Reduced circulating BMP9 and pBMP10 in hospitalized COVID‐19 patients

Abstract Similar to other causes of acute respiratory distress syndrome, coronavirus disease 2019 (COVID‐19) is characterized by the aberrant expression of vascular injury biomarkers. We present the first report that circulating plasma bone morphogenetic proteins (BMPs), BMP9 and pBMP10, involved in vascular protection, are reduced in hospitalized patients with COVID‐19.


INTRODUCTION
The rapid spread of a severe acute respiratory syndrome coronavirus (SARS-CoV-2), first identified in 2019, led to a major worldwide public health crisis. For many individuals, coronavirus disease 2019 (COVID-19) led to no or mild symptoms. However, for some patients, severe COVID-19 resulted in respiratory failure, admission to intensive care, the requirement for mechanical ventilation, and death. According to the World Health Organization (WHO), infection with SARS-CoV-2 was observed in over 600 million individuals worldwide and accounted for more than 6 million deaths by October 2022.
Originally, SARS-CoV-2 was reported to directly infect pulmonary endothelial cells (ECs) via the angiotensin-converting enzyme 2 receptors with sustained infection resulting in the destruction of ECs and vascular leak, causing tissue edema and thromboinflammation. 1 Patients with severe COVID-19 often present with refractory hypoxemia arising from EC damage and pulmonary vasculature dysfunction. Severe cases have been associated with the upregulation of prothrombotic/ inflammatory proteins. Two members of the transforming growth factor-β family, bone morphogenetic protein 9 (BMP9) and BMP10, are recognized as vascular quiescence factors that guard against endothelial dysfunction. We have previously established that BMP9 protects against excess endothelial permeability associated with pulmonary arterial hypertension. 2 Furthermore, BMP9 administration protected mice from lung injury and vascular permeability in a murine experimental model of acute lung injury (ALI). In addition, plasma BMP9 concentrations were shown to be markedly reduced in both patients with sepsis and endotoxemic mice. 3 Given that COVID-19 is a cause of acute respiratory distress syndrome (ARDS), we investigated circulating concentrations of BMP9 and prodomain-BMP10 (pBMP10) in a cohort of patients with COVID-19.   Table 1). Healthy controls were designated 0. Blood samples were drawn into EDTA blood tubes (BD Biosciences) upon study enrollment and plasma samples were processed on the same day as receipt. Acellular banked plasma aliquots were stored at −80°C. A stock aliquot per participant was then defrosted and subaliquoted for this study, meaning the plasma provided had one freeze-thaw cycle. Enzyme-linked immunosorbent assays (ELISAs) for BMP9 and pBMP10 were conducted as previously described. 5,6 The soluble endoglin Quantikine assay (R&D Systems) was performed according to the manufacturer's instructions. A range of angiogenesis and vascular injury biomarkers were measured using three MesoScale Discovery multiplex immunoassay panels following the manufacturer's protocol:

