore, elucidating the potential mechanisms underlying platelet activation in ACS may provide novel insights into the prevention and management of this condition.

Proprotein convertase subtilisin/kexin type 9 (PCSK9), a serine protease first identified in 2003,¹ is a key regulator of cholesterol homeostasis through its inhibitory effect on the low-density lipoprotein receptor pathway. PCSK9 is primarily synthesized and secreted by the liver.^2,3 Inhibition of PCSK9 has been shown to reduce circulating low-density lipoprotein cholesterol (LDL-C) levels and significantly decrease the risk of cardiovascular events.⁴ Beyond its lipid-regulatory function, PCSK9 also contributes to the progression of cardiovascular disease through multiple mechanisms.⁵ Although its pleiotropic effects are still being elucidated, accumulating evidence indicates that PCSK9 can modulate platelet activation through several pathways, potentially independent of LDL-C metabolism.

The role of PCSK9 in promoting platelet activation and thrombosis through its binding to platelet CD36 molecule/scavenger receptor class B member 3/thrombospondin receptor (CD36) and subsequent activation of downstream CD36 signaling pathways has been well established.⁶ However, beyond this well-characterized CD36-mediated mechanism, it remains unclear whether additional pathways contribute to PCSK9-induced platelet activation, either independently or through modulation of the CD36 pathway. Previous studies have shown that Toll-like receptor 4 (TLR4)-myeloid differentiation primary response 88 (MyD88)-nuclear factor kappa-B (NF-κB) signaling plays a critical role in regulating PCSK9 expression.⁷ Likewise, elevated PCSK9 levels in sepsis have been reported to activate the TLR4/MyD88/NF-κB and NACHT, LRR, and PYD domains-containing protein 3 (NLRP3) pathways, thereby inducing inflammation and promoting vascular endothelial dysfunction.⁸ Considering that platelet TLR4 can mediate activation and aggregation through multiple mechanisms,^9-11 we hypothesized that a potential PCSK9-TLR4 axis may be involved in PCSK9-mediated platelet activation.

Accordingly, the present study was designed to investigate auxiliary pathways contributing to PCSK9-induced platelet activation beyond the established CD36 mechanism, with a particular focus on elucidating the role of the interaction between PCSK9 and TLR4 in this process.

METHODS

Human study

This study involved human participants, was conducted in full accordance with the principles of the Declaration of Helsinki, and received approval from the Institutional Review Board of Tianjin Fourth Central Hospital. Written informed consent was obtained from all participants prior to enrollment. Between September 2021 and December 2022, 120 consecutive patients diagnosed with ACS were recruited at Tianjin Fourth Central Hospital. The diagnosis of ACS was established based on the Emergency Rapid Diagnosis and Treatment Guidelines for ACS. All participants received standardized doses of antiplatelet agents and statins. For the control group, blood samples were collected within 3-5 days after diagnosis. The experimental group received a subcutaneous injection of evolocumab within 48 hours of diagnosis, and blood samples were collected 72 hours post-injection, prior to the peak lipid-lowering effect of evolocumab. Pharmacokinetic studies have indicated that plasma concentrations of evolocumab peak approximately 3-4 days after injection, with a maximal LDL-C reduction observed at around 1 week, and then gradually diminishing over 2 weeks.¹² To minimize potential confounding from its lipid-lowering activity and ensure consistent timing relative to hospitalization, the 72-hour post-injection window was selected for sample collection. Blood samples were drawn into sodium citrate tubes and centrifuged at 180 × g for 15 minutes at room temperature (RT) to obtain platelet-rich plasma (PRP). Washed platelets were subsequently prepared by centrifuging PRP at 300 × g for 10 minutes at RT. The expression of TLR4 on the platelet surface was quantified by flow cytometry. Data are presented as mean ± standard error of the mean (SEM).

In vitro study

Blood samples anticoagulated with sodium citrate were collected between 7:00 and 8:00 a.m. from healthy subjects (HS) (n = 5; 3 males, 2 females; age 35.6 ± 5.6 years) under fasting conditions. PRP was prepared by centrifugation at 180 × g for 15 minutes at RT. The PRP aliquots were then incubated for 30 minutes at RT under the following conditions: PRP + normal saline, PRP + PCSK9 (0.8 μg/mL), PRP + PCSK9 (0.8 μg/mL) + evolocumab (100 μg/mL), and PRP + PCSK9 (0.8 μg/mL) + resatorvid (TAK-242) (1 μmol/L). Platelet aggregation was subsequently induced by adenosine diphosphate (ADP, 10 μmol/L), and the maximum aggregation rate in 5 minutes was recorded immediately using a platelet aggregometer. The extent of platelet aggregation was expressed as a percentage (Figure 1). Data are presented as mean ± SEM (n = 5).

