科研论文

当前位置: 检验科首页 > 科研教学 > 科研论文 > 正文

国家自然科学基金委员会官网报道曹炬教授团队重要研究成果

时间：2024-03-04 09:24:54 作者：重庆医科大学附属第一医院 浏览次数：


近日，国家自然科学基金委员会在其官网首页以《我国学者在脓毒症生物标志物及其免疫治疗研究方面取得重要进展》为题，报道了重医附一院医学检验科主任曹炬教授团队在脓毒症免疫机制研究领域取得的重要研究成果。该团队通过磷酸化蛋白组学、转录组学分析等手段，揭示了BMP9-ALK1轴活化Smad1/5信号通路促进巨噬细胞募集和细菌吞噬，控制宿主原发感染并减轻器官炎症损伤，为开展基于生物标志物的脓毒症精准治疗提供了新策略。






据悉，今年1月31日，曹炬教授团队的研究成果以《骨形态发生蛋白9是脓毒症的候选预后生物标志物和宿主定向治疗靶点》为题，在Science Translational Medicine杂志发表后，立即被JAMA正刊（IF=120.7)作为杂志当期要闻报道，同时，被期刊Nature Reviews Drug Discovery(IF=120.1）作为专题进行全篇亮点述评，凸显了国际医学顶级期刊对重医附一院该项重要科研成果的高度认可。据了解，《科技日报》《健康报》和《重庆日报》等媒体也对这一研究成果进行了报道。








当前，国家自然科学基金申报工作正在进行中。国家自然科学基金委员会官网、JAMA正刊和Nature子刊对重医附一院科研成果的报道和述评对医院的科技工作给予了极大的鼓舞，期待全院科研工作者在“科技强院”战略的指引下，提升科技创新能力，推动医院高质量发展，为服务人民健康提供坚实保障。




原文链接：

https://www.science.org/doi/10.1126/scitranslmed.adi3275



Edito’s summary

Current therapeutic strategies for sepsis are limited, and additional therapies tailored to individual patients are urgently needed. Here, Bai et al. identified bone morphogenetic protein 9 (BMP9) as a protein that was decreased in two cohorts of patients with sepsis. BMP9 concentrations at admission were higher in survivors, prompting the authors to ask whether exogenous BMP9 treatment could be a therapeutic option. BMP9 treatment improved outcomes in murine sepsis models by promoting macrophage recruitment, phagocytosis, and bacterial killing. Together, these data highlight the potential value of BMP9 both as a prognostic biomarker and as a host-directed therapy. —Courtney Malo

Abstract

Defining next-generation immune therapeutics for the treatment of sepsis will involve biomarker-based therapeutic decision-making. Bone morphogenetic protein 9 (BMP9) is a cytokine in the transforming growth factor–β superfamily. Here, circulating BMP9 concentrations were quantified in two independent cohorts of patients with sepsis. Decreased concentrations of serum BMP9 were observed in the patients with sepsis at the time of admission as compared with healthy controls. Concentrations of BMP9 at the time of admission were also associated with 28-day mortality, because patients with sepsis at a higher risk of death had lower BMP9 concentrations. The mechanism driving the contribution of BMP9 to host immunity was further investigated using in vivo murine sepsis models and in vitro cell models. We found that BMP9 treatment improved outcome in mice with experimental sepsis. BMP9-treated mice exhibited increased macrophage influx into the peritoneal cavity and more efficient bacterial clearance than untreated mice. In vitro, BMP9 promoted macrophage recruitment, phagocytosis, and subsequent bacterial killing. We further found that deletion of the type 1 BMP receptor ALK1 in macrophages abolished BMP9-mediated protection against polymicrobial sepsis in vivo. Further experiments indicated that the regulation of macrophage activation by the BMP9-ALK1 axis was mainly mediated through the suppressor of mother against decapentaplegic 1/5 signaling pathway. Together, these results suggest that BMP9 can both serve as a biomarker for patient stratification with an independent prognostic value and be developed as a host-directed therapy for sepsis.

SIGN UP FOR THE SCIENCEADVISER NEWSLETTER

The latest news, commentary, and research, free to your inbox daily
SIGN UP

INTRODUCTION

Sepsis is defined as a life-threatening organ dysfunction syndrome caused by a dysregulated host response to infection ( 1, 2). More than half of sepsis cases are caused by bacterial infections ( 3, 4). A recent Global Burden of Disease report highlights that there are nearly 50 million sepsis cases globally per year ( 5). Sepsis develops when the early host immune response fails to contain the infection, triggering an excessive cytokine-mediated host inflammatory response ( 6– 8). As sepsis progresses, it causes a compensatory anti-inflammatory response that can become immunosuppressive, which is associated with particularly poor outcomes and increased mortality either because of the inability to clear the initial infection or from increased risk of secondary infections ( 9, 10). Autopsy results have shown that most patients admitted to intensive care unit (ICU) for treatment of sepsis had unresolved septic foci postmortem, suggesting that the host immune system is compromised, resulting in persistent stimulation of host cells and organ injury that fails to return to homeostasis ( 11, 12). Discovery of host factors driving the unbalanced immune response in sepsis constitutes an important area of research that should offer effective treatment strategies.
Bone morphogenetic proteins (BMPs) are members of the transforming growth factor–β (TGF-β) family of cytokines ( 13). BMP9, also known as growth differentiation factor 2, is mainly produced from the liver and constantly circulated into the blood ( 14). BMP9 exerts its biological effects through binding to a receptor complex comprising the high-affinity type 1 BMP receptor ALK1 (activin receptor–like kinase 1) and the type 2 BMP receptor, BMPR2 (BMP receptor type 2) ( 15 – 17). Although BMP9-knockout (KO) mice are viable and fertile, they have abnormal lymphatic development and display fatty liver and obesity-related metabolic complications ( 18, 19). In addition, mice lacking ALK1 die early during embryonic development because of a severe lack of vascular remodeling ( 20, 21). BMP9 can promote ALK1-mediated phosphorylation of suppressor of mother against decapentaplegic 1/5/8 (Smad1/5/8), which bind to Smad4 and translocate to the nucleus to activate Smad-response elements in BMP gene targets, such as the inhibitors of differentiation ( Id1 to Id3) in endothelial cells ( 15, 22). Recently, circulating BMP9 has been shown to protect pulmonary endothelium during inflammation-induced lung injury ( 23). However, a role for BMP9 in sepsis has not been established. Here, we illustrated that BMP9 is a candidate prognostic biomarker and protective factor in sepsis.

