Prediction of methicillin-resistant Staphylococcus aureus and carbapenem-resistant Klebsiella pneumoniae from flagged blood cultures by combining rapid Sepsityper MALDI-TOF mass spectrometry with machine learning.

This study investigated combination of the Rapid Sepsityper Kit and a machine learning (ML)-based matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) approach for rapid prediction of methicillin-resistant Staphylococcus aureus (MRSA) and carbapenem-resistant Klebsiella pneumoniae (CRKP) from positive blood culture bottles. The study involved 461 patients with monomicrobial bloodstream infections. Species identification was performed using the conventional MALDI-TOF MS Biotyper system and the Rapid Sepsityper protocol. The data underwent preprocessing steps, and ML models were trained using preprocessed MALDI-TOF data and corresponding labels. The interpretability of the model was enhanced using SHapely Additive exPlanations values to identify significant features. In total, 44 S. aureus isolates comprising 406 MALDI-TOF MS files and 126 K. pneumoniae isolates comprising 1249 MALDI-TOF MS files were evaluated. This study demonstrated the feasibility of predicting MRSA among S. aureus and CRKP among K. pneumoniae isolates using MALDI-TOF MS and Sepsityper. Accuracy, area under the receiver operating characteristic curve, and F1 score for MRSA/methicillin-susceptible S. aureus were 0.875, 0.898 and 0.904, respectively; for CRKP/carbapenem-susceptible K. pneumoniae, these values were 0.766, 0.828 and 0.795, respectively. In conclusion, the novel ML-based MALDI-TOF MS approach enables rapid identification of MRSA and CRKP from flagged blood cultures within 1 h. This enables earlier initiation of targeted antimicrobial therapy, reducing deaths due to sepsis. The favourable performance and reduced turnaround time of this method suggest its potential as a rapid detection strategy in clinical microbiology laboratories, ultimately improving patient outcomes.

Direct prediction of carbapenem-resistant, carbapenemase-producing, and colistin-resistant Klebsiella pneumoniae isolates from routine MALDI-TOF mass spectra using machine learning and outcome evaluation.

The objective of this study was to develop a rapid prediction method for carbapenem-resistant Klebsiella pneumoniae (CRKP) and colistin-resistant K. pneumoniae (ColRKP) based on routine MALDI-TOF mass spectrometry (MS) results in order to formulate a suitable and rapid treatment strategy. A total of 830 CRKP and 1462 carbapenem-susceptible K. pneumoniae (CSKP) isolates were collected; 54 ColRKP isolates and 1592 colistin-intermediate K. pneumoniae (ColIKP) isolates were also included. Routine MALDI-TOF MS, antimicrobial susceptibility testing, NG-Test CARBA 5, and resistance gene detection were followed by machine learning (ML). Using the ML model, the accuracy and area under the curve for differentiating CRKP and CSKP were 0.8869 and 0.9551, respectively, and those for ColRKP and ColIKP were 0.8361 and 0.8447, respectively. The most important MS features of CRKP and ColRKP were m/z 4520-4529 and m/z 4170-4179, respectively. Of the CRKP isolates, MS m/z 4520-4529 was a potential biomarker for distinguishing KPC from OXA, NDM, IMP, and VIM. Of the 34 patients who received preliminary CRKP ML prediction results (by texting), 24 (70.6%) were confirmed to have CRKP infection. The mortality rate was lower in patients who received antibiotic regimen adjustment based on the preliminary ML prediction (4/14, 28.6%). In conclusion, the proposed model can provide rapid results for differentiating CRKP and CSKP, as well as ColRKP and ColIKP. The combination of ML-based CRKP with preliminary reporting of results can help physicians alter the regimen approximately 24 h earlier, resulting in improved survival of patients with timely antibiotic intervention.

Predicting microvascular invasion in hepatocellular carcinoma: a deep learning model validated across hospitals.

