Advertisement

Hantaan virus-induced elevation of plasma osteoprotegerin and its clinical implications in hemorrhagic fever with renal syndrome

Open AccessPublished:November 09, 2022DOI:https://doi.org/10.1016/j.ijid.2022.11.005

      Highlights

      • Plasma osteoprotegerin levels were elevated in patients with severe/critical hemorrhagic fever with renal syndrome.
      • Tumor necrosis factor-α synergizes with the Hantaan virus to promote osteoprotegerin production in endothelial cells.
      • Plasma osteoprotegerin correlated with hemorrhagic disorder indicators of hemorrhagic fever with renal syndrome.

      Abstract

      Objectives

      The bleeding tendency is a hallmark of hemorrhagic fever with renal syndrome (HFRS) after Hantaan virus (HTNV) infection. Growing reports indicate the importance of osteoprotegerin (OPG) in vascular homeostasis, implying OPG might be involved in the pathogenesis of coagulopathy in patients with HFRS.

      Methods

      Acute and convalescence plasmas of 32 patients with HFRS were collected. Enzyme-linked immunosorbent assays (ELISA) were used to detect plasma OPG levels and other parameters. The human umbilical vein endothelial cells were stimulated with HTNV and/or tumor necrosis factor-α (TNF-α) to explore the source of OPG.

      Results

      Plasma OPG levels of patients with HFRS were elevated and correlated positively with the severity of HFRS and negatively with platelet counts. Abundant OPG was released from endothelial cells in response to TNF-α stimuli, along with HTNV infection, which was in accordance with the findings of positive correlations between plasma OPG and TNF-α or c-reactive protein. Importantly, plasma OPG levels correlated positively with activated partial thromboplastin time and the content of d-dimer.

      Conclusion

      These findings suggested that increased plasma OPG levels induced by HTNV might be an important factor for the severity of HFRS, and was likely involved in endothelium dysfunction and hemorrhagic disorder of HFRS, which might contribute to the pathogenesis of hemorrhage in HFRS.

      Keywords

      Introduction

      Hemorrhagic fever with renal syndrome (HFRS) has gained widespread attention as a major public health concern (
      • Brocato RL
      • Hooper JW.
      Progress on the prevention and treatment of hantavirus disease.
      ). China is a severe endemic area where patients with HFRS, especially those infected with the Hantaan virus (HTNV) and Seoul virus, account for 90% of the total global cases (
      • Jiang H
      • Zheng X
      • Wang L
      • Du H
      • Wang P
      • Bai X.
      Hantavirus infection: a global zoonotic challenge.
      ). Although hantavirus infections do not disrupt the vascular endothelium, the underlying pathogenesis of HFRS is undoubtedly related to increased vascular permeability and pronounced thrombocytopenia (
      • Noack D
      • Goeijenbier M
      • Reusken CBEM
      • Koopmans MPG
      • Rockx BHG.
      Orthohantavirus pathogenesis and cell tropism.
      ). Remarkably, a decline in platelet numbers usually starts very early from the onset of infection. The platelet numbers negatively correlate with HTNV RNA load (
      • Yi J
      • Xu Z
      • Zhuang R
      • Wang J
      • Zhang Y
      • Ma Y
      • et al.
      Hantaan virus RNA load in patients having hemorrhagic fever with renal syndrome: correlation with disease severity.
      ). However, there is little understanding of the precise mechanism of the decreased platelets and abnormal coagulation in HFRS.
      The vascular effects of osteoprotegerin (OPG) have received increased attention in recent years. OPG is a soluble glycoprotein belonging to the tumor necrosis factor (TNF) receptor superfamily (
      • Rochette L
      • Meloux A
      • Rigal E
      • Zeller M
      • Cottin Y
      • Vergely C.
      The role of osteoprotegerin and its ligands in vascular function.
      ). Functionally, OPG acts as a decoy receptor for receptor activator of nuclear factor-κB ligand (RANKL) and TNF-related apoptosis-inducing ligand (TRAIL) to regulate processes, such as cell apoptosis/survival and necroptosis, immune surveillance, and host defense (
      • Bernardi S
      • Bossi F
      • Toffoli B
      • Fabris B.
      Roles and clinical applications of OPG and TRAIL as biomarkers in cardiovascular disease.
      ;
      • Udagawa N
      • Koide M
      • Nakamura M
      • Nakamichi Y
      • Yamashita T
      • Uehara S
      • et al.
      Osteoclast differentiation by RANKL and OPG signaling pathways.
      ). OPG also has direct ligand-independent effects on the vasculature and immune system (
      • Bernardi S
      • Bossi F
      • Toffoli B
      • Fabris B.
      Roles and clinical applications of OPG and TRAIL as biomarkers in cardiovascular disease.
      ;
      • Zhang R
      • Liu J
      • Yu S
      • Sun D
      • Wang X
      • Fu J
      • et al.
      Osteoprotegerin (OPG) promotes recruitment of endothelial progenitor cells (EPCs) via CXCR4 signaling pathway to improve bone defect repair.
      ). Notably, the specific effects of OPG on the vasculature, including stimulation of vasculogenesis, and expression of adhesion molecules on endothelium have been widely reported (
      • Rochette L
      • Meloux A
      • Rigal E
      • Zeller M
      • Cottin Y
      • Vergely C.
      The role of osteoprotegerin and its ligands in vascular function.
      ). The importance of OPG in regulating endothelial cells implies the possible association of OPG with the pathogenesis of HFRS.
      In this study, we determined the concentrations of plasma OPG from patients with HFRS and confirmed the possible sources for the OPG after HTNV infection. We also estimated the relationship between plasma OPG and inflammatory factors, or between plasma OPG and clinical coagulation-fibrinolysis parameters. These findings indicated that plasma OPG might be involved in endothelium injury, inflammation, and coagulation hemostasis in patients with HFRS after HTNV infection.