Figure 1.

Platelet aggregation (A) and serum TXA₂ (B) in vitro study. * p < 0.05 vs. the NS group, ^# p < 0.05 vs. the PCSK9 group. NS, normal saline; PCSK9, proprotein convertase subtilisin/kexin type 9; TAK-242, resatorvid; TXA₂, thromboxane A₂.

Thromboxane A₂ (TXA₂) assay

Blood samples collected from both clinical and in vitro experiments were centrifuged, and the resulting plasma was stored at -80 °C until analysis. The plasma concentration of TXA₂ was measured using a commercial enzyme-linked immunosorbent assay (ELISA) kit according to the manufacturer’s instructions and expressed in pg/mL.

Immunofluorescence analysis

Cell culture coverslips were coated with fibrinogen (50 μg/mL) at 4 °C for 24 hours and maintained in a moist environment at 4 °C until use. Before the experiment, the coverslips were placed at the bottom of 12-well plates and equilibrated to 37 °C prior to cell seeding. Megakaryoblastic leukemia cell line (MEG-01) cells (CL-0498, Pricella, China) were co-incubated with PCSK9 (0.8 μg/mL) at 37 °C for 4 hours, followed by fixation with 4% paraformaldehyde for 20 minutes and subsequent immunofluorescence processing. Primary staining was performed using anti-Cluster of differentiation antigen 41 (CD41) (a platelet surface marker) and anti-PCSK9 antibodies, with CD41 visualized as green fluorescence and PCSK9 as red fluorescence. Nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI). After confirming cell identity, dual immunofluorescence staining with anti-TLR4 and anti-PCSK9 antibodies was conducted, where TLR4 and PCSK9 were visualized as green and red fluorescence, respectively. Nuclei were again counterstained with DAPI.

Statistical analysis

All data are presented as mean ± SEM. Comparisons between two groups were performed using the Student’s t-test. A p value of < 0.05 was considered statistically significant. Statistical analyses were conducted using SPSS version 19.0 (IBM Corp., Armonk, NY, USA).

RESULTS

Human study

The clinical characteristics of the ACS patients enrolled in the study are summarized in Table 1. Compared with the control group, patients in the evolocumab-treated group had a lower expression of TLR4 on the platelet surface. This decrease was consistent across all ACS subtypes, with no significant differences observed among them (Figure 2). Furthermore, compared with the patients who did not receive evolocumab, those who received evolocumab had lower levels of platelet activation markers, including TXA₂ (Figure 3).

Table 1. Clinical characteristics of ACS patients enrolled in the study.

	Control group (n = 60)	Evolocumab group (n = 60)	p value
Age (years)	59.88 ± 8.96	60.53 ± 10.48	0.72
Female sex (%)	20	16.67	0.32
Smoking (%)	58.33	60	0.37
SBP (mmHg)	128.83 ± 19.03	123.10 ± 23.69	0.15
DBP (mmHg)	79.97 ± 14.32	75.28 ± 14.66	0.08
PLT (×109/L)	241.90 ± 55.22	231.97 ± 73.56	0.40
Glu (mmol/L)	6.97 ± 2.63	7.03 ± 3.52	0.91
Cr (mmol/L)	70.77 ± 18.49	71.25 ± 15.65	0.88
TG (mmol/L)	1.80 ± 1.21	1.71 ± 0.86	0.65
TC (mmol/L)	4.94 ± 1.09	4.71 ± 1.09	0.24
LDL-C (mmol/L)	3.15 ± 0.89	2.99 ± 0.88	0.29
HDL-C (mmol/L)	1.03 ± 0.25	0.96 ± 0.20	0.05
UA (mmol/L)	322.20 ± 80.43	338.56 ± 88.74	0.30
EF (%)	52.95 ± 7.46	55.20 ± 6.06	0.39
Arterial hypertension (%)	60	63.33	0.41
Diabetes mellitus (%)	33.33	36.67	0.37
β-blocker (%)	90	95	0.47
Proton pomp inhibitors (%)	98.33	96.67	0.59

ACS, acute coronary syndrome; Cr, creatinine; DBP, diastolic blood pressure; EF, ejection fraction; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; PLT, platelet; SBP, systolic blood pressure; TC, total cholesterol; TG, triglyceride; UA, uric acid.

Figure 2.