RESULTS

Characteristics of the human participants

The discovery cohort (Sichuan) comprised 68 patients with sepsis and 30 healthy volunteers (table S1). The 28-day mortality was 30.88% in the Sichuan cohort patients. There were no differences in gender, age, body mass index (BMI), white blood cell (WBC) numbers, C-reactive protein (CRP) and procalcitonin (PCT) concentrations, or ICU stay days between survivors and non-survivors. However, a significant ( P = 0.0001) difference was observed between these patients when considering sequential (sepsis-related) organ failure assessment (SOFA) scores or septic shock on ICU admission.
The validation cohort (Chongqing) comprised 384 patients with sepsis and 50 healthy volunteers (table S2). The 28-day mortality was 31.51% in the Chongqing cohort patients. Similar to Sichuan cohort, there were no differences in gender, age, BMI, WBC numbers, CRP concentrations, or ICU days between survivors and non-survivors. However, significant differences were observed between survivors and non-survivors in SOFA scores ( P < 0.0001), PCT concentrations ( P = 0.0008), and septic shock status at the time of ICU admission ( P < 0.0001).

Sepsis results in reduced serum BMP9 concentrations in patients

In the Sichuan cohort, patients with sepsis had significantly ( P < 0.0001) lower serum concentrations of BMP9 on the day of ICU admission versus healthy volunteers (Fig. 1A). There were no differences in the serum concentrations of BMP2, BMP4, BMP6, BMP7, or BMP10 between patients and healthy controls. Serum BMP9 concentrations in the patients with septic shock were significantly ( P = 0.0214) lower compared with those in patients with sepsis without shock (Fig. 1B). Further, non-survivors showed significantly ( P < 0.0001) lower BMP9 concentrations than survivors (Fig. 1C). Furthermore, concentrations of serum BMP9 at admission negatively correlated with SOFA scores and CRP concentrations, although no correlations were observed between serum BMP9 concentrations and PCT concentrations or WBC counts (Fig. 1D). In addition, we analyzed the correlation of BMP9 with another inflammatory mediator, C-C motif chemokine ligand 2 (CCL2), an indicator of sepsis severity ( 3, 7, 8). The concentrations of serum CCL2 in patients with sepsis at admission were significantly ( P < 0.0001) higher than those in the healthy controls (fig. S1A). Serum CCL2 concentrations were also significantly increased in patients with septic shock compared with those in patients without shock ( P < 0.0001, fig. S1B) and in non-survivors compared with those in survivors ( P = 0.0004, fig. S1C). However, there was no correlation between serum concentrations of BMP9 and CCL2 in patients with sepsis (Fig. 1D).

图片

Expand for more
OPEN IN VIEWER

BMP9 concentrations predict survival in the patients with sepsis

Univariate and multivariate Cox regression analyses were performed to evaluate the association between admission concentrations of serum BMP9 and clinical outcome (28-day mortality). After Cox regression analyses, associations were observed between 28-day mortality and serum BMP9 concentrations, SOFA scores, or CCL2 concentrations. BMP9 concentrations, SOFA scores, and CCL2 concentrations remained in our multivariate model, and these parameters could independently predict 28-day mortality (table S3). To estimate the performance of serum BMP9 as a prognostic biomarker, the areas under the receiver operating characteristic (ROC) curves (AUCs) for prediction of 28-day mortality were calculated for BMP9 concentrations, SOFA scores, CRP concentrations, PCT concentrations, WBC counts, and CCL2 concentrations (Fig. 1E). The highest AUC was observed for serum BMP9 concentrations on day of ICU admission [AUC, 0.83; 95% confidence interval (CI), 0.72 to 0.91]. BMP9 concentrations below 19.82 pg/ml predicted 28-day mortality with a sensitivity of 76.19% and a specificity of 76.60% (table S4). Patients with lower serum BMP9 concentrations had poorer survival than patients with higher BMP9 concentrations when using a cutoff value of 19.82 pg/ml (Fig. 1F).
As observed in the Sichuan cohort, serum BMP9 concentrations at admission were significantly ( P < 0.0001) lower in patients with sepsis compared with those in healthy controls in the Chongqing cohort (Fig. 2A). Serum BMP9 concentrations were significantly ( P < 0.0001) decreased in the patients with septic shock compared with those in patients with sepsis without shock (Fig. 2B), and serum BMP9 concentrations were significantly ( P < 0.0001) lower in non-survivors compared with those in survivors (Fig. 2C). We also collected serum samples at days 3, 7, and 14 after admission in randomized survivors and non-survivors to study dynamic changes of BMP9. A gradual and significant ( P < 0.001) increase in serum BMP9 concentrations was noted in the survivors after sepsis (Fig. 2D). However, non-survivors showed an opposite trend, with significantly ( P < 0.01) decreased concentrations of serum BMP9 compared with those on day of ICU admission (Fig. 2E). Admission concentrations of serum BMP9 negatively correlated with SOFA scores, but there was no correlation of BMP9 concentrations with CRP concentrations, PCT concentrations, or WBC counts in patients with sepsis (Fig. 2F). In addition, serum concentrations of CCL2 were increased in patients with sepsis at admission compared with those in healthy controls (fig. S2A). The serum CCL2 concentrations in septic shock patients were higher than those in patients with sepsis without shock (fig. S2B), and non-survivors had higher admission concentrations of serum CCL2 than survivors (fig. S2C). Consistent with the Sichuan cohort, serum BMP9 did not correlate with CCL2 (Fig. 2F).