      Materials and methods

      Patients

      The study was carried out at the Department of Infectious Diseases in Tangdu Hospital of the Fourth Military Medical University (Xi'an, China) from October 2020 to January 2021. A total of 32 patients confirmed to have HFRS via serological testing of immunoglobulin M (IgM) and IgG in serum specimens were enrolled in the study. According to the diagnostic criteria from the Prevention and Treatment Strategy of HFRS promulgated by the Ministry of Health in the People's Republic of China, the patients were classified into four clinical types (mild, moderate, severe, and critical) (
      • Ma Y
      • Yuan B
      • Zhuang R
      • Zhang Y
      • Liu B
      • Zhang C
      • et al.
      Hantaan virus infection induces both Th1 and ThGranzyme B+ cell immune responses that associated with viral control and clinical outcome in humans.
      ). Mild HFRS was defined as a mild renal failure without an obvious oliguric stage and moderate for obvious symptoms of uremia, effusion (bulbar conjunctiva), hemorrhage (skin and mucous membrane), and renal failure with a typical oliguric stage. Patients with severe uremia, effusion (bulbar conjunctiva and either peritoneum or pleura), hemorrhage (skin and mucous membrane), and renal failure with oliguria (urine output, 50-500 ml/day) for ≤5 days or anuria (urine output, <50 ml/day) for ≤2 days were defined as severe HFRS, and critical patients were considered as those with ≥1 of the following symptoms: refractory shock, visceral hemorrhage, heart failure, pulmonary edema, brain edema, severe secondary infection, and severe renal failure with either oliguria (urine output, 50-500 ml/day) for >5 days, anuria (urine output, <50 ml/day) for >2 days, or a blood urea nitrogen level of >42.84 mmol/l. To ensure the sample size in some statistical analyses, we combined the patients according to the disease severity into mild/moderate and severe/critical groups for comparison. According to the clinical observation, the illness could be divided into acute (febrile, hypotensive, oliguric) and convalescent (diuretic, convalescent) stages. The phase from the fever onset to the early oliguric stage was defined as the acute or early stage of the disease. The patients who had other kidney diseases, cardiovascular diseases, autoimmune diseases, viral hepatitis, and other liver diseases were excluded from this study. Patients who received whole blood transfusions and transfusion of any blood components or blood products before collection were also excluded.
      The detailed characteristics of the patients enrolled in the present study are summarized in Table 1. A control group of 16 healthy volunteers showing anti-HTNV negative or no HTNV risk factors was also enrolled as a negative control group.
      Table 1Clinical characteristics of the study subjects.
      HFRS patients
      TotalMild/moderateSevere/critical
      n = 32n = 9n = 23
      Age (year)
      There was no difference of age and male percent between total HFRS patients and controls or between milder group and more severe group.
      44(49-49)40(24-55)46(40-51)
      Male (%)
      There was no difference of age and male percent between total HFRS patients and controls or between milder group and more severe group.
      75.0073.9177.78
      Course (days)
      The days from fever onset to convalescence discharge from the hospital.
      19(15-22)12(9-16)
      P <0.01 and
      21(17-25)
      P <0.01 and
      Patients with shock (%)18.750
      P <0.05 between mild/moderate group and severe/critical group.
      26.09
      P <0.05 between mild/moderate group and severe/critical group.
      Patients with hemorrhage (%)68.7555.5673.91
      Patients with dialysis (%)68.7533.33
      P <0.01 and
      82.61
      P <0.01 and
      BUNmax (mmol/l)25.86(21.25-30.48)18.00(7.77-28.24)
      P <0.05 between mild/moderate group and severe/critical group.
      28.94(23.98-33.89)
      P <0.05 between mild/moderate group and severe/critical group.
      Creamax (μmol/l)561.47(439.73-683.20)362.63(175.18-550.09)
      P <0.05 between mild/moderate group and severe/critical group.
      639.27(491.66-786.88)
      P <0.05 between mild/moderate group and severe/critical group.
      Leukmax (
      P <0.05 between mild/moderate group and severe/critical group.
      109/l)
      20.59(15.84-25.33)16.00(7.90-24.10)
      P <0.05 between mild/moderate group and severe/critical group.
      22.38(16.40-28.37)
      P <0.05 between mild/moderate group and severe/critical group.
      Thrommin (
      P <0.05 between mild/moderate group and severe/critical group.
      109/l)
      35.94(26.21-45.67)49.44(29.12-69.77)
      P <0.05 between mild/moderate group and severe/critical group.
      30.65(19.47-41.83)
      P <0.05 between mild/moderate group and severe/critical group.
      ALBmin (g/l)25.78(23.56-28.00)29.81(25.64-33.99)
      P <0.05 between mild/moderate group and severe/critical group.
      24.20(21.70-26.70)
      P <0.05 between mild/moderate group and severe/critical group.
      APTTmax (sec)47.33(40.34-54.31)36.73(32.16-41.31)
      P <0.05 between mild/moderate group and severe/critical group.
      51.47(42.27-60.67)
      P <0.05 between mild/moderate group and severe/critical group.
      D-dimermax (mg/l)1.67(1.17-2.17)0.96(0.45-1.46)
      P <0.05 between mild/moderate group and severe/critical group.
      1.95(1.30-2.61)
      P <0.05 between mild/moderate group and severe/critical group.
      NOTE. Continuous variables are presented as the mean (95% confidence interval). The Mann-Whitney U test was used to compare groups; the χ2 test was used for the percentage of male, patients with shock, patients with hemorrhage and patients with dialysis.
      ALB, serum albumin concentration; APTT, activated partial thromboplastin time; BUN, blood urea nitrogen; Crea, serum creatinine concentration; HFRS, hemorrhagic fever with renal syndrome; Leuk, blood leukocyte count; max, maximum; min, minimumn; number of patients; Throm, blood thrombocyte count.
      a There was no difference of age and male percent between total HFRS patients and controls or between milder group and more severe group.
      b The days from fever onset to convalescence discharge from the hospital.
      low asterisklow asterisk P <0.01 and
      low asterisk P <0.05 between mild/moderate group and severe/critical group.

      Plasma and clinical data collection

      Two peripheral blood samples per patient were collected during the hospitalization for the disease. The first was obtained on admission as early as possible, 4-11 (median 6) days after fever onset. The second was obtained at the full recovery phase (ranging 19-39 days after fever onset, median 26 days). The separated plasma was stored at -70℃ until use. Complete blood count and renal function tests were performed daily in all patients as part of routine clinical care. A blood coagulation test was conducted at the Laboratory Centre of the Tangdu Hospital using standard methods.

      Viruses and cell lines

      The HTNV 76-118 strain was kindly provided by the Department of Microbiology of our university. The human umbilical vein endothelial cell (HUVEC) line was stored and provided by Dr. Yusi Zhang in our department. The Vero E6 cell line was stored and provided by Dr. Boquan Jin in our department.