Expression rate of TLR4 on platelets. Patients with ACS (A), patients with STEMI (B), patients with NSTEMI (C), patients with UAP (D). (Data are represented as median and IQR. * p < 0.05, ** p < 0.01). ACS, acute coronary syndrome; IQR, interquartile range; NSTEMI, non-ST-segment elevation myocardial infarction; STEMI, ST-segment elevation myocardial infarction; TLR4, toll-like receptor 4; UAP, unstable angina pectoris.

Figure 3.

Serum TXA₂. Patients with ACS (A), patients with STEMI (B), patients with NSTEMI (C), patients with UAP (D). (Data are represented as median and IQR. * p < 0.05, ** p < 0.01). ACS, acute coronary syndrome; IQR, interquartile range; NSTEMI, non-ST-segment elevation myocardial infarction; STEMI, ST-segment elevation myocardial infarction; TXA₂, thromboxane A₂; UAP, unstable angina pectoris.

In vitro study

Recombinant human PCSK9 enhanced adenosine diphosphate(ADP)-induced platelet aggregation. This pro-aggregatory effect of PCSK9 was effectively neutralized by evolocumab, a PCSK9 inhibitor, and by TAK-242, a TLR4 inhibitor.

Platelets co-incubated with PCSK9 and subsequently stimulated with ADP exhibited a significantly higher aggregation response compared with platelets stimulated with ADP alone (Figure 1). Notably, this enhancement was abolished when either evolocumab or TAK-242 was added during incubation. Platelet aggregation in samples co-incubated with both inhibitors showed no significant difference from that in the control group. In addition, PCSK9 markedly increased plasma TXA₂ production, an effect that was also attenuated by treatment with evolocumab and TAK-242 (Figure 1). Of note, there were slight variations in the baseline optical density (OD) readings of the technical controls across different experimental plates. This is recognized as inherent plate-to-plate variability common in ELISA assays. To account for this, all raw OD values from the same plate were normalized to the mean value of the technical controls on that respective plate prior to inter-group comparisons and statistical analysis.

Immunofluorescence analysis

Immunofluorescence staining with anti-CD41 and anti-PCSK9 antibodies confirmed the identity of the cells as megakaryocytes (platelet precursor cells) and demonstrated intracellular localization of PCSK9 following incubation with recombinant PCSK9. Dual immunofluorescence staining with anti-TLR4 and anti-PCSK9 antibodies further revealed co-localization of TLR4 and PCSK9 in MEG-01 cells (Figure 4).

Figure 4.

Immunofluorescence staining and confocal microscopy analysis demonstrated the colocalization of TLR4 (green) and PCSK9 (red) in MEG-01 cells. A in MEG-01 cells, green fluorescence signal indicates CD41, red fluorescence marks PCSK9, nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI). B at 400 × magnification, in MEG-01 cells, green fluorescence signal indicates TLR4, red fluorescence marks PCSK9, nuclei were stained with DAPI. C at 1000 × magnification, in MEG-01 cells, green fluorescence signal indicates TLR4, red fluorescence marks PCSK9, nuclei were stained with DAPI. CD41, Cluster of differentiation antigen 41; MEG-01, megakaryoblastic leukemia cell line; PCSK9, proprotein convertase subtilisin/kexin type 9; TLR4, Toll-like receptor 4.

DISCUSSION

The present study yielded several key findings: 1) In vitro, PCSK9 significantly enhanced platelet aggregation, an effect that was abolished by evolocumab and the TLR4 inhibitor TAK-242. 2) PCSK9 increased the release of platelet activation markers, including TXA₂, which was similarly attenuated by evolocumab and TAK-242. 3) In the patients with ACS, evolocumab treatment reduced the surface expression of TLR4 on platelets. 4) Evolocumab also decreased the plasma levels of platelet activation markers such as TXA₂ in the ACS patients. 5) Immunofluorescence analysis revealed co-localization of PCSK9 and TLR4 in MEG-01 megakaryocytic cells, strongly suggesting a potential molecular interaction between the two proteins.

