Role of LL-37 in thrombotic complications in patients with COVID-19

Blood clot formation induced by dysfunctional coagulation is a frequent complication of coronavirus disease 2019 (COVID-19) and a high-risk factor for severe illness and death. Neutrophil extracellular traps (NETs) are implicated in COVID-19-induced immunothrombosis. Furthermore, human cathelicidin, a NET component, can perturb the interaction between the SARS-CoV-2 spike protein and its ACE2 receptor, which mediates viral entry into cells. At present, however, the levels of cathelicidin antimicrobial peptides after SARS-CoV-2 infection and their role in COVID-19 thrombosis formation remain unclear. In the current study, we analyzed coagulation function and found a decrease in thrombin time but an increase in fibrinogen level, prothrombin time, and activated partial thromboplastin time in COVID-19 patients. In addition, the cathelicidin antimicrobial peptide LL-37 was upregulated by the spike protein and significantly elevated in the plasma of patients. Furthermore, LL-37 levels were negatively correlated with thrombin time but positively correlated with fibrinogen level. In addition to platelet activation, cathelicidin peptides enhanced the activity of coagulation factors, such as factor Xa (FXa) and thrombin, which may induce hypercoagulation in diseases with high cathelicidin peptide levels. Injection of cathelicidin peptides promoted the formation of thrombosis, whereas deletion of cathelicidin inhibited thrombosis in vivo. These results suggest that cathelicidin antimicrobial peptide LL-37 is elevated during SARS-CoV-2 infection, which may induce hypercoagulation in COVID-19 patients by activating coagulation factors. Supplementary Information The online version contains supplementary material available at 10.1007/s00018-022-04309-y.

Cathelicidins, which belong to the family of host defense peptides, play an important role in innate immunity [31]. They exhibit a broad-spectrum effect against pathogens via direct microbicidal and immunomodulatory activities [32]. LL-37 is the only human member of the cathelicidin antimicrobial peptide family and is derived from human cathelicidin antimicrobial protein 18 (hCAP18) by the cleavage of proteinase 3 [33]. Although cathelicidins have mostly been studied with respect to their antibacterial and immunomodulatory activity, elevated levels are reported to aggravate diseases, such as psoriasis [34], atherosclerosis [35] and ulcerative colitis [36], by induction of inflammation. Cathelicidins may initiate and propagate thrombosis by activating platelets [37,38]; however, their role in coagulation cascade activation remains unclear.
Here, we investigated the role of LL-37 in COVID-19 patient coagulopathy. Results showed that increased LL-37 was correlated with COVID-19-related coagulation dysfunction. LL-37 may potentiate the activity of coagulation factors, such as FXa and thrombin, thereby contributing to hypercoagulation in COVID-19.

Experimental ethics
All human specimens and clinical information were collected with informed consent of the patients prior to the study from the Chongqing Public Health Medical Center (CPHMC) and Department of Infectious Diseases, Southwest Hospital, Third Military Medical University (Army Medial University). Patients with laboratory-confirmed COVID-19 (n = 62) and age-and sex-matched healthy controls (HCs, n = 21) were included in this study. For the measurement of LL-37, thrombin time (TT), fibrinogen, prothrombin time (PT) and activated partial thromboplastin time (APTT), numbers of each group indicated in the text. COVID-19 patients were divided into mild or moderate (MM, n = 40) and severe or critical (SC, n = 22) groups according to the Chinese Clinical Guidance for COVID-19 Pneumonia Diagnosis and Treatment (6th edition). Briefly, patients with mild or moderate (MM) disease were defined based on the following clinical symptoms: (1) Mild clinical symptoms, with no sign of pneumonia on chest imaging; (2) Fever and respiratory symptoms, with signs of pneumonia through radiological assessment. Patients with severe or critical (SC) disease were defined based on the following clinical symptoms: (1) Shortness of breath, respiratory rate (RR) ≥ 30 times/min, oxygen saturation ≤ 93% at rest, alveolar oxygen partial pressure/fraction of inspiration O 2 (PaO 2 /FiO 2 ) ≤ 300 mmHg; (2) Respiratory failure requiring mechanical ventilation, shock, combined with other organ failure needed ICU monitoring and treatment. Determination of LL-37 in the plasma of COVID-19 patients and HCs was approved by the Ethics Committee of Chongqing Public Health Medical Center (2020-002-01-KY, 2020-003-01-KY). The study and all animal experiments were approved by the Institutional Review Board and Animal Care and Use Committee at Kunming Institute of Zoology (SMKX-20201021-15).