      Human umbilical vein endothelial cell culture and treatments

      Cultures of HUVECs grown on fibronectin-coated plates (Millipore, USA) were maintained in endothelial cell growth medium (Lonza, USA) supplemented with 10% fetal bovine serum (PAA, Austria), 100 IU of penicillin/ml, and 100 μg of streptomycin/ml. In total, four groups of HUVECs with different treatments were carried out, including control, HTNV infection, TNF-α stimulation, and HTNV infection combined with TNF-α stimulation. For all infections, the virus was allowed to adsorb to HUVECs at a multiplicity of infection of approximately 1.0 in serum-free endothelial cell growth medium maintenance medium for 2 hours at 37℃. HTNV was added to cells at a final concentration of tissue culture infectious dose 10−3 (1:50 diluted) and TNF-α was added to cells at different concentrations. HUVECs were cultured for the indicated time. Culture supernatants of the cells were harvested for protein detection.

      ELISA for detecting osteoprotegerin protein, tumor necrosis factor-α and C-reactive protein

      Concentrations of plasma OPG and cell culture supernatants were determined using a commercial ELISA kit (eBioscience, North America). The concentration of plasma c-reactive protein (CRP) and TNF-α were also determined using a commercial CRP ELISA kit (BioVendor, Heidelberg, Germany) and a TNF-α ELISA kit (R&D Systems, Minneapolis, Minessotta).

      Statistical analysis

      Statistical analyses and graphing were performed using SPSS 11.5 (SPSS Inc., Chicago, IL, USA) and Prism software, version 5.0 (GraphPad; La Jolla, California). Means with corresponding 95% CI were calculated for normally distributed data, whereas medians with ranges were determined for data with a skewed distribution. Comparisons between the groups were based on the nonparametric Mann-Whitney U test for the numerical data and the chi-square test for the categorical data. Correlations between various markers were tested by calculating Spearman's correlation coefficient. A two-tailed P-value below 0.05 (P ≤0.05) was statistically significant.

      Results

      Patient characteristics

      The details of clinical parameters during the hospitalization of the patients are shown in Table 1. Notably, the overall hospitalization course and the percentage of patients with shock or with dialysis were higher in the severe/critical group than in the mild/moderate group. The level of maximum levels of blood urea nitrogen (BUNmax), serum creatinine (Creamax), and leukocyte counts (Leukmax) during the period of hospitalization in severe/critical patients were statistically higher compared with the mild/moderate patients, whereas the minimum serum albumin (ALBmin) level and the nadir count of blood thrombocytes (Thrommin) were lower in the severe/critical group compared with the mild/moderate group. For the hemorrhage parameters, the maximum activated partial thromboplastin time (APTTmax) and the maximum concentration of d-dimer were higher in the severe/critical group compared with the mild/moderate group.

      Dramatic elevation of the plasma osteoprotegerin levels in hemorrhagic fever with renal syndrome patients

      We first detected the concentration of plasma OPG in HFRS patients. The mean (95% CI) level of plasma OPG in healthy controls was 0.0086 (0.0036-0.0135) μg/ml. Compared with the healthy controls, the mean (95% CI) level of plasma OPG in patients with HFRS, despite disease severity and stages, was elevated to 2.3640 (1.3907-3.3373) μg/ml (P <0.001). When analyzing the OPG level at different stages of patients with HFRS, the concentration of plasma OPG at the acute stage and convalescence of patients with HFRS were 3.2364 (1.4304-5.0424) μg/ml and 1.4917 (0.7536-2.2298) μg/ml, respectively, both of which were elevated compared with healthy controls (P <0.001) (Figure 1a). In addition, the OPG level at the acute stage was statistically higher compared with the convalescence of HFRS (P-value = 0.005) (Figure 1a). There was a decrease in the tendency of the plasma OPG level from the acute stage to convalescence in most patients with HFRS(Figure 1b). The mean (95% CI) level of plasma OPG was 0.7287 (-0.5680-2.0255) μg/ml in the mild/moderate group and 4.2177 (1.8200-6.6153) μg/ml in the severe/critical group at the acute stage of HFRS. The concentration of plasma OPG in the severe/critical group showed a higher level compared with the mild/moderate group at the acute stage of HFRS (P-value = 0.044) (Figure 1c).
      Figure 1
      Figure 1Comparison of plasma OPG levels in HFRS patients with different disease severities at different stages. (a) Comparison of the plasma OPG concentrations between control group and HFRS patients at an acute stage or at convalescence, or comparison of the plasma OPG concentrations between acute stage and convalescence of HFRS patients. (b) The connection diagram for plasma OPG concentration changes tendency for each HFRS patient from acute stage to convalescence (c) Comparison of the plasma OPG concentrations between control group and mild/moderate HFRS patients or severe/critical HFRS patients, or comparison of acute stage plasma OPG concentrations between mild/moderate and severe/critical HFRS patients. The Mann-Whitney U test was used for comparisons between the groups.
      HFRS, hemorrhagic fever with renal syndrome; OPG, osteoprotegerin.
      (*) P < 0.05, (**) P < 0.01, (***) P < 0.001.

      The concentration of plasma osteoprotegerin is associated with inflammatory levels in patients with HFRS

      Given the importance of inflammatory parameters used as indicators for the severity of HFRS, we next set out to determine whether there was a relationship between OPG level and inflammatory parameters in HFRS patients. Two important inflammatory parameters, TNF-α and acute phase reactant CRP were analyzed (Table 2). Interestingly, the concentrations of plasma OPG appeared to be positively associated with the levels of plasma CRP (P-value = 0.001, r = 0.394) and plasma TNF-α (P-value = 0.003, r = 0.371) (Figures 2a and 2c). Similarly, there were still positive associations between plasma OPG levels and the concentration of plasma CRP (P-value = 0.001, r = 0.556) or plasma TNF-α (P-value = 0.014, r = 0.429) at the acute stage of the disease (Figures 2b and 2d).
      Table 2Comparison of plasma variables in 32 patients with hemorrhagic fever with renal syndrome.
      Normal controlMild/moderateSevere/criticalAcute phaseRecovery phase
      n = 16n = 9n = 23n = 32n = 32
      Osteoprotegerin (mg/l)0.0086 (0.0036-0.0135)0.7287 (-0.5680-2.0255)
      P <0.01 between normal control and mild/moderate, severe/critical group, or between normal control and acute phase, recovery phase.
      ,
      P <0.05 between mild/moderate group and severe/critical group.
      4.2177