图片

Expand for more
OPEN IN VIEWER
The Chongqing cohort study confirmed the association between 28-day mortality and admission concentrations of serum BMP9 (table S5), and this association was still significant ( P = 0.0001) in our multivariate model. BMP9 concentrations, SOFA scores, and CCL2 concentrations independently predicted 28-day mortality in patients with sepsis. We assessed the performance of serum BMP9 concentrations at admission to predict 28-day mortality and obtained an AUC of 0.81 (95% CI, 0.77 to 0.85). This AUC value was higher than those for SOFA scores, CRP concentrations, PCT concentrations, WBC counts, and CCL2 concentrations (Fig. 2G). Using the threshold from the Sichuan cohort, a sensitivity of 66.94% and a specificity of 83.65% were obtained in the Chongqing cohort (table S6). Last, the survival rates of patients with sepsis were significantly ( P < 0.0001) different when stratified according to admission concentrations of serum BMP9 identified in the Chongqing cohort (Fig. 2H).

BMP9 treatment improves outcomes in a murine sepsis model

Using the cecal ligation and puncture (CLP)–induced polymicrobial sepsis model, we found that circulating BMP9 concentrations were significantly ( P < 0.01) reduced after CLP (Fig. 3A). Moreover, the protein expression of BMP9 was decreased in the hearts, livers, spleens, lungs, and kidneys of septic mice as compared with that of naive mice (Fig. 3B). These results indicate a suppression of BMP9 protein production during sepsis in mice, similar to what was observed in patients.

图片

Expand for more
OPEN IN VIEWER
To investigate whether replenishment of BMP9 during sepsis could modify the progress of disease, we treated mice that had undergone CLP with recombinant murine BMP9 (rmBMP9; Fig. 3C). In a CLP sepsis model that consistently caused 80 to 100% mortality, early treatment with rmBMP9 (6.25, 12.5, or 25.0 μg/kg) at the time of CLP increased mouse survival (Fig. 3D) when compared with vehicle-treated controls. Delayed treatment with rmBMP9 (25.0 μg/kg) after the mice displayed clinical signs (at 6 or 12 hours after CLP) also prolonged survival (Fig. 3E). BMP9 protected against CLP-induced mortality in a sex-independent manner (Fig. 3F). Furthermore, resistance to CLP-induced mortality was observed in BMP9-treated mice in the presence of antibiotics (Fig. 3G). Given these results, our subsequent experiments were performed in male mice unless otherwise stated using rmBMP9 (25.0 μg/kg) at the time of CLP in the absence of antibiotics to assess systemic inflammation, dissemination of polyintestinal bacteria ( 24 – 27), and whether BMP9 treatment could normalize the circulating concentrations of BMP9 in septic mice (fig. S3A). To confirm target engagement, we measured the changes in BMP9-regulated genes in the peritoneal lavage fluids (PLFs). As expected, BMP9 administration led to an enhancement in BMP9 activity, as evidenced by the elevated mRNA expression of the BMP9 target gene, Id1 (fig. S3B).
The decreased mortality of septic mice with BMP9 treatment was associated with reduced evidence of organ injury. Serum concentrations of aspartate aminotransferase (AST) and alanine aminotransferase (ALT), markers for hepatocellular injury, lactate dehydrogenase (LDH), a marker for general cellular injury, and blood creatinine and urea, markers for kidney injury, were reduced in mice treated with BMP9 (Fig. 3H). Tissue inflammation and damage were also analyzed by hematoxylin and eosin staining; BMP9 treatment decreased the histopathological scores of the lungs, livers, spleens, and kidneys after CLP (Fig. 3I).
Effective and rapid bacterial clearance is a fundamental determinant of outcomes in sepsis ( 28, 29). Therefore, we examined the effects of BMP9 treatment on bacterial burdens. Administration of BMP9 reduced bacterial loads in the PLF and blood after CLP (Fig. 3J). Both Staphylococcus aureus (a Gram-positive bacterium) and Pseudomonas aeruginosa (a Gram-negative bacterium) are major contributors to sepsis in the clinic ( 3, 4). We thus tested the effects of BMP9 treatment on clinical outcomes in S. aureus– or P. aeruginosa–dependent peritonitis models, finding that BMP9 treatment protected mice against S. aureus– or P. aeruginosa–mediated mortality (fig. S4, A and B). Moreover, bacterial overgrowth in the PLF and blood was decreased in BMP9-treated mice when compared with that in vehicle control (fig. S4, C and D). BMP9 treatment also protected against intranasal infection with S. aureus or P. aeruginosa (fig. S4, E and F), and bacterial overgrowth in the lung and blood was reduced in BMP9-treated mice (fig. S4, G and H).