Cell culture and treatment
The human lung epithelial cell line A549 was purchased from the Kunming Cell Bank and maintained in complete Dulbecco's Modified Eagle Medium (Corning, 10-013-CVR) supplemented with 10% fetal bovine serum, 100 U/ ml penicillin, and 100 μg/ml streptomycin (Gibco BRL, Gaithersburg, MD, USA) at 37 °C in 5% CO 2 .
To determine the effects of SARS-CoV-2 infection on LL-37 expression, A549 cells were stimulated with SARS-CoV-2 (MOI: 0.01, 0.05, 0.25) for 2 h at the biosafety level-3 laboratory of the Kunming High-level Biosafety Primate Research Center, Yunnan, China, with LL-37 expression then detected using confocal microscopy and enzyme linked immunosorbent assay (ELISA) after 24 h. To confirm whether the effects of SARS-CoV-2 on LL-37 expression were dependent on the spike protein, we stimulated A549 cells with spike protein (0.4-10 μg/ml), bovine serum albumin (BSA, 10 μg/ml, negative control), lipopolysaccharides (LPS, positive control) from Escherichia coli O111:B4 (10 μg/ml) for 24 h, then measured LL-37 expression using confocal microscopy and Western blot analysis.

ELISA
The levels of LL-37 in the plasma of COVID-19 patients and supernatant of SARS-CoV-2 stimulated A549 cells were analyzed using a LL-37 ELISA kit (Hycult Biotech, HK321-01) according to the manufacturer's instructions. A binding assay between cardiolipin and LL-37 was carried out by ELISA according to previously described methods [39]. Briefly, a 96-well white plate (Corning, Kennebunk ME, USA) was filled with 50 μl of 50 μg/ml cardiolipin diluted in ethanol and evaporated at 4 °C. After washing with phosphate-buffered saline (PBS; pH 7.4), the wells were blocked with 2% BSA in PBS (1 h at room temperature). Then, 100 μl of FITC-labeled LL-37 or FITC-labeled Cramp (10 μg/ml) was added, followed by incubation for 1 h at 37 °C. After washing with PBS (pH 7.4), fluorescence was detected using a Cytation 3 Cell Imaging Multi-Mode Reader (Biotek), and the binding of LL-37 with cardiolipin was calculated. In all assays, ethanol-treated wells were used as negative controls.

Coagulation functional assay and enzymatic activity assay of coagulation factors
Coagulation function assays (TT, APTT, PT) were conducted by detecting absorbance at 650 nm using the SUN-BIO kit according to the manufacturer's instructions. Enzymatic activity of coagulation factors was measured using chromogenic substrate. Briefly, coagulation factors (FXa, 0.1 nM, thrombin, 10 nM) and different amounts of peptides (final concentrations ranging from 0.0 to 16.2 μg/ml) were pre-incubated for 10 min at 37 °C. After incubation, the reaction was initiated by the addition of 0.5 mM substrate 309 Page 4 of 14 F3301 (CH3OCO-D-CHA-Gly-Arg-pNA-AcOH, Sigma) for FXa and H-D-Phe-Pip-Arg-pNa·2HCl (Hyphen Biomed, Neuville-sur-Oise, France) for thrombin, with the reaction monitored continuously at 405 nm.
To determine the enzymatic activity of thrombin and FXa in plasma from Cramp knockout (Cramp −/− ) and C57BL/6 J mice (n = 6-7 per group), thrombin and FXa chromogenic substrates were added to the plasma, with the reaction initiated immediately and monitored continuously at 405 nm.