      (1.8200-6.6153)
      P <0.01 between normal control and mild/moderate, severe/critical group, or between normal control and acute phase, recovery phase.
      ,
      P <0.05 between mild/moderate group and severe/critical group.
      3.2364 (1.4304-5.0424)
      P <0.01 between normal control and mild/moderate, severe/critical group, or between normal control and acute phase, recovery phase.
      ,
      P <0.01 between acute phase and recovery phase.
      1.4917 (0.7536-2.2298)
      P <0.01 between normal control and mild/moderate, severe/critical group, or between normal control and acute phase, recovery phase.
      ,
      P <0.01 between acute phase and recovery phase.
      Tumor necrosis factor-α (ng/l)44.29

      (36.39-52.19)
      83.82

      (66.90-100.74)
      P <0.01 between normal control and mild/moderate, severe/critical group, or between normal control and acute phase, recovery phase.
      ,
      P <0.05 between mild/moderate group and severe/critical group.
      105.94

      (94.97-116.91)
      P <0.01 between normal control and mild/moderate, severe/critical group, or between normal control and acute phase, recovery phase.
      ,
      P <0.05 between mild/moderate group and severe/critical group.
      99.72

      (90.29-109.15)
      P <0.01 between normal control and mild/moderate, severe/critical group, or between normal control and acute phase, recovery phase.
      ,
      P <0.01 between acute phase and recovery phase.
      70.74

      (61.70-79.78)
      P <0.01 between normal control and mild/moderate, severe/critical group, or between normal control and acute phase, recovery phase.
      ,
      P <0.01 between acute phase and recovery phase.
      C-reactive protein (mg/l)8.83

      (-2.87-20.53)
      30.50

      (12.37-48.63)
      P <0.01 between normal control and mild/moderate, severe/critical group, or between normal control and acute phase, recovery phase.
      ,
      P <0.05 between mild/moderate group and severe/critical group.
      58.58

      (43.64-73.52)
      P <0.01 between normal control and mild/moderate, severe/critical group, or between normal control and acute phase, recovery phase.
      ,
      P <0.05 between mild/moderate group and severe/critical group.
      50.68

      (34.42-62.93)
      P <0.01 between normal control and mild/moderate, severe/critical group, or between normal control and acute phase, recovery phase.
      ,
      P <0.01 between acute phase and recovery phase.
      19.45

      (11.70-27.19)
      P <0.01 between normal control and mild/moderate, severe/critical group, or between normal control and acute phase, recovery phase.
      ,
      P <0.01 between acute phase and recovery phase.
      NOTE. Continuous variables are presented as the mean (95% confidence interval). The Mann-Whitney U test was used to compare groups. n, number of plasma samples. The comparison between mild/moderate group and severe/critical group were acute phase samples.
      a P <0.01 between normal control and mild/moderate, severe/critical group, or between normal control and acute phase, recovery phase.
      b P <0.05 between mild/moderate group and severe/critical group.
      c P <0.01 between acute phase and recovery phase.
      Figure 2
      Figure 2The relationships between the concentration of plasma OPG and inflammatory indicators in HFRS patients. (a) Analysis of the relationship between the level of plasma OPG and CRP in HFRS patients at both acute stage and convalescence. (b) Analysis of the relationship between the level of plasma OPG and CRP at acute stage of HFRS patients. (c) Analysis of the relationship between the level of plasma OPG and TNF-α in HFRS patients at both the acute stage and convalescence. (d) Analysis of the relationship between the level of plasma OPG and TNF-α at acute stage of HFRS patients. Spearman's correlation coefficient analysis was used for the statistical tests.
      CRP, C-reactive protein; HFRS, hemorrhagic fever with renal syndrome; OPG, osteoprotegerin; TNF-α, tumor necrosis factor-α.

      The synergistic effect of Hantaan virus infection and tumor necrosis factor-α on osteoprotegerin production from human umbilical vein endothelial cell

      Next, we selected HUVEC as the HTNV infection model to ascertain whether the elevation of OPG production would be modulated by HTNV infection and proinflammatory cytokines. First, the concentration of OPG in the supernatant of HUVEC was detected at different stimulating time points. As shown in Figure 3a, at each stimulating time point, the highest level of OPG in the supernatant of HUVEC was always observed in the group stimulated with 10 ng/ml TNF-α combined with HTNV infection. Compared with the control group with no stimulation, there was a slight increase of OPG level in HUVEC supernatant in the single HTNV infection group and an obvious elevation of OPG level in the 10 ng/ml single TNF-α stimulation group or in 10 ng/ml TNF-α stimulation combined with HTNV infection group. Importantly, as the stimulating time prolonged from 12 hours to 48 hours, an apparent increasing trend of the concentration of OPG in the supernatant was observed in all three groups with different stimulations. We further compared the concentration of OPG in HUVEC supernatant at fixed 48 hours of stimulation under the different concentrations of TNF-α. As expected, there was an evident increasing trend of the OPG level in HUVEC supernatant detected from 2 ng/ml to 20 ng/ml of TNF-α stimulation in single TNF-α stimulation group and TNF-α stimulation combined with the HTNV infection group. Notably, when the concentration of TNF-α was increased to 10 ng/ml or 20 ng/ml, the OPG level in TNF-α stimulation combined with the HTNV infection group was higher compared with a single TNF-α stimulation group (Figure 3b), suggesting that inflammatory condition after HTNV infection could synergize with HTNV infection to promote OPG secretion from endothelial cells, which might be one of the main sources of OPG production.
      Figure 3
      Figure 3Detection of OPG production from cultured HUVEC under different HTNV infection or TNF-α stimuli conditions. (a) Detection of the change tendency of OPG concentration in the supernatant of HUVEC at 12 hours, 24 hours and 48 hours stimulating time points in four different treatment groups including control group, single HTNV infection group, single 10 ng/ml TNF-α stimulation group and HTNV infection combined with 10 ng/ml TNF-α stimulation group, respectively. (b) Detection of the change tendency of OPG concentration in the supernatant of HUVEC in single TNF-α stimulation group and HTNV infection combined with TNF-α stimulation group after 48 hours at the TNF-α concentration of 2 ng/ml, 5 ng/ml, 10 ng/ml and 20 ng/ml.
      HUVEC, human umbilical vein endothelial cell; HTNV, Hantaan virus; OPG, osteoprotegerin; TNF-α, tumor necrosis factor-α.