BMP9 increases CCL2 but decreases IL-6 concentrations in experimental sepsis

The effect of BMP9 on CLP-induced cytokines and chemokines was evaluated after CLP. Among the cytokines, including interleukin-1β (IL-1β), IL-6, and IL-27, and chemokines, including CCL2, CCL3, CXCL1, and CXCL10, that were tested, CCL2 was the only chemokine whose concentration was increased in the PLF and blood of BMP9-treated mice at 6 and 24 hours after CLP (Fig. 3K). Although BMP9 had no effect on IL-6 production at 6 hours, it resulted in a decrease in IL-6 concentrations in the PLF and blood at 24 hours after CLP (Fig. 3K). There were no differences in IL-1β, IL-27, CCL3, CXCL1, and CXCL10 concentrations between BMP9-treated and vehicle-treated mice at 6 and 24 hours after CLP (fig. S5). Collectively, these data indicate that BMP9 had a specific and selective effect on the production of CCL2 and IL-6 that may regulate the pathology of polymicrobial sepsis.

Inhibition of endogenous BMP9 increases mortality and decreases bacterial clearance in experimental sepsis

To further demonstrate the role of BMP9 in protecting mice against sepsis, we examined whether systemic treatment with a BMP9-neutralizing monoclonal antibody would affect the development of sepsis in a CLP sepsis model using a 24-gauge needle that consistently caused 40 to 50% mortality ( 23, 30). Immediately after CLP, mice were treated with the anti-BMP9 antibody (fig. S6A). Anti-BMP9–treated mice exhibited increased mortality compared with mice treated with control immunoglobulin G (IgG; fig. S6B), which was associated with increased serum concentrations of AST, ALT, LDH, creatinine, and urea (fig. S6C). Histological analysis demonstrated greater pathology in anti-BMP9–treated mice after CLP (fig. S6D). Anti-BMP9–treated mice also had increased bacterial burdens in the PLF and blood (fig. S6E). Furthermore, the concentrations of CCL2 in PLF ( P < 0.05) and blood ( P < 0.01) from anti-BMP9–treated mice were significantly decreased, but IL-6 concentrations in PLF and blood were increased in anti-BMP9–treated mice (fig. S6F). There were no differences in the serum concentrations of IL-1β, IL-27, CCL3, CXCL1, or CXCL10 between anti-BMP9–treated and IgG-treated mice after CLP. We confirmed that the anti-BMP9 antibody treatment had decreased circulating BMP9 concentrations (fig. S7A). To further confirm that the anti-BMP9 antibody had inhibited BMP9 activity, we measured Id1 gene expression in PLF, which was also reduced in anti-BMP9–treated mice (fig. S7B).

BMP9 protects mice against sepsis by promoting macrophage recruitment and activation

To investigate the mechanisms underlying BMP9-mediated protection against polymicrobial sepsis, we performed single-cell RNA sequencing (scRNA-seq) of PLF (Fig. 4A). After quality filtering, we obtained 63,028 cells for analysis. A total of 16 cell clusters (clusters 0 to 15) were characterized, and there was no discernible difference in the profiles of major immune cells between BMP9- and vehicle-treated mice after CLP (Fig. 4B). Next, by annotating each cluster according to the marker genes, 16 cell types were obtained from these 16 clusters, mainly including neutrophils, monocytes, macrophages, and B cells (fig. S8A). We first examined BMP9-related differences in cell-type composition of each group. Cluster 0 was the largest cell subgroup, and its proportion was increased in the BMP9-treated mice (Fig. 4C). For each individual cluster, we identified the top five genes that were differentially expressed. Specifically, Ccl2 was used as a marker to identify cluster 0 (Fig. 4D and fig. S8B). This population highly expressed the immune cell marker Cd45 and macrophage markers including Cd11b and F4/80 (Fig. 4E). Cluster 0 was therefore identified as Ccl2 High macrophages; these cells also expressed Ccr2, indicating that these macrophages could be recruited to the site of infection in response to CCL2 in the microenvironment (fig. S8C). As mentioned above, CCL2 concentrations in CLP mice treated with BMP9 protein or anti-BMP9 antibody differed from those in the control group (Fig. 3K and fig. S6F). To better elucidate the function of cluster 0 cells, Gene Ontology Biological Processes (GO-BP) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses of differentially expressed genes (DEGs) were performed (fig. S8D). The overrepresented GO-BP terms enriched by DEGs in cluster 0 were “immune system process” and “response to bacterium” (Fig. 4F), and the KEGG terms enriched were “NOD-like receptor signaling pathway” and “chemokine signaling pathway” (Fig. 4G). Together, these results suggested a potential role of BMP9 in macrophage recruitment and activation in sepsis.

图片

Expand for more
OPEN IN VIEWER
Flow cytometry analysis confirmed that the frequency of infiltrating CD11b +F4/80 +macrophages in the peritoneum was significantly ( P = 0.0159) higher in BMP9-treated mice relative to the vehicle-treated mice after CLP (Fig. 4H). Thus, we hypothesized that macrophages might mediate the protection effects of BMP9 on sepsis. Depleting macrophages by clodronate liposomes abrogated the beneficial effects of BMP9 on survival of septic mice after CLP (Fig. 4, I and J). Furthermore, the survival rate of septic mice receiving BMP9-treated macrophages was significantly ( P = 0.0026) higher than that of septic mice receiving saline-treated macrophages (Fig. 4K). However, BMP9 treatment did not influence CD11b +Ly6G + neutrophil frequency in the peritoneum after CLP (fig. S9A). There was no change in the survival of BMP9-treated septic mice after depleting neutrophils by anti-Ly6G monoclonal antibody (fig. S9, B and C). There were also no differences in the frequency of CD45 +CD3 + T lymphocytes or CD4 +CD25 +Foxp3 + regulatory T cells in PLF after CLP (fig. S9, D and E).