BACKGROUND: The accuracy of estimating microvascular invasion (MVI) preoperatively in hepatocellular carcinoma (HCC) by clinical observers is low. Most recent studies constructed MVI predictive models utilizing radiological and/or radiomics features extracted from computed tomography (CT) images. These methods, however, rely heavily on human experiences and require manual tumor contouring. We developed a deep learning-based framework for preoperative MVI prediction by using CT images of arterial phase (AP) with simple tumor labeling and without the need of manual feature extraction. The model was further validated on CT images that were originally scanned at multiple different hospitals. METHODS: CT images of AP were acquired for 309 patients from China Medical University Hospital (CMUH). Images of 164 patients, who took their CT scanning at 54 different hospitals but were referred to CMUH, were also collected. Deep learning (ResNet-18) and machine learning (support vector machine) models were constructed with AP images and/or patients' clinical factors (CFs), and their performance was compared systematically. All models were independently evaluated on two patient cohorts: validation set (within CMUH) and external set (other hospitals). Subsequently, explainability of the best model was visualized using gradient-weighted class activation map (Grad-CAM). RESULTS: The ResNet-18 model built with AP images and patients' clinical factors was superior than other models achieving a highest AUC of 0.845. When evaluating on the external set, the model produced an AUC of 0.777, approaching its performance on the validation set. Model interpretation with Grad-CAM revealed that MVI relevant imaging features on CT images were captured and learned by the ResNet-18 model. CONCLUSIONS: This framework provide evidence showing the generalizability and robustness of ResNet-18 in predicting MVI using CT images of AP scanned at multiple different hospitals. Attention heatmaps obtained from model explainability further confirmed that ResNet-18 focused on imaging features on CT overlapping with the conditions used by radiologists to estimate MVI clinically.

Targeting positive feedback between BASP1 and EGFR as a therapeutic strategy for lung cancer progression.

Rationale: Brain metastasis in patients with lung cancer is life-threatening. However, the molecular mechanism for this catastrophic disease remains elusive, and few druggable targets are available. Therefore, this study aimed to identify and characterize proteins that could be used as therapeutic targets. Methods: Proteomic analyses were conducted to identify differentially expressed membrane proteins between brain metastatic lung cancer cells and primary lung cancer cells. A neuronal growth-associated protein, brain acid soluble protein 1 (BASP1), was chosen for further investigation. The clinical relevance of BASP1 in lung adenocarcinoma was first assessed. Tyrosine kinase activity assays and in vitro and in vivo functional assays were conducted to explore the oncogenic mechanisms of BASP1. Results: The protein levels of BASP1 were positively associated with tumor progression and poor prognosis in patients with lung adenocarcinoma. Membrane-bound BASP1 increased EGFR signaling and stabilized EGFR proteins by facilitating their escape from the ubiquitin-proteasome pathway. Reciprocally, activation of EGFR recruited more BASP1 to the plasma membrane, generating a positive feedback loop between BASP1 and EGFR. Moreover, the synergistic therapeutic effects of EGFR tyrosine kinase inhibitor and arsenic trioxide led to a reduction in the level of BASP1 protein observed in lung cancer cells with acquired resistance to EGFR inhibitors. Conclusions: The reciprocal interaction between BASP1 and EGFR facilitates EGFR signaling in brain metastatic lung cancer. Targeting the newly identified BASP1-EGFR interaction could open new venues for lung cancer treatment.

Plasminogen/thrombomodulin signaling enhances VEGF expression to promote cutaneous wound healing.

Plasminogen (Plg) and thrombomodulin (TM) are glycoproteins well known for fibrinolytic and anticoagulant functions, respectively. Both Plg and TM are essential for wound healing. However, their significance during the reparative process was separately demonstrated in previous studies. Here, we investigate the interaction between Plg and epithelial TM and its effect on wound healing. Characterization of the wound margin revealed that Plg and TM were simultaneously upregulated at the early stage of wound healing and the two molecules were bound together. In vitro, TM silencing or knockout in keratinocytes inhibited Plg activation. Plg treatment enhanced keratinocyte proliferation and migration, and these actions were abolished by TM antibody. Keratinocyte-expressed vascular endothelial growth factor (VEGF), which presented a dose-response relationship with Plg treatment, can be suppressed by TM silencing. Moreover, treatment with VEGF antibody inhibited Plg-enhanced keratinocyte proliferation and wound recovery. In vivo, TM antibody treatment and keratinocyte-specific TM knockout can impede Plg-enhanced wound healing in mice. In high-glucose environments, Plg-enhanced VEGF expression and wound healing were suppressed due at least in part to downregulation of keratinocyte-expressed TM. Taken together, our findings suggest that activation of Plg/TM signaling may hold therapeutic potential for chronic wounds in diabetic or non-diabetic individuals. KEY MESSAGES: Plg binds to TM in cutaneous wound healing. TM facilitates the activation of Plg to Plm in keratinocytes. Epithelial TM regulates Plg-enhanced wound healing through VEGF expression.

ADAM9 promotes lung cancer progression through vascular remodeling by VEGFA, ANGPT2, and PLAT.