Surface plasmon resonance (SPR) analysis
SPR analysis was performed as described previously, with some modifications [40]. Briefly, thrombin and FXa were immobilized on the activated sensor chip CM-5 by amine coupling. LL-37 or Cramp in HBS-EP + running buffer was applied to the immobilized ligand at a flow rate of 30 μl/min and the real-time binding signal was recorded using BIAcore 3000 (GE, USA). The equilibrium dissociation constant (K D ) was calculated using the Langmuir model with Biacore evaluation software provided by the manufacturer.

FeCl 3 -induced mouse thrombosis model
Male C57BL/6 mice with or without cathelicidin antimicrobial peptides administration and Cramp −/− mice (6-8 weeks old, n = 3-7 per group) were anesthetized with 2% isoflurane. The left carotid artery was exposed and visualized through a dissecting microscope. A Doppler microvascular probe (RWD Life Science, Shenzhen, China) was placed on the exposed artery to measure vascular blood flow. Thrombosis was induced by directly placing a small piece of filter paper saturated with 20% FeCl 3 on the artery for 1.5 min. Time to vessel occlusion was measured when blood flow was completely stopped.

Acute pulmonary thromboembolism in mice
LL-37 (100 μl, 30 mg/kg) or Cramp (100 μl, 30 mg/kg) in saline was injected into male C57BL/6 mice (6-8 weeks old, n = 5 per group) via the caudal vein. Control mice received the same volume of saline. At 10 min after the injection, the mice were anesthetized with pentobarbital sodium (50 mg/ kg, intraperitoneal injection). The chest cavity was then exposed, and the right side of the heart was perfused with saline to remove blood from pulmonary circulation. After perfusion, the whole lung was excised and fixed in 4% paraformaldehyde dissolved in PBS at 4 °C overnight for histopathological examination.

Statistical evaluation
Data obtained from independent experiments were presented as mean ± standard deviation (SD). For normal continuous variables, one-way analysis of variance (ANOVA) was used. Comparisons of more than two groups were performed using Kruskal-Wallis one-way ANOVA followed by Dunn's multiple comparison test using GraphPad Prism v5. Differences were considered significant at p < 0.05.

Elevated LL-37 in plasma of COVID-19 patients
Cathelicidin peptides increase during viral infection, which induces the formation of thrombosis by activating platelets [37,38]. The human cathelicidin antimicrobial peptide LL-37 is reported to inhibit SARS-CoV-2 infection [30]. To investigate the role of cathelicidin peptides in SARS-CoV-2 infection, the level of LL-37 in the plasma of COVID-19 patients (MM and SC) and HCs was measured. As illustrated in Fig. 2A, the concentration of LL-37 in the plasma of COVID-19 patients (140 ± 46.47 ng/ml in MM and 147.6 ± 64.24 ng/ml in SC) was significantly higher than that in HCs (93.62 ± 48.14 ng/ml).

SARS-CoV-2 upregulates LL-37 expression through the spike protein
To further investigate the association between LL-37 overexpression and SARS-CoV-2 infection, the expression of hCAP18 (precursor of LL-37) in A549 cells after SARS-CoV-2 spike protein (0.4-10 μg/ml) incubation for 24 h at 37 ℃ was determined by Western blot analysis. From the results of Fig. 2B, hCAP18 expression in A549 cells significantly increased after incubation with the spike protein.

LL-37 promotes thrombosis formation through potentiation of coagulation factor activity
As LL-37 was correlated with TT and fibrinogen, the effects of LL-37 on coagulation factor activity were determined. Based on the chromogenic substrate assay, LL-37 enhanced the activity of thrombin (Fig. 4A) and FXa (Fig. 4B) in a dose-dependent manner. SPR was conducted to analyze the binding capacity of LL-37 with thrombin and FXa. As their physiological substrates, fibrinogen (Fig. 4C) and prothrombin (Fig. 4D), respectively.
Given the effect of LL-37 on thrombin and FXa, we further investigated the effects of cathelicidin on thrombosis formation in vivo. As seen in Fig. 4E, LL-37 significantly promoted thrombosis formation and shortened the time of arterial occlusion in the FeCl 3 -induced carotid artery thrombosis mouse model. LL-37 administration also induced lung thrombosis directly (Fig. 4G).