BMP9 protects against experimental sepsis through CCL2 production

CCL2 is a primary chemokine of monocytes and macrophages ( 31). Having observed that BMP9 augmented the production of CCL2 and Ccl2 High macrophages mainly infiltrated into the peritoneum of BMP9-treated mice (Fig. 3K and Fig. 4, C and D), we next studied whether the CCL2-macrophage axis was directly responsible for BMP9-mediated protection against sepsis. We first characterized the cellular source of CCL2 in BMP9-treated mice during CLP, finding that depleting macrophages could abrogate BMP9-mediated CCL2 increase in the PLF and blood after CLP (fig. S10A). In vitro studies also demonstrated that BMP9 could enhance CCL2 production in macrophages but not neutrophils and lymphocytes (fig. S10B). These data suggest that macrophages are the major sources of CCL2 elicited by BMP9 in CLP-induced sepsis.
To further investigate the contribution of CCL2 to BMP9-mediated protection against sepsis, we used a specific CCL2 monoclonal antibody, which resulted in a decrease in macrophage recruitment (Fig. 5A). Anti-CCL2 treatment significantly ( P < 0.0125 at 1 mg/kg; P = 0.0005 at 2 mg/kg) increased mortality in BMP9-treated CLP mice compared with that in IgG isotype control (Fig. 5B). Anti-CCL2 also caused an increase in organ injury of BMP9-treated mice, as shown by elevated serum concentrations of AST, ALT, LDH, creatinine, and urea after CLP (Fig. 5C). Tissue inflammation of the lung, liver, spleen, and kidney was increased by anti-CCL2 antibody in BMP9-treated mice, which was reflected by greater pathology scores after CLP (Fig. 5D). As predicted, an increase of bacterial burden in the PLF and blood from BMP9-treated septic mice was also observed after anti-CCL2 treatment (Fig. 5E).

图片

Expand for more
OPEN IN VIEWER

BMP9 can directly enhance bacterial phagocytosis and killing by macrophages

Because macrophages are the primary effector cells for bacterial clearance during sepsis ( 28, 31, 32), we tested whether BMP9 modulates intrinsic antibacterial functions of macrophages. BMP9 treatment significantly ( P < 0.01) enhanced phagocytosis of Escherichia coli by macrophages (Fig. 6A and fig. S11A). BMP9-mediated enhancement of phagocytosis was associated with an increase of phagosome formation in macrophages. Scanning electron microscopy showed that no zymosan particle was detected on the surface of BMP9-treated macrophages (fig. S11B), and transmission electron microscopy showed that zymosan particles were all intracellular in BMP9-treated macrophages (fig. S11C). By contrast, in vehicle-treated macrophages, many zymosan particles were observed on the cell surface in phagocytic cups, in which distal margins remained open. BMP9 treatment also increased intracellular killing of live E. coli by macrophages (Fig. 6B), and the production of reactive oxygen species (ROS) was augmented in E. coli–stimulated, BMP9-treated macrophages when compared with vehicle control (Fig. 6C). However, a direct killing effect of BMP9 on bacteria was not observed (fig. S12). BMP9 did not influence bacterial phagocytosis and killing in neutrophils on E. coli infection (fig. S13, A and B). Because the frequency of peritoneum-infiltrating macrophages was increased in BMP9-treated mice (Fig. 4H), we also evaluated potential chemokine properties of BMP9 for macrophages. An in vitro migration assay showed that BMP9 yielded similar chemotactic activity of macrophages to the positive control, CCL2 treatment (Fig. 6D).

图片

Expand for more
OPEN IN VIEWER

ALK1 signaling in macrophages plays a critical role in BMP9-induced protection against experimental sepsis

Activin A receptor–like type 1 ( Acvrl1)/ALK1, activin A receptor type 1 ( Acvr1)/ALK2, BMP receptor 1A ( Bmpr1a)/ALK3, and BMP receptor 1B ( Bmpr1b)/ALK6 have been reported to be receptors for BMP9 ( 15, 22, 33). Our scRNA-seq data showed that the expression of Acvrl1 in PLF cells was higher than that of Acvr1, Bmpr1a, or Bmpr1b, and it was mainly distributed in macrophages (Fig. 7A). To explore the functional role of macrophage ALK1 in BMP9-mediated protection against sepsis, we first confirmed the protein expression of ALK1 in primary peritoneal macrophages (fig. S14A). We next generated myeloid-specific ALK1 KO mice (LyzM-Cre +-ALK1 flox/flox mice, hereinafter defined as ALK1 f/f Cre + mice) to delete ALK1 in macrophages. Their littermates (LyzM-Cre −-ALK1 flox/flox mice, hereinafter defined as ALK1 f/f Cre − mice) served as controls (Fig. 7B). ALK1 expression was successfully deleted in macrophages (fig. S14, B and C).

图片

Expand for more
OPEN IN VIEWER
ALK1 f/f Cre + and ALK1 f/f Cre − mice were then subjected to CLP in the presence or absence of BMP9 treatment. Deletion of ALK1 on macrophages abolished BMP9-induced protection against sepsis, as shown by mortality rate (Fig. 7C); serum concentrations of AST, ALT, LDH, creatinine, and urea (Fig. 7D); tissue pathology scores (Fig. 7E); bacterial burdens (Fig. 7F and fig. S14D); and CCL2 production (Fig. 7G). In vitro studies also demonstrated that BMP9 treatment had no influence on bacterial phagocytosis and killing (Fig. 7, H and I; and fig. S14, E and F), chemotaxis activity (Fig. 7J), or CCL2 induction (Fig. 7K) in macrophages isolated from ALK1 f/f Cre + mice. Furthermore, we used CRISPR-Cas9 to generate RAW264.7 cell line with ALK1 knocked out (fig. S14G); ALK1 ablation in RAW264.7 cell line resulted in the abolishment of bacterial phagocytosis and killing (fig. S14, H and I), chemotaxis activity (fig. S14J), and CCL2 production (fig. S14K) upon BMP9 treatment.