RESULTS
We confirmed many of the vascular injury and proangiogenesis markers previously reported to be associated with disease severity and mortality in COVID-19. First, SAA and CRP were significantly elevated in all patients with COVID-19 (Figure 1a and b, respectively).
We also assessed the circulating concentrations of proangiogenesis markers that have been reported as predictive of COVID-19 disease and severity. 9,10 VEGF pathway activation is associated with ARDS. Both nonventilated (4, 5, 6) and ventilated patients (7,8,9) had elevated plasma concentrations of VEGF-A ( Figure 1g). Increased VEGFR-1/Flt-1 has previously been reported to correlate with disease severity, but in this cohort both nonventilated (4, 5, 6) and ventilated (7,8,9) patients had significantly increased concentrations of the VEGF receptor, VEGFR-1/Flt-1 (Figure 1h). 11 PlGF release was increased in both nonventilated (4,5,6) and ventilated (7,8,9) patients (Figure 1i), and FGF (basic) was significantly increased in ventilated patients ( Figure 1j). The plasma concentration of the vessel maturity receptor, Tie-2 (Tek), 12 was markedly decreased in ventilated patients compared to healthy controls and nonventilated patients (Figure 1k). Interestingly, pro-BMP9 treatment in a murine model of ALI increased the transcriptional expression of Tek. 3 Soluble endoglin (sEng) is associated with inflammation/endothelial dysfunction, 13 and increased sEng levels have been reported in people who do not survive COVID-19 infection. 14 Surprisingly, sEng concentrations in this cohort were decreased in both nonventilated (4, 5, 6) and ventilated (7,8,9) subjects (Figure 1l). A reported increase of sEng was observed in nonsurvivors of COVID-19. 14 In fact, sEng concentrations in survivors increased 14 days after study inclusion but were actually lower in patients than healthy controls at Day 0. 14 Our plasma samples were collected as close to hospital admission as possible, and possibly closer to the initial infection. Therefore, secretion of sEng may increase over the duration of COVID-19, and may be associated with sustained endothelial dysfunction in patients with persistent illness. The data from our cohort expands the previous evidence that biomarkers associated with vascular injury and EC dysfunction are increased in COVID-19. Given the clear association between COVID-19 severity and inflammation/endothelial dysfunction, we hypothesized that endothelial-selective BMP ligands may be novel biomarkers for endothelial injury in COVID-19. BMP9 plasma concentrations were significantly decreased in only nonventilated (4, 5, 6) patients compared to ventilated (7,8,9) patients and healthy controls (Figure 1m). Similarly, circulating pBMP10 concentrations were reduced in nonventilated patients ( Figure 1n). As previously reported, there was a strong correlation between plasma BMP9 and pBMP10 concentrations ( Figure 1o). 6 This is the first report that circulating concentrations of BMP9 and pBMP10 are decreased during COVID-19. This is perhaps unsurprising given the association of inflammation and vascular dysfunction with BMP9 and pBMP10. In fact, BMP9 and pBMP10 have been shown to inhibit chemokine (C-C motif) ligand 2 secretion by vascular ECs, while endogenous circulating BMP9 is elevated in inflammation. 3,15 Induced endotoxemia by lipopolysaccharide (LPS) treatment revealed that liver BMP9 expression was reduced by 3-6 h, returning to normal levels by 18-24 h. 3 Plasma BMP9 levels gradually declined 24 h post-LPS administration, but whether plasma BMP9 remained reduced was not investigated. 3 Therefore, it is plausible that circulating levels of BMP9 and pBMP10 are reduced by systemic inflammation. However, in this study, BMP9 and pBMP10 did not F I G U R E 1 Circulating vascular injury and angiogenesis biomarkers after coronavirus disease 2019 (COVID-19) infection. Plasma samples from healthy controls (0; n = 29), patients whose maximal respiratory support was supplemental oxygen (4, 5, 6; n = 49), and patients who required assisted ventilation (7, 8, 9;  correlate with CRP and SAA, markers of systemic inflammation. Although not addressed here, the release of neutrophil elastase (NE) could be assessed and correlated with BMP9 plasma levels to further investigate the role of inflammation. It is known that neutrophils isolated from COVID-19 patients have increased NE release, and BMP9 is a substrate of NE. 3,16 However, it is still unclear whether the downregulation of BMP9 and pBMP10 is due to SARS-CoV-2 infection or associated inflammatory factors.
Interestingly, plasma concentrations did not correlate with disease severity as individuals who required mechanical ventilation had similar plasma concentrations to control subjects. The principal limitations of our study are the small number of patients recruited into the ventilated (7,8,9) cohort and the lack of longitudinal follow-up. Upon examination of the COVID-19 medications prescribed, we discovered that all ventilated patients were administered dexamethasone for significantly longer before study enrollment than those individuals requiring noninvasive oxygen therapy (Figure 1p). Successful treatment of COVID-19 with dexamethasone was originally highlighted by the Recovery Trial in the United Kingdom, 17 and we observed elevated BMP9 concentrations mildly correlated with the duration of dexamethasone treatment (Figure 1q).
Furthermore, due to the nature of patient recruitment, we also observed that several individuals in both the nonventilated and ventilated groups were not enrolled in the study until 7 days after hospital admission. Interestingly, plasma BMP9 (and pBMP10; data not shown) concentrations were decreased in patients recruited within 7 days of hospital admission, when compared to those recruited after 7 days (Figure 1r). We therefore cannot rule out that the administration of COVID-19 therapies (Table 1) might lead to the normalization of BMP9 (or pBMP10) concentrations. We hypothesize that plasma BMP9 and pBMP10 concentrations may be novel biomarkers of endothelial injury observed in hospitalized patients with COVID-19. However, longitudinal analysis would be required to determine whether these could predict disease severity and clinical outcome. The EpiCov database collated electronic health record data. Paul A. Lyons devised and supervised the collection and processing of COVID-19 samples. Kenneth G. C. Smith devised and supervised the collection and processing of COVID-19 samples. Stefan Gräf devised and supervised the collation of electronic health records and supervised the analysis of clinical data. Charlotte Summers supervised the analysis of clinical data and wrote the manuscript. Nicholas W. Morrell devised the study and wrote the manuscript.

ACKNOWLEDGMENTS
We thank NIHR BioResource volunteers for their participation, and gratefully acknowledge NIHR BioResource Centres, NHS Trusts, and staff for their contribution. In particular, we thank Kathy Stirrups for sample collection and curation. We thank the National Institute for Health and Care Research, NHS Blood and Transplant, and Health Data Research UK as part of the Digital Innovation Hub Program. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR, or the Department of Health and Social Care. The work was funded by awards from NIHR to the NIHR BioResource (RG94028 and RG85445). Furthermore, we thank Keith Burling and Peter Barker at the NIHR Cambridge BRC Core Biochemical Assay Laboratory (CBAL), Cambridge University Hospitals NHS Foundation Trust for the metabolite analysis. We also thank Laura Bergamaschi and Federica Mescia for their help in collecting samples and clinical analysis. The results reported in this publication are in part or entirely based on the analysis of electronic health record (EHR) data collated in the EpiCov database. The EpiCov database has been established by Cambridge University Hospitals in partnership with the Research Computing Services (RCS) at the University of Cambridge and other stakeholders to support COVID-19 relevant strategic monitoring and research analysis of EHR data with the aim to improve the care of individuals infected with SARS-CoV-2 at Cambridge University Hospitals and in other health and social care settings in the United Kingdom and beyond. Members of the clinical informatics team at CUH have been instrumental in preparing and pseudonymization of EHR data before transfer to the EpiCov database. Benjamin J. Dunmore and Paul D.