PCSK9 inhibitors are best known for their ability to lower LDL-C, and they are widely recognized as being safe and effective lipid-lowering agents. Beyond their lipid-regulatory function, a growing body of evidence indicates that PCSK9 exerts pro-atherogenic effects independent of its influence on plasma cholesterol levels. Tiller et al. reported a significant increase in PCSK9 in patients with ST-segment elevation myocardial infarction between 24 and 48 hours after percutaneous coronary intervention.¹³ Several clinical studies have also implicated PCSK9 in processes such as atherosclerotic inflammation,¹⁴ oxidative stress,¹⁵ and platelet activation. Given the central role of platelets in atherothrombosis, the present study focused on the interaction between PCSK9 and platelets. Accumulating clinical and experimental data support the hypothesis that PCSK9 can modulate platelet activation through multiple mechanisms. These findings are supported by studies involving patients with ACS,¹⁶ symptomatic coronary artery disease,¹⁷ atrial fibrillation,⁶ and hypercholesterolemia,^18,19 as well as by results from animal models.²⁰ In this study, we extended these observations by examining ACS patients, supported by in vitro experiments and immunofluorescence analysis, to further elucidate the mechanisms underlying PCSK9-mediated platelet activation.

Numerous studies have demonstrated that circulating PCSK9 directly promotes platelet activation.⁶ The in vivo administration of PCSK9 has been shown to markedly enhance agonist-induced platelet aggregation accompanied by significant elevations in platelet activation and oxidative stress markers.^6,21 PCSK9 has also been shown to augment ADP-induced human platelet aggregation in a concentration-dependent manner. Accordingly, we used PCSK9 (0.8 μg/mL) and evolocumab (100 μg/mL) in the in vitro experimental model in the present study. The results revealed that, compared with the control group, PCSK9 incubation significantly increased the platelet aggregation rate and TXA₂ release. These effects were completely abolished by the addition of evolocumab. Moreover, no significant differences were observed in platelet aggregation or TXA₂ levels between the control group and the group receiving combined PCSK9 and evolocumab treatment. These findings are consistent with previous reports. Mechanistic studies have shown that PCSK9 binds to platelet scavenger receptors CD36 and lectin-like oxidized low-density lipoprotein receptor-1 (LOX-1), activating downstream signaling cascades that enhance platelet activation. Recent studies further showed that the dual inhibition of oxidized LDL receptors, CD36 and LOX-1, led to substantial reductions in platelet activation and oxidative stress.^18,21

Beyond the established CD36 pathway, the present study explored TLR4 as a potential alternative pathway mediating PCSK9-induced platelet activation. Foundational work by Andonegui et al. first demonstrated the presence of functional TLR4 receptors on platelets.²¹ As a platelet surface receptor, TLR4 contributes to thrombosis through multiple mechanisms.^9,10,22 For example, a study examining the effects of histones on platelet procoagulant activity revealed that histone-stimulated platelets promoted plasma thrombin generation, with TLR2 and TLR4 jointly mediating this process.²² Subsequently, research by Vogel and colleagues showed that genetic deletion of platelet TLR4 led to reduced caspase-1 activation and diminished platelet aggregation, underscoring the critical role of platelet TLR4/NLRP3 signaling in promoting aggregation.¹¹ Furthermore, an in vitro study on oxLDL-stimulated RAW264.7 macrophages demonstrated that PCSK9 enhanced the production of pro-inflammatory cytokines primarily through the upregulation of TLR4 expression and activation of the TLR4/NF-κB signaling pathway.²³ These findings suggest that the pro-inflammatory actions of PCSK9 may be partially mediated through TLR4/NF-κB signaling. In nucleated cells, PCSK9 can trigger pro-inflammatory cytokine production via TLR4, which in turn may influence platelet activation. Building upon this evidence, the present study aimed to determine whether PCSK9 similarly modulates platelet activation through the TLR4-mediated pathway in platelets.

In the present study, we used flow cytometry to assess TLR4 expression on platelets. Our findings revealed that platelet TLR4 expression was significantly reduced following evolocumab administration in the patients with ACS. This suggests that evolocumab may downregulate platelet TLR4 expression in patients with ACS, indicating a potential interaction between PCSK9 and platelet TLR4 that could modulate platelet activation. Consistently, plasma TXA₂ levels were markedly decreased after evolocumab treatment, implying that evolocumab may suppress the production of platelet activation-related mediators, corroborating previous reports. Whether the antiplatelet effects of PCSK9 inhibitors are mediated independently of their potent LDL-C-lowering efficacy remains incompletely understood and is a subject of ongoing debate.²⁴ To minimize confounding by the lipid-lowering effects of evolocumab, we measured TLR4 expression prior to the expected onset of its maximal pharmacodynamic activity. Considering that TLR4 is an inflammation-associated receptor and that a PCSK9-TLR4/MyD88/NF-κB signaling axis has been previously described, the interaction between PCSK9 and TLR4 is likely mediated through inflammatory rather than lipid-metabolic pathways. Nevertheless, the absence of concurrent plasma PCSK9 and LDL-C measurements precludes a definitive distinction between direct anti-inflammatory effects and those secondarily linked to lipid lowering. Future studies incorporating simultaneous assessments of PCSK9, LDL-C, and TLR4 expression are warranted to clarify whether PCSK9 inhibition modulates TLR4 via a mechanism independent of lipid metabolism.