      The inverse relationship between platelet number and osteoprotegerin level in patients with HFRS

      The platelet number of the patients with HFRS could be used as a hallmark for disease severity. As shown in Figure 4a, the concentration of plasma OPG was inversely associated with the number of platelets detected at the same plasma collection time point (P-value = 0.002, r = -0.380). Furthermore, the concentration of plasma OPG at the acute stage of HFRS also showed an inverse association with the nadir of platelets (P <0.001, r = -0.592) (Figure 4b).
      Figure 4
      Figure 4Analysis of the associations between plasma OPG concentration and number of platelets in HFRS patients. (a) The associations between the concentration of plasma OPG and the number of platelets at the same day of each blood collection during hospitalization in HFRS patients. (b) The associations between the concentration of plasma OPG at acute stage HFRS and the nadir number of platelets during hospitalization in HFRS patients. Spearman's correlation coefficient analysis was used for the statistical tests.
      HFRS, hemorrhagic fever with renal syndrome; OPG, osteoprotegerin.

      Positive relationships between coagulation-fibrinolysis indicators and osteoprotegerin level in patients with HFRS

      We then analyzed the relationships between plasma OPG and certain coagulation-fibrinolysis factors to better assess the possible effects of OPG on the hemorrhage state of patients with HFRS. The prolonged APTT and elevated level of d-dimer were the typical laboratory parameters of HFRS disease and reflect the abnormal function of coagulation and fibrinolysis, respectively (
      • Du H
      • Wang PZ
      • Li J
      • Bai L
      • Li H
      • Yu HT
      • et al.
      Clinical characteristics and outcomes in critical patients with hemorrhagic fever with renal syndrome.
      ). Importantly, the concentrations of plasma OPG in acute-stage HFRS patients positively correlated with the maximum APTT (P-value = 0.013, r = 0.434) (Figure 5a), as well as the maximum levels of plasma d-dimer (P-value = 0.003, r = 0.510) (Figure 5b).
      Figure 5
      Figure 5Associations between concentration of plasma OPG and coagulation indicator or fibrinolysis indicator in HFRS patients. (a) The association between the concentration of plasma OPG at the acute stage of HFRS and the maximum APTT during hospitalization in HFRS patients. (b) The associations between the concentration of plasma OPG at acute stage HFRS and the maximum levels of d-dimer during hospitalization in HFRS patients. Spearman's correlation coefficient analysis was used for the statistical tests.
      APTT, prolonged activated partial thromboplastin time; HFRS, hemorrhagic fever with renal syndrome; OPG, osteoprotegerin.