BMP9 promotes antibacterial functions of macrophages through the Smad1/5 pathway

Having demonstrated that BMP9 acted on macrophages through ALK1 to protect against experimental sepsis, we next performed phosphoproteomics and RNA-seq to investigate downstream signaling (Fig. 8A). BMP9 activated TGF-β phosphorylation signaling pathways (Fig. 8B), and RNA-seq confirmed that TGF-β signaling was the main pathway activated by BMP9 in macrophages (Fig. 8C). To gain further insight into the intracellular signaling pathways in macrophages triggered by BMP9, we evaluated the phosphorylation of Smad1/5. BMP9 induced the phosphorylation of Smad1/5 in a dose- and time-dependent manner but not of nuclear factor κB (NF-κB) p65, p38 mitogen-activated protein kinase (p38MAPK), c-Jun N-terminal kinase (JNK), extracellular signal–regulated kinase (ERK), or Akt (fig. S15, A and B). Macrophage-specific ALK1 loss suppressed Smad1/5 phosphorylation induced by BMP9 (fig. S15C). Heat-killed E. coli stimulated the phosphorylation of NF-κB p65, p38MAPK, JNK, ERK, or Akt but not Smad1/5, and the phosphorylation of Smad1/5 in macrophages triggered by heat-killed E. coli was only observed after treatment with BMP9 (Fig. 8D). ML347 is a highly selective ALK1/ALK2 inhibitor, which can block the phosphorylation of Smad1/5 by TGF-β1 ( 34, 35). Following these studies and toxicity threshold values from the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay (fig. S15D), we used an optimal concentration of ML347 that could suppress Smad1/5 activation by BMP9 (fig. S15E). BMP9-mediated increases of bacterial phagocytosis and killing (Fig. 8, E and F), chemotaxis activity (Fig. 8G), and CCL2 production (Fig. 8H) were suppressed in macrophages by ML347. Furthermore, ML347 treatment impaired BMP9-induced protection against death in mice (fig. S15F), as well as BMP9-mediated CCL2 increase after CLP (fig. S15G).

图片

Expand for more
OPEN IN VIEWER

BMP9 enhances antibacterial functions of human monocytes and monocyte-derived macrophages through the ALK1/ Smad1/5 signaling pathway

To assess whether the preceding observations applied to human macrophages, human monocytes were differentiated into monocyte-derived macrophages. Treatment with recombinant human BMP9 (rhBMP9) could significantly ( P < 0.01) enhance bacterial phagocytosis and killing in human macrophages (Fig. 8, I and J), and BMP9 acted as a chemoattractant to human monocytes (Fig. 8K). Besides, BMP9 could augment CCL2 production from human macrophages but not from neutrophils and lymphocytes (Fig. 8L). Anti-ALK1 treatment could inhibit bacterial phagocytosis and killing in human macrophages elicited by BMP9 (Fig. 8, M and N) and impair the chemotaxis of human monocytes triggered by BMP9 (Fig. 8O). Anti-ALK1 also suppressed BMP9-induced CCL2 production in human macrophages (Fig. 8P). Similarly, ML347 inhibited BMP9-mediated increase of bacterial phagocytosis and killing in human macrophages (Fig. 8, Q and R), chemotaxis of human monocytes induced by BMP9 (Fig. 8S), and enhanced CCL2 production elicited by BMP9 (Fig. 8T). Collectively, our results suggest that BMP9 protects against the development of sepsis and is able to increase macrophage recruitment and to enhance bacterial phagocytosis and killing capacity of macrophages through ALK1 receptor and Smad1/5 signaling (fig. S16).