To further verify the involvement of TLR4 in PCSK9-mediated platelet activation, we conducted in vitro experiments using TAK242 and evolocumab. The results demonstrated that the PCSK9-induced enhancement of platelet aggregation was abolished following treatment with either TAK242 or evolocumab. Correspondingly, TXA₂ levels declined significantly, reaching values comparable to the control group. These findings suggest that both TAK242 and evolocumab effectively suppress PCSK9-facilitated platelet aggregation and the release of platelet activation markers. Hence, it is plausible that PCSK9 promotes platelet activation, at least in part, through modulation of platelet TLR4. To further elucidate the mechanism of whether PCSK9 regulates TLR4 expression or directly interacts with the receptor, immunofluorescence co-localization assays were performed in MEG-01 megakaryocytic cells. The results revealed co-localization of PCSK9 and TLR4, suggesting a potential direct interaction that may influence platelet activation. However, whether this pathway is independent of the established CD36 signaling mechanism remains to be determined.

Given the aforementioned mechanisms and empirical evidence, we hypothesize that in patients with ACS, PCSK9 may modulate platelet aggregation and the secretion of platelet activation-related markers by regulating or interacting with TLR4 receptors, an effect that appears to be inhibited by evolocumab and TAK242.

Nevertheless, this study has several limitations and important implications, and the findings should therefore be interpreted with caution. A minor technical limitation is the observed baseline variation in OD values among the technical controls across different assay plates, which reflects inherent inter-plate variability in ELISA. We mitigated this potential bias by normalizing all raw data to the plate-specific control values, and therefore, this variability does not affect the relative treatment-induced changes or the validity of our primary conclusions. In addition, the absence of plasma PCSK9 concentration data and measurements of evolocumab-induced LDL-C reduction precluded stratified analyses based on circulating PCSK9 levels. Consequently, it remains uncertain whether the observed reduction in TLR4 expression reflects a direct lipid-independent effect or one secondary to lipid lowering. In addition, the immunofluorescence co-localization results suggest a potential interaction between PCSK9 and TLR4; however, we recognize that this constitutes preliminary evidence. Definitive proof of a direct interaction requires further validation through co-immunoprecipitation experiments in future studies. Moreover, the in vitro experiments used platelets derived from HS; future studies incorporating platelets from ACS patients would enhance the clinical relevance and translational value of the results. Finally, the use of CD36^–/– mice in future investigations could provide further mechanistic insights into whether PCSK9 influences platelet activation through a TLR4-dependent pathway that is independent of the established CD36 signaling axis.

CONCLUSIONS

This study provides the first evidence that PCSK9 inhibition can downregulate TLR4 expression on the platelet surface in patients with ACS, and that the pro-activating effect of PCSK9 on platelets can be effectively counteracted by a TLR4 inhibitor.

DECLARATION OF CONFLICT OF INTEREST

All authors declare that they have no conflict of interests.

ETHICS APPROVAL AND CONSENT TO PARTICIPATE

The present investigation involving human subjects strictly adhered to the stipulations of the Declaration of Helsinki and procured approval from the Institutional Review Board of the Tianjin Fourth Central Hospital, approval number: SZXLL-2021-K040.

CONSENT FOR PUBLICATION

Written informed consent was obtained from a legally authorized representative(s) for anonymized patient information to be published in this article.

AVAILABILITY OF DATA AND MATERIALS

The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.

FUNDING

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

AUTHORS’ CONTRIBUTIONS

Conceptualization, Kangyin Chen, Yue Zhang and Lingshan Qi; methodology, Yue Zhang, Meng Yuan and Lingshan Qi; software, Lingshan Qi; validation, Lingshan Qi; formal analysis, Xing Liu and Lingshan Qi; investigation, Xing Liu and Lingshan Qi; resources, Kangyin Chen; data curation, Lingshan Qi; writing — original draft preparation, Lingshan Qi; writing — review and editing, Kangyin Chen and Yue Zhang; visualization, Lingshan Qi; supervision, Kangyin Chen; project administration, Kangyin Chen and Tong Liu; funding acquisition, Kangyin Chen. All authors have read and agreed to the published version of the manuscript.