      Discussion

      In this study, we showed that plasma OPG levels at the acute stage of HFRS were significantly higher in severe/critical patients when compared with patients with mild/moderate HFRS. Moreover, the inflammatory conditions after HTNV infection could synergize with HTNV infection to promote OPG production from endothelial cells. Importantly, the elevated plasma OPG level was negatively associated with the platelet count, while positively correlated with coagulation-fibrinolysis indicators. These findings suggested that increased plasma OPG levels may be an important factor for the disease severity of HFRS, which might be involved in the pathogenesis of hemorrhage in HFRS.
      The OPG could bind to TRAIL and inhibit the TRAIL-induced proapoptotic signaling pathway (
      • Soto-Gamez A
      • Wang Y
      • Zhou X
      • Seras L
      • Quax W
      • Demaria M.
      Enhanced extrinsic apoptosis of therapy-induced senescent cancer cells using a death receptor 5 (DR5) selective agonist.
      ). A recent study has shown that TRAIL treatment could significantly reduce HTNV viral load, alleviate virus-induced tissue lesions and increase apoptotic cells of suckling mice, while TRAIL interference could inhibit apoptosis of HUVECs and enhance the production of HTNV as well as reduce interferon-β production (
      • Chen QZ
      • Wang X
      • Luo F
      • Li N
      • Zhu N
      • Lu S
      • et al.
      HTNV sensitizes host toward TRAIL-mediated apoptosis-a pivotal anti-hantaviral role of TRAIL.
      ). Here, we found that plasma OPG levels at the acute stage of HFRS were significantly higher in severe/critical patients compared with patients with mild/moderate HFRS, and the increased plasma OPG might inhibit the proapoptotic function of TRAIL and enhance the production of HTNV to induce more serious HFRS. Increased OPG predicting poor virological outcomes during anti-cytomegalovirus therapy also indicates that OPG might inhibit antiviral responses (
      • Ueland T
      • Rollag H
      • Hartmann A
      • Jardine A
      • Humar A
      • Bignamini AA
      • et al.
      Increased osteoprotegerin predicts poor virological outcome during anticytomegalovirus therapy in solid organ transplant recipients.
      ). In addition, OPG has ligand-independent effects, such as the direct actions on the endothelium, mediated by the interaction of its heparin-binding domain with cellular heparin sulfate proteoglycans (
      • Bernardi S
      • Bossi F
      • Toffoli B
      • Fabris B.
      Roles and clinical applications of OPG and TRAIL as biomarkers in cardiovascular disease.
      ). Therefore, OPG has been considered a molecular biomarker for several disease statuses, such as cardiovascular disease and infectious diseases (
      • Bernardi S
      • Bossi F
      • Toffoli B
      • Fabris B.
      Roles and clinical applications of OPG and TRAIL as biomarkers in cardiovascular disease.
      ;
      • Pichler G
      • Haller MC
      • Kainz A
      • Wolf M
      • Redon J
      • Oberbauer R.
      Prognostic value of bone- and vascular-derived molecular biomarkers in hemodialysis and renal transplant patients: a systematic review and meta-analysis.
      ). In patients with malaria, elevated plasma OPG was considered a biomarker of endothelial activation and microvascular dysfunction (
      • Woodford J
      • Yeo TW
      • Piera KA
      • Butler K
      • Weinberg JB
      • McCarthy JS
      • et al.
      Early endothelial activation precedes glycocalyx degradation and microvascular dysfunction in experimentally induced Plasmodium falciparum and Plasmodium vivax infection.
      ). The increased circulating OPG levels were also found in severe dengue fever and COVID-19 patients, which was likely to contribute to thrombocytopenia and activation of endothelial cells (
      • Djamiatun K
      • van der Ven AJ
      • de Groot PG
      • Faradz SM
      • Hapsari D
      • Dolmans WM
      • et al.
      Severe dengue is associated with consumption of von Willebrand factor and its cleaving enzyme ADAMTS-13.
      ;
      • Fogarty H
      • Ward SE
      • Townsend L
      • Karampini E
      • Elliott S
      • Conlon N
      • et al.
      Sustained VWF-ADAMTS-13 axis imbalance and endotheliopathy in long COVID syndrome is related to immune dysfunction.
      ). Hantaviruses infect endothelial cells and subsequently cause direct injury. The hemorrhage in HFRS patients may be a consequence of vascular injury, a deficit of functional platelets, and sometimes disseminated intravascular coagulation (
      • Vaheri A
      • Strandin T
      • Hepojoki J
      • Sironen T
      • Henttonen H
      • Mäkelä S
      • et al.
      Uncovering the mysteries of hantavirus infections.
      ). It is possible that OPG would have effects on endothelium dysfunction and the pathogenesis of coagulation disorder after HTNV infection.
      The finding that abundant OPG expression could be detected in HUVEC further supported our study on OPG production using HUVEC as an infection model (
      • Zannettino AC
      • Holding CA
      • Diamond P
      • Atkins GJ
      • Kostakis P
      • Farrugia A
      • et al.
      Osteoprotegerin (OPG) is localized to the Weibel-Palade bodies of human vascular endothelial cells and is physically associated with von Willebrand factor.
      ). We showed that endothelial cells might be one of the major contributors to the elevated plasma OPG after HTNV infection in HFRS. In endothelial cells, OPG is associated with von Willebrand factor (VWF) and localized in the Weibel-Palade bodies. Notably, VWF is a marker for endothelial activation and is involved in blood coagulation through binding with platelets (
      • Zannettino AC
      • Holding CA
      • Diamond P
      • Atkins GJ
      • Kostakis P
      • Farrugia A
      • et al.
      Osteoprotegerin (OPG) is localized to the Weibel-Palade bodies of human vascular endothelial cells and is physically associated with von Willebrand factor.
      ), the increased OPG exposed the binding site for VWF, which could subsequently prevent the combination of VWF to platelets. As a result, OPG acts to regulate the hemostatic role of VWF (
      • Wohner N
      • Sebastian S
      • Muczynski V
      • Huskens D
      • de Laat B
      • de Groot PG
      • et al.
      Osteoprotegerin modulates platelet adhesion to von Willebrand factor during release from endothelial cells.
      ). Elevated plasma OPG may therefore be an important inducer of bleeding during early HFRS.
      Although elevated plasma OPG indicated that OPG might be involved in the disorder of endothelium in patients with HFRS, subsequent studies showed that elevated OPG was closely related to inflammation after HTNV infection. Previous studies showed that levels of proinflammatory cytokines started to increase at the acute stage during HFRS (
      • Guo J
      • Guo X
      • Wang Y
      • Tian F
      • Luo W
      • Zou Y.
      Cytokine response to Hantaan virus infection in patients with hemorrhagic fever with renal syndrome.
      ). Inflammatory cytokines are known endothelium agonists (
      • Hepojoki J
      • Vaheri A
      • Strandin T.