DISCUSSION

Here, we report four important findings for the role of BMP9 and propose BMP9 as a candidate treatment for sepsis: (i) Sepsis was characterized by a defect of BMP9 induction, and lower admission concentrations of BMP9 were associated with higher mortality in patients; (ii) BMP9 restoration therapy protected mice with polymicrobial infection from sepsis in a macrophage-dependent manner; (iii) inactivation of ALK1 in macrophages abrogated the protection effects of BMP9 on experimental sepsis; and (iv) activation of Smad1/5 signaling pathway was crucial for BMP9-mediated augment of antibacterial functions in macrophages and improvement of survival in experimental sepsis. Hence, BMP9 not only could serve as a biomarker to predict the outcome of sepsis but also represents a potential host-directed therapy to restore immune homeostasis in combating sepsis.
In a hope for more successful patient outcomes, it is critical to efficiently and accurately determine the immune status of patients with sepsis for specific pathobiological abnormalities that could be targeted ( 36 – 41). This study showed that BMP9 production was impaired in patients with sepsis and that BMP9 concentrations at admission were associated with 28-day mortality. The strength of this study is that the findings were replicated in two independent and geographically dispersed cohorts of patients with sepsis. This is important because available biomarkers of host immune response usually present with predictive value in regard to outcome prediction later on after the onset of shock. One example is the decreased expression on circulating monocytes of human leukocyte antigen–DR (HLA-DR), which has been shown to be associated with increased risk of death after septic shock ( 42, 43). Another is mRNA expression of the gene encoding IL-7 receptor ( IL7R), which was lower in non-survivors on day 3 after septic shock and was further associated with 28-day mortality ( 44 ). In contrast to HLA-DR, which requires quantification by flow cytometry, and IL7R expression, which requires quantification by polymerase chain reaction ( 44, 45), BMP9 protein concentration can likely be easily measured in serum using automated tests that are available in routine and emergency laboratories. To further support BMP9 as a biomarker, a prior study found that circulating BMP9 concentrations were markedly lower in portopulmonary hypertension (PoPH), which could act as a biomarker of PoPH, predicting transplant-free survival ( 46).
Early and effective infection control is a critical element of treatment for sepsis ( 47 – 50). This pilot study demonstrated that BMP9 was protective in sepsis through improving survival and host control of infection. We demonstrated that BMP9 directly stimulated the chemokinesis of macrophages and indirectly empowered macrophages to migrate to the site of infection by augmenting chemokine CCL2 production from macrophages. Although production of CCL2 was essential for host defense against bacteria ( 31, 51, 52), overproduction of CCL2 has been shown to correlate with organ dysfunction and mortality after sepsis ( 53, 54). Therefore, the immune response induced by BMP9-CCL2 axis in macrophages during sepsis needs to be tightly controlled. Consistent with previous studies ( 55, 56), we found that CCL2 revealed prognostic value for 28-day mortality in the patients with sepsis. Although macrophages are found to be the major source of CCL2 ( 57), other cell types, including fibroblasts, endothelial, epithelial, and smooth muscle cells, may produce CCL2 after induction by oxidative stress, cytokines, or growth factors ( 58). It is thus reasonable that there was no correlation between BMP9 and CCL2 concentrations in patients with sepsis. The evidence that CCL2 was protective in bacterial clearance during sepsis may be viewed at first glance as paradoxical, given that high CCL2 concentrations were associated with increased mortality in patients with sepsis. These high CCL2 concentrations may be a reactive process to sepsis-mediated failure to contain bacterial infection that does not achieve full protection. Similar observations have been made for other mediators, such as progranulin ( 31, 59), which can promote protection against bacterial clearance in sepsis. The positive relationship between high concentrations of CCL2 and septic death in our analysis of the prognostic value of CCL2 concentrations is consistent with this scenario. In addition, BMP9 could directly enhance phagocytic function by promoting phagosome formation and bacterial killing by increasing ROS production in macrophages. Therefore, BMP9 improved bacterial clearance in sepsis through macrophage recruitment as well as by enhancing bacterial phagocytosis and killing in macrophages.
ALK1 is a type I receptor predominantly expressed in endothelial cells, mediating the signals from BMP9 ( 22, 60). We found that ALK1 was highly expressed on macrophages in BMP9-treated mice and identified a crucial role for ALK1 in macrophages in promoting bacterial clearance and improving survival mediated by BMP9 upon sepsis. This is consistent with a recent study showing that ALK1 was important for Listeria monocytogenes capture by Kupffer cells, because the loss of ALK1 led to a failure of bacterial capture and overwhelming disseminated infections ( 61). Although loss of Smad3 in macrophages has been shown to perturb phagocytic activity ( 62), our study reported an additional role for Smad1/5 signaling in bacterial clearance through BMP9-ALK1 pathway in macrophages.
Our study has several limitations. First, it may still be enriched compared with the Western population in general clinical practice to examine the clinical role of circulating BMP9. Second, BMP9-mediated survival improvement in response to murine sepsis was observed in the presence of antibiotics, suggesting that mechanisms other than bacterial clearance may also participate in BMP-mediated protection. Broken barriers have been recognized as a major take on sepsis pathogenesis ( 63), and a recent study has identified endogenous BMP9 as a pulmonary endothelial–protective factor ( 23). We thus hypothesize that BMP9 may participate in reducing sepsis severity by maintaining endothelial barrier function. Third, future in vivo imaging studies would further substantiate our findings from in vitro assays of macrophages activated by BMP9. Last, the cross-talk between BMP9-CCL2 axis in macrophages and other immune cells or factors during sepsis is still incompletely understood. Similarly, other immune cells or factors have been shown to have a specific and selective effect on the production of key cytokines and chemokines that drive the pathology of polymicrobial sepsis ( 26, 32, 64, 65).
In summary, our study revealed that sepsis was associated with a decrease in BMP9 production. BMP9 measurement could represent an attractive biomarker for identification of septic individuals presenting with higher risk of mortality. In complementary experimental studies, recombinant BMP9 replacement therapy enhanced bacterial clearance and improved survival in sepsis by activating macrophages in an ALK1 and Smad1/5-dependent manner. Thus, these data lay the foundation for future studies guiding the development of personalized BMP9-based immune therapies for sepsis.