Cramp promotes thrombosis formation through promotion of coagulation factor activity
To confirm the effects of LL-37 on thrombosis formation through activation of coagulation factors, we detected the effects of Cramp (LL-37 homolog from Mus musculus) on coagulation factor activation. Similar to the LL-37 results, the chromogenic substrate assay showed that Cramp significantly enhanced the enzymatic activity of thrombin (Fig. 5A) and FXa (Fig. 5B) in a dose-dependent manner. SPR analysis also indicated interactions of Cramp with thrombin ( Fig. S1C) and FXa (Fig. S1D), with K D values of 6.78 × 10 -6 M and 3.17 × 10 -4 M, respectively. Based on the K D results, the binding capacity of LL-37 with coagulation factors was higher than that of Cramp. The results of thrombin and FXa on their natural substrates (fibrinogen and prothrombin) confirmed that Cramp promoted the enzymatic activity of thrombin (Fig. 5C) and FXa (Fig. 5D).
Similar to the LL-37 results, Cramp administration significantly enhanced lung thrombosis in acute pulmonary thromboembolism mouse model (Fig. 5G) and shortened the time of arterial occlusion in the FeCl 3 -induced carotid artery thrombosis mouse model (Fig. 5H). Furthermore, Cramp deletion was resistant to arterial occlusion in the FeCl 3 -induced carotid artery thrombosis mouse model (Fig. 5I).

LL-37 and Cramp show similar effects on PT and APTT in COVID-19 patients.
LL-37 and Cramp enhanced coagulation factor (thrombin and FXa) activity, which should shorten PT and APTT. However, LL-37 had no effect on PT (Fig. S2A) and prolonged APTT (Fig. S2B) at higher concentrations. Somewhat inconsistent with LL-37, Cramp prolonged PT (Fig.  S2C) and APTT (Fig. S2D) in a dose-dependent manner. Thus, these findings were contradictory to expectation, but were consistent with the PT and APTT values found in COVID-19 patients.
Basic histone, which can induce hypercoagulation, prolongs PT and APTT by binding with phospholipids [41], suggesting that basic LL-37 or Cramp may prolong PT and APTT through interaction with phospholipids. Indeed, based on the ELISA results, LL-37 and Cramp could bind with cardiolipin (Fig. S3A).Furthermore, we detected the effects of pre-incubation with cardiolipin on cathelicidin peptideinduced PT/APTT prolongation. LL-37 induced PT prolongation tendency at concentration of 25 μM, which could be inhibited by pre-incubation with cardiolipin (Fig. S3B). In addition, LL-37-induced APTT prolongation was significantly inhibited by cardiolipin pre-incubation (Fig. S3C). Similar to LL-37, pre-incubation with cardiolipin significantly abolished Cramp-induced PT (Fig. S3D) and APTT (Fig. S3E) prolongation. From these results, PT/APTT prolongation induced by LL-37 or Cramp may be caused by interactions with phospholipids, which can inhibited by the pre-incubation with cardiolipin.