      The fundamental role of endothelial cells in hantavirus pathogenesis.
      ). We found that OPG in HUVEC could rapidly release in response to inflammatory stimuli and HTNV infection, which was in accordance with the previous reports that OPG could be translocated and externalized to the extracellular environment in response to proinflammatory cytokines, such as TNF-α (
      • Zannettino AC
      • Holding CA
      • Diamond P
      • Atkins GJ
      • Kostakis P
      • Farrugia A
      • et al.
      Osteoprotegerin (OPG) is localized to the Weibel-Palade bodies of human vascular endothelial cells and is physically associated with von Willebrand factor.
      ).
      Notably, positive associations between plasma OPG levels and TNF-α or CRP in patients with HFRS were found in our study. OPG delivery could increase the circulating levels of interleukin-6 and TNF-α (
      • Bernardi S
      • Fabris B
      • Thomas M
      • Toffoli B
      • Tikellis C
      • Candido R
      • et al.
      Osteoprotegerin increases in metabolic syndrome and promotes adipose tissue proinflammatory changes.
      ) as well as the expression of adhesion molecules on endothelial cells and leukocyte adhesion to endothelial cells (
      • Mangan SH
      • Van Campenhout A
      • Rush C
      • Golledge J.
      Osteoprotegerin upregulates endothelial cell adhesion molecule response to tumor necrosis factor-alpha associated with induction of angiopoietin-2.
      ), which demonstrated that OPG has proinflammatory and profibrotic effects in vasculature. Therefore, it is possible that OPG could damage the endothelium by promoting inflammation and fibrosis (
      • Bernardi S
      • Bossi F
      • Toffoli B
      • Fabris B.
      Roles and clinical applications of OPG and TRAIL as biomarkers in cardiovascular disease.
      ). The interactions between OPG and RANKL or TRAIL are relevant to inflammatory pathways. OPG could be regulated by inflammatory signals, such as NF-κB (
      • Toruner M
      • Fernandez-Zapico M
      • Sha JJ
      • Pham L
      • Urrutia R
      • Egan LJ.
      Antianoikis effect of nuclear factor-kappaB through up-regulated expression of osteoprotegerin, BCL-2, and IAP-1.
      ). High levels of inflammatory cytokines contribute to elevated OPG levels in patients with HIV-1 (
      • Gallego-Escuredo JM
      • Lamarca MK
      • Villarroya J
      • Domingo JC
      • Mateo MG
      • Gutierrez MDM
      • et al.
      High FGF21 levels are associated with altered bone homeostasis in HIV-1-infected patients.
      ). Proinflammatory cytokines, such as TNF-α, could induce the synthesis of OPG from immune cells or endothelium. OPG may, in turn, serve to modulate the extent of the inflammatory response by binding to RANKL and TRAIL.
      The mechanisms that lead to severe forms of HFRS are complex but undoubtedly related to increased coagulation and fibrinolysis activity during hantaviruses infection (
      • Jiang H
      • Du H
      • Wang LM
      • Wang PZ
      • Bai XF.
      Hemorrhagic fever with renal syndrome: pathogenesis and clinical picture.
      ). Patients with HFRS exhibit various degrees of abnormal hematologic indices, among which, low platelets, prolonged APTT, and thrombin time are indicators closely correlated with hemorrhage and coagulopathy. Notably, platelet counts begin to fall during the febrile stage of patients with HFRS. The increased platelet consumption from the interaction between platelets and HTNV-infected endothelial cells and the defective platelet activation after HTNV infection were confirmed to be the reasons for acute thrombocytopenia (
      • Cosgriff TM
      • Lee HW
      • See AF
      • Parrish DB
      • Moon JS
      • Kim DJ
      • et al.
      Platelet dysfunction contributes to the haemostatic defect in haemorrhagic fever with renal syndrome.
      ). Under physiological conditions, the active conformation of VWF could bind with platelets to take part in platelet aggregation and adhesion to the injured endothelium during primary hemostasis and pathological thrombus formation (
      • Ruggeri ZM
      • Mendolicchio GL.
      Interaction of von Willebrand factor with platelets and the vessel wall.
      ). However, it was found that OPG could bind selectively to the A1 domain of VWF to interfere with the binding of platelets to activated VWF, preventing platelet adhesion and aggregation (
      • Lenting PJ
      • Pegon JN
      • Groot E
      • de Groot PG.
      Regulation of von Willebrand factor-platelet interactions.
      ;
      • Shahbazi S
      • Lenting PJ
      • Fribourg C
      • Terraube V
      • Denis CV
      • Christophe OD.
      Characterization of the interaction between von Willebrand factor and osteoprotegerin.
      ). Moreover, OPG may aid in tethering VWF multimers at the site of vascular injury through binding with thrombospondin-1, the reductase of VWF, thereby participating in thrombus formation and facilitating the proteolysis of the endothelium (
      • Zannettino AC
      • Holding CA
      • Diamond P
      • Atkins GJ
      • Kostakis P
      • Farrugia A
      • et al.
      Osteoprotegerin (OPG) is localized to the Weibel-Palade bodies of human vascular endothelial cells and is physically associated with von Willebrand factor.
      ). In addition to the inverse association between OPG level and platelet count, we also observed prolonged APTT and increased concentration of d-dimer in patients with HFRS. The APTT reflects the function of an endogenous coagulative pathway, and d-dimer is an indirect marker for fibrinolysis. The activation of coagulation pathways and fibrinolysis have also been reported previously in dengue virus infection (
      • de Azeredo EL
      • Monteiro RQ
      • de-Oliveira Pinto LM
      Thrombocytopenia in dengue: interrelationship between virus and the imbalance between coagulation and fibrinolysis and inflammatory mediators.
      ). In Puumala virus-infected nephropathia epidemica patients, the disseminated intravascular coagulation score was not only based on platelet count but also based on prothrombin time and concentrations of d-dimer and fibrinogen (
      • Laine O
      • Mäkelä S
      • Mustonen J
      • Huhtala H
      • Szanto T
      • Vaheri A
      • et al.
      Enhanced thrombin formation and fibrinolysis during acute Puumala hantavirus infection.
      ). Based on our result that a positive association was found between OPG level and APTT or the concentration of d-dimer in patients with HFRS, it is tempting to speculate that the elevated plasma OPG may play important roles in coagulation disorders and thrombus formation in patients with HFRS after HTNV infection.
      The major limitation of our study was that only a small number of patients from a single center were recruited for detection. For the limited samples, the continuous dynamic changes of plasma OPG levels during HFRS and the time point at which OPG may play a more important role were not obtained in this study. Although our results showed obvious elevated OPG levels during the acute stage of HFRS, especially in patients with severe/critical disease, further investigation including a large number of samples for the role of OPG during HTNV infection will still be needed.