MATERIALS AND METHODS

Study design

This study was designed to determine the prognostic and functional role of BMP9 in sepsis. Identification and quantification of BMP9 was performed in the serum of patients with sepsis in two different independent and prospectively collected cohorts: one from Sichuan Provincial People’s Hospital (Sichuan cohort) and one from the First Affiliated Hospital of Chongqing Medical University (Chongqing cohort). The Sichuan cohort comprised 68 patients. All patients met the requirements according to the clinical criteria of Sepsis-3 ( 1). Patients who were pregnant or breast-feeding; who were diagnosed with malignancy, organ transplantation, HIV infection, or autoimmune disease; or who were using immunosuppressive medication were excluded from the Sichuan cohort study ( 66 – 69). The Chongqing cohort comprised 384 patients. Patients with sepsis with preexistent immunosuppression history or malignancy were excluded in the discovery cohort of Sichuan but were included in the validation cohort of Chongqing ( 23, 70, 71).
Functional studies were performed in vivo in mouse models (polymicrobial sepsis and single-bacterial sepsis model) and in vitro in cell culture models (human and mouse macrophages) to determine whether BMP9 is a candidate therapeutic target for treating sepsis and to understand the underlying regulatory mechanisms by which BMP9 modulates the activation of macrophages. For all in vivo experiments, mice were randomly assigned to experimental groups. The in vivo experiments were designed to detect differences between treatment groups or genotype-dependent effects at 80% power (α = 0.05), and our in vivo sample sizes are also similar to those generally used in the field. The primary endpoint was survival. The parameters of in vitro experiments were determined according to related published papers and preliminary experiments. Sample sizes for in vitro assays were those used by other laboratories in the field. All samples of mouse, cell, histological, or other experiments were blinded to investigators before the images or results were obtained and quantified. Data were not excluded in any studies. The number of samples and the number of experimental replicates for each experiment are reported in the figure legends. Further details are described in Supplementary Materials and Methods.

Statistical analysis

All individual-level data are presented in data file S1. Chi-square or Fisher exact tests were used for assessment of qualitative variables. Unless otherwise stated, quantitative variables were compared with Mann-Whitney U test or Kruskal-Wallis test followed by Dunn’s multiple comparisons posttest as appropriate. Spearman correlation coefficient ( r) was used to evaluate the correlations. The association between serum BMP9 concentration and 28-day mortality was analyzed by univariate and multivariate Cox models. For multivariate analysis, confounding factors were selected using the following criteria: a correlation between the variable and serum BMP9 concentration (Spearman r ≥ 0.15), a significant association with 28-day mortality ( P < 0.1, Cox model), and the absence of missing values. ROC curves and AUC values were calculated to assess the predictive performance regarding 28-day mortality. The DeLong method was used for comparison of AUC. In the Chongqing cohort, survival curves with 95% CIs were drawn in patient groups defined according to BMP9 concentration threshold on day of ICU admission in the Sichuan cohort. Differences in survival (Kaplan-Meier survival curves) were analyzed by log-rank (Mantel-Cox) test. All statistics were performed with SPSS (version 17.0) and GraphPad Prism (version 5.03) software. The level of statistical significance was set at P < 0.05, and all tests were two-tailed. Details of statistical tests are included in the figure legends.

Acknowledgments

We acknowledge X. Zhang (Key Laboratory of Laboratory Medical Diagnostics Designated by the Ministry of Education, School of Laboratory Medicine, Chongqing Medical University) for technical expertise and support.
Funding: This work was supported by grants from National Natural Science Foundation of China (grant nos. 82070014 and 82370009 to J. C.), Natural Science Foundation Project of Chongqing (grant no. CSTB2022NSCQ-LZX0017 to J.C.), Sichuan Science and Technology Program (grant no. 2022YFS0238 to C.W.), and Scientific and Technological Research Program of Chongqing Municipal Education Commission (grant no. KJZD-K201900405 to J.C.).
Author contributions: J.C., X.L., H.B., and Q.L. conceived and designed the study. H.B., Q.L., X.L., C.W., F.X., J.L., K.W., H.D., and Y.L. performed the experiments. J.C., X.L., C.W., F.X., J.L., Y.L., and Y.Y. coordinated the clinical evaluation of the patient and their parents. J.C., H.B., Q.L., and J.L. drafted the paper. J.C., H.B., and Q.L. proofread the paper. All authors reviewed and edited the manuscript.
Competing interests: The authors declare that they have no competing interests.
Data and materials availability: All data associated with this study are present in the paper or the Supplementary Materials. RNA-seq data are available at GEO under accession number GSE231894 (ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE231894). scRNA-seq data are available at GEO under accession number GSE230030 (ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE230030). Proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE with accession number PXD045566 (https://proteomecentral.proteomexchange.org/cgi/GetDataset?ID=PXD045566). For data or material requests, please contact the corresponding authors (J.C. and X.L.).







向下滑动查看论文全文


专家简介


曹炬 研究员


研究员，二级教授，博士生导师， 现任重医附一院医学检验科主任，国家级人才，国家重点研发计划首席科学家，国家科技部中青年科技创新领军人才，重庆市首席专家工作室领衔专家，重庆市杰出青年基金获得者，重庆市特支计划青年拔尖人才，重庆市中青年医学卓越团队负责人，中华医学会检验医学分会临床微生物学组副组长，中国医师协会检验医师分会委员。


主要从事重症感染诊疗标志物研究与临床检验工作。主持国家重点研发计划项目1项，国家自然科学基金7项，以通讯作者在Sci Transl Med,Am J Respir Crit Care Med，EMBO Mol Med和Plos Pathog等国际权威期刊发表SCI论文30余篇，被引用1000余次，授权国家发明专利7项，获中华医学科技奖二等奖，重庆市自然科学二等奖和浙江省科技进步二等奖。

擅长领域：感染性疾病的分子免疫诊断。


撰稿 | 赖晓霏
编辑 | 李丹 李萌
审核 | 曹炬 周芳



下一条：2022年文章

【关闭】

地址：重庆市渝中区袁家岗友谊路1号
【来院交通路线】

门诊部：023-89012255
住院部：023-89012513

扫描二维码

联系我们

重庆医科大学附属第一医院医学检验科 版权所有 E-mail:nic@hospital-cqmu.com 地址：重庆市渝中区袁家岗友谊路1号  站长统计