Discussion
To the best of our knowledge, this is the first study to report on the correlation between elevated LL-37 levels and hypercoagulation in COVID-19 patients. LL-37 was upregulated by SARS-CoV-2 infection to cause elevated concentration in the plasma of COVID-19 patients and showed the ability to directly activate coagulation factors. The upregulation of LL-37 was associated with clinical hypercoagulation manifestations induced by SARS-CoV-2 infection and likely contributes to the hypercoagulation frequently observed in COVID-19 patients (Fig. 6).
Here, we showed that LL-37 induced hypercoagulability through enhancement of coagulation factor activity. LL-37 has also been found to induce endothelial cell dysfunction, inflammation, NETs formation, platelet activation, which may promote thrombosis in COVID-19.
The hCAP18 protein contains a conserved cathelin-like domain and a highly variable C-terminal peptide (LL-37). The cathelin-like domain of the cathelicidin is classified into the same superfamily as cystatins, the cysteine protease inhibitors, and the cathelin-like domain of hCAP18 can inhibit cathepsin L activity [44]. Although few studies have explored the function of the cathelicidin C-terminal peptide on protease activity, a recent study on cathelicidin-MH (cath-MH) from the skin of Microhyla heymonsivogt frog was found to suppress coagulation by affecting enzymatic activities [45], inconsistent with the effects of LL-37. Through sequence alignment analysis, LL-37 displays low sequence identity with cath-MH and lacks the loop formed by the intramolecular disulfide bond [45], which may explain their discrepancy in coagulation factor activity.
Disseminated intravascular coagulopathy (DIC) has been reported in COVID-19 patients [46]. In addition, consistent with classic DIC caused by bacterial sepsis, prolonged APTT, thrombocytopenia, elevated D-dimer, and multiorgan microangiopathic thrombosis have also been found in COVID-19 patients. The prolongation of APTT is difficult to explain, but our results suggest it may be induced by elevated LL-37. Notably, LL-37 activates the thrombin and FXa coagulation factors, leading to hypercoagulability, with the binding of phospholipids likely prolonging APTT.
As a major family of antimicrobial peptides, cathelicidin peptides are expressed over a broad range of sites during infection and inflammation, and are primarily generated by neutrophils and epithelial cells [47]. Elevated cathelicidins form part of the body's defense against pathogens, with the antiviral activity of cathelicidin peptides reported in many viruses e.g., human immunodeficiency virus (HIV)-1 [48], influenza A virus (IAV) [31,49], respiratory syncytial virus (RSV) [47], rhinovirus (HRV) [50], vaccinia virus (VACV) [51], herpes simplex virus (HSV) [52], zika virus (ZIKV) [53], dengue virus (DENV) [54] and hepatitis C virus (HCV) [55]. Recently, research reported the inhibition of LL-37 on SARS-CoV-2 infection using biochemical and pseudovirus entry assays [30]. While, the direct effects of LL-37 on SARS-CoV-2 remain unclear. Here we found LL-37 induced hypercoagulation through the potentiation of coagulation factor activities, which was consistent with the clinical symptoms of COVID-19 patients. Moreover, elevated LL-37 has been found to activate platelets. Therefore, although the elevation of LL-37 during SARS-CoV-2 infection may be a protective mechanism of the innate immune system, the increase in LL-37 may also aggravate disease progress by inducing thrombosis, which may explain the controversy of vitamin D (inducer of LL-37 production) treatment in COVID-19 [56].
In addition to its antimicrobial activity, LL-37 also exhibits various biological effects, such as regulation of  Data are mean ± SD of at least three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001 inflammation, cell proliferation and apoptosis. However, overexpression of LL-37 or LL-37 complexes with other molecules may contribute to progression of diseases. For example, the LL-37-DNA/RNA complex can aggravate psoriasis [34,57], atherosclerosis [35,58], ulcerative colitis [36], sepsis [59], thrombosis [37], and chronic obstructive pulmonary disease [60] through the induction of inflammation. Cytokine storms are another clinical marker of SARS-CoV-2, which induce higher morbidity and mortality in COVID-19 patients [61]. Circulating cell-free DNA [62], cell-free mitochondrial DNA [63], and cell-free microbial DNA [64] are increased in SARS-CoV-2 infection and are associated with disease severity and mortality in COVID-19 patients. Therefore, elevated LL-37 may induce the inflammation via interactions with increased cell-free DNA of COVID-19 patients, thereby exacerbating disease process.
In conclusion, we observed a close correlation between LL-37 and the hypercoagulation frequently observed in COVID-19 patients. The level of LL-37 was increased in the plasma of COVID-19 patients with the induction of SARS-CoV-2 spike protein. Elevated LL-37 may contribute to thrombosis via potentiation of coagulation factor activity. As the results of our research, although LL-37 has been found to perturb SARS-CoV-2 infection, it is not suitable for the treatment of SARS-CoV-2 infection, especially for patients in hypercoagulability.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.