      Conclusion

      Our results might provide new insights into understanding the pathogenesis of hemorrhage in HFRS after HTNV infection. However, the only research on the increased OPG levels and the relationships with clinical indicators is just a fraction of the research work. Further studies about the functional mechanisms of elevated plasma OPG used as a soluble biomarker reflecting the abnormal coagulation in HFRS need to be investigated further to make a comprehensive explanation of the coagulopathy of HFRS after HTNV infection.

      Declaration of competing interest

      The authors have no competing interests to declare.

      Funding

      This work was supported by the National Natural Science Foundation of China, grant number 81871239; Technical Field of Foundation Strengthening Plan Projects, grant number 2019-JCJQ-JJ-094; and National Natural Science Foundation of China, grant number 81771705 and 81901600.

      Ethical approval

      The study was conducted in accordance with the Declaration of Helsinki, and approved by the Institutional Review Board of the Tangdu Hospital and the Fourth Military Medical University. The research involving humans and human materials was also approved by the Ethical Review Board of the Fourth Military Medical University (KY20183312-1), and the related data were anonymized before use.

      Declaration of interests

      The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

      References

        • Bernardi S
        • Fabris B
        • Thomas M
        • Toffoli B
        • Tikellis C
        • Candido R
        • et al.
        Osteoprotegerin increases in metabolic syndrome and promotes adipose tissue proinflammatory changes.
        Mol Cell Endocrinol. 2014; 394: 13-20
        • Bernardi S
        • Bossi F
        • Toffoli B
        • Fabris B.
        Roles and clinical applications of OPG and TRAIL as biomarkers in cardiovascular disease.
        BioMed Res Int. 2016; 20161752854
        • Brocato RL
        • Hooper JW.
        Progress on the prevention and treatment of hantavirus disease.
        Viruses. 2019; 11: 610
        • Chen QZ
        • Wang X
        • Luo F
        • Li N
        • Zhu N
        • Lu S
        • et al.
        HTNV sensitizes host toward TRAIL-mediated apoptosis-a pivotal anti-hantaviral role of TRAIL.
        Front Immunol. 2020; 11: 1072
        • Cosgriff TM
        • Lee HW
        • See AF
        • Parrish DB
        • Moon JS
        • Kim DJ
        • et al.
        Platelet dysfunction contributes to the haemostatic defect in haemorrhagic fever with renal syndrome.
        Trans R Soc Trop Med Hyg. 1991; 85: 660-663
        • de Azeredo EL
        • Monteiro RQ
        • de-Oliveira Pinto LM
        Thrombocytopenia in dengue: interrelationship between virus and the imbalance between coagulation and fibrinolysis and inflammatory mediators.
        Mediators Inflamm. 2015; 2015313842
        • Djamiatun K
        • van der Ven AJ
        • de Groot PG
        • Faradz SM
        • Hapsari D
        • Dolmans WM
        • et al.
        Severe dengue is associated with consumption of von Willebrand factor and its cleaving enzyme ADAMTS-13.
        PLoS Negl Trop Dis. 2012; 6: e1628
        • Du H
        • Wang PZ
        • Li J
        • Bai L
        • Li H
        • Yu HT
        • et al.
        Clinical characteristics and outcomes in critical patients with hemorrhagic fever with renal syndrome.
        BMC Infect Dis. 2014; 14: 191
        • Fogarty H
        • Ward SE
        • Townsend L
        • Karampini E
        • Elliott S
        • Conlon N
        • et al.
        Sustained VWF-ADAMTS-13 axis imbalance and endotheliopathy in long COVID syndrome is related to immune dysfunction.
        J Thromb Haemost. 2022; 20: 2429-2438
        • Gallego-Escuredo JM
        • Lamarca MK
        • Villarroya J
        • Domingo JC
        • Mateo MG
        • Gutierrez MDM
        • et al.
        High FGF21 levels are associated with altered bone homeostasis in HIV-1-infected patients.
        Metabolism. 2017; 71: 163-170
        • Guo J
        • Guo X
        • Wang Y
        • Tian F
        • Luo W
        • Zou Y.
        Cytokine response to Hantaan virus infection in patients with hemorrhagic fever with renal syndrome.
        J Med Virol. 2017; 89: 1139-1145
        • Hepojoki J
        • Vaheri A
        • Strandin T.
        The fundamental role of endothelial cells in hantavirus pathogenesis.
        Front Microbiol. 2014; 5: 727
        • Jiang H
        • Du H
        • Wang LM
        • Wang PZ
        • Bai XF.
        Hemorrhagic fever with renal syndrome: pathogenesis and clinical picture.
        Front Cell Infect Microbiol. 2016; 6: 1
        • Jiang H
        • Zheng X
        • Wang L
        • Du H
        • Wang P
        • Bai X.
        Hantavirus infection: a global zoonotic challenge.
        Virol Sin. 2017; 32: 32-43
        • Laine O
        • Mäkelä S
        • Mustonen J
        • Huhtala H
        • Szanto T
        • Vaheri A
        • et al.
        Enhanced thrombin formation and fibrinolysis during acute Puumala hantavirus infection.
        Thromb Res. 2010; 126: 154-158
        • Lenting PJ
        • Pegon JN
        • Groot E
        • de Groot PG.
        Regulation of von Willebrand factor-platelet interactions.
        Thromb Haemost. 2010; 104: 449-455
        • Ma Y
        • Yuan B
        • Zhuang R
        • Zhang Y
        • Liu B
        • Zhang C
        • et al.
        Hantaan virus infection induces both Th1 and ThGranzyme B+ cell immune responses that associated with viral control and clinical outcome in humans.
        PLoS Pathog. 2015; 11e1004788
        • Mangan SH
        • Van Campenhout A
        • Rush C
        • Golledge J.
        Osteoprotegerin upregulates endothelial cell adhesion molecule response to tumor necrosis factor-alpha associated with induction of angiopoietin-2.
        Cardiovasc Res. 2007; 76: 494-505
        • Noack D
        • Goeijenbier M
        • Reusken CBEM
        • Koopmans MPG
        • Rockx BHG.
        Orthohantavirus pathogenesis and cell tropism.
        Front Cell Infect Microbiol. 2020; 10: 399
        • Pichler G
        • Haller MC
        • Kainz A
        • Wolf M
        • Redon J
        • Oberbauer R.
        Prognostic value of bone- and vascular-derived molecular biomarkers in hemodialysis and renal transplant patients: a systematic review and meta-analysis.
        Nephrol Dial Transplant. 2017; 32: 1566-1578
        • Rochette L
        • Meloux A
        • Rigal E
        • Zeller M
        • Cottin Y
        • Vergely C.
        The role of osteoprotegerin and its ligands in vascular function.
        Int J Mol Sci. 2019; 20: 705
        • Ruggeri ZM
        • Mendolicchio GL.
        Interaction of von Willebrand factor with platelets and the vessel wall.
        Hamostaseologie. 2015; 35: 211-224
        • Shahbazi S
        • Lenting PJ
        • Fribourg C
        • Terraube V
        • Denis CV
        • Christophe OD.
        Characterization of the interaction between von Willebrand factor and osteoprotegerin.
        J Thromb Haemost. 2007; 5: 1956-1962
        • Soto-Gamez A
        • Wang Y
        • Zhou X
        • Seras L
        • Quax W
        • Demaria M.
        Enhanced extrinsic apoptosis of therapy-induced senescent cancer cells using a death receptor 5 (DR5) selective agonist.
        Cancer Lett. 2022; 525: 67-75
        • Toruner M
        • Fernandez-Zapico M
        • Sha JJ
        • Pham L
        • Urrutia R
        • Egan LJ.
        Antianoikis effect of nuclear factor-kappaB through up-regulated expression of osteoprotegerin, BCL-2, and IAP-1.
        J Biol Chem. 2006; 281: 8686-8696
        • Udagawa N
        • Koide M
        • Nakamura M
        • Nakamichi Y
        • Yamashita T
        • Uehara S
        • et al.
        Osteoclast differentiation by RANKL and OPG signaling pathways.
        J Bone Miner Metab. 2021; 39: 19-26
        • Ueland T
        • Rollag H
        • Hartmann A
        • Jardine A
        • Humar A
        • Bignamini AA
        • et al.
        Increased osteoprotegerin predicts poor virological outcome during anticytomegalovirus therapy in solid organ transplant recipients.
        Transplantation. 2015; 99: 100-105
        • Vaheri A
        • Strandin T
        • Hepojoki J
        • Sironen T
        • Henttonen H
        • Mäkelä S
        • et al.
        Uncovering the mysteries of hantavirus infections.
        Nat Rev Microbiol. 2013; 11: 539-550
        • Wohner N
        • Sebastian S
        • Muczynski V
        • Huskens D
        • de Laat B
        • de Groot PG
        • et al.
        Osteoprotegerin modulates platelet adhesion to von Willebrand factor during release from endothelial cells.
        J Thromb Haemost. 2022; 20: 755-766
        • Woodford J
        • Yeo TW
        • Piera KA
        • Butler K
        • Weinberg JB
        • McCarthy JS
        • et al.
        Early endothelial activation precedes glycocalyx degradation and microvascular dysfunction in experimentally induced Plasmodium falciparum and Plasmodium vivax infection.
        Infect Immun. 2020; 88: e00819-e00895
        • Yi J
        • Xu Z
        • Zhuang R
        • Wang J
        • Zhang Y
        • Ma Y
        • et al.
        Hantaan virus RNA load in patients having hemorrhagic fever with renal syndrome: correlation with disease severity.
        J Infect Dis. 2013; 207: 1457-1461
        • Zannettino AC
        • Holding CA
        • Diamond P
        • Atkins GJ
        • Kostakis P
        • Farrugia A
        • et al.
        Osteoprotegerin (OPG) is localized to the Weibel-Palade bodies of human vascular endothelial cells and is physically associated with von Willebrand factor.
        J Cell Physiol. 2005; 204: 714-723
        • Zhang R
        • Liu J
        • Yu S
        • Sun D
        • Wang X
        • Fu J
        • et al.
        Osteoprotegerin (OPG) promotes recruitment of endothelial progenitor cells (EPCs) via CXCR4 signaling pathway to improve bone defect repair.
        Med Sci Monit. 2019; 25: 5572-5579