Lung cancer has a very high prevalence of brain metastasis, which results in a poor clinical outcome. Up-regulation of a disintegrin and metalloproteinase 9 (ADAM9) in lung cancer cells is correlated with metastasis to the brain. However, the molecular mechanism underlying this correlation remains to be elucidated. Since angiogenesis is an essential step for brain metastasis, microarray experiments were used to explore ADAM9-regulated genes that function in vascular remodeling. The results showed that the expression levels of vascular endothelial growth factor A (VEGFA), angiopoietin-2 (ANGPT2), and tissue plasminogen activator (PLAT) were suppressed in ADAM9-silenced cells, which in turn leads to decreases in angiogenesis, vascular remodeling, and tumor growth in vivo. Furthermore, simultaneous high expression of ADAM9 and VEGFA or of ADAM9 and ANGPT2 was correlated with poor prognosis in a clinical dataset. These findings suggest that ADAM9 promotes tumorigenesis through vascular remodeling, particularly by increasing the function of VEGFA, ANGPT2, and PLAT.

Thrombomodulin promotes diabetic wound healing by regulating toll-like receptor 4 expression.

Keratinocyte-expressed thrombomodulin (TM) and the released soluble TM (sTM) have been demonstrated to promote wound healing. However, the effects of high glucose on TM expression in keratinocytes and the role of TM in diabetic ulcer remain unclear. In this study, we demonstrated that expressions of TM and Toll-like receptor 4 (TLR4) were both downregulated in high-glucose cultured human keratinocytes and in skin keratinocytes of diabetic patients. In addition, the wound-triggered upregulation of TM and sTM production was abolished in both high-glucose cultured human keratinocytes and streptozotocin-induced diabetic mouse skin. Furthermore, supplementation of recombinant sTM could increase TLR4 expression and promote cutaneous wound healing in both high-glucose cultured human keratinocytes and diabetic mice. However, in Tlr4-deleted mice, which exhibited delayed wound healing, the therapeutic benefit of recombinant sTM was abrogated. Moreover, our results showed that tumor necrosis factor-α (TNF-α) expression in keratinocytes was dose-dependently upregulated by glucose, and TNF-α treatment downregulated the expression of TM and TLR4. Taken together, high-glucose environment reduces the expression of TM and TLR4 in keratinocytes possibly through the action of TNF-α, and recombinant sTM can increase the TLR4 expression and promote wound healing under diabetic condition.

FGFR1 mediates recombinant thrombomodulin domain-induced angiogenesis.

AIMS: The recombinant epidermal growth factor-like domain plus the serine/threonine-rich domain of thrombomodulin (rTMD23) promotes angiogenesis and accelerates the generation of activated protein C (APC), which facilitates angiogenesis. The aim of this study was to elucidate the molecular mechanisms underlying the angiogenic activity of rTMD23. METHODS AND RESULTS: We prepared rTMD23 and its mutants that did not possess the ability to promote APC generation and investigated their angiogenic activities in vitro and in vivo. rTMD23 mutants promoted proliferation, migration, and tube formation of human umbilical vein endothelial cells in vitro and induced neovascularization in vivo; these effects were similar to those exerted by wild-type rTMD23. To investigate its interaction with rTMD23, Type I fibroblast growth factor receptor (FGFR1) was precipitated along with syndecan-4 by rTMD23-conjugated Sepharose in human umbilical vein endothelial cells and FGFR1-expressing human embryonic kidney 293 cells. Additionally, the kinetics of the interaction between rTMD23 and FGFR1 were analysed using surface plasmon resonance. rTMD23-induced FGFR1 activation and tube formation were inhibited by an FGFR1-specific tyrosine kinase inhibitor, PD173074, or by knockdown of FGFR1 using siRNA technology. We observed an improvement in rat hindlimb recovery in an ischaemic model following rTMD23 treatment, and this was associated with increased neovascularization and FGFR1 phosphorylation. CONCLUSION: rTMD23 induced angiogenesis via FGFR1, a process that is independent of the APC pathway.

Thrombomodulin functions as a plasminogen receptor to modulate angiogenesis.

Urokinase-type plasminogen activator (uPA) activates plasminogen (Plg) through a major pericellular proteolytic system involved in cell migration and angiogenesis; however, the Plg receptor that participates in uPA-mediated Plg activation has not yet been identified. In this study, we demonstrated that thrombomodulin (TM), a type I transmembrane glycoprotein, is a novel Plg receptor that plays a role in pericellular proteolysis and cell migration. Plg activation at the cell surface and the extent of its cell migration- and invasion-promoting effect are cellular TM expression dependent. Direct binding of Plg and the recombinant TM extracellular domain, with a KD of 0.1-0.3 µM, was determined through surface plasmon resonance analysis. Colocalization of TM, Plg, and the uPA receptor within plasma membrane lipid rafts, at the leading edge of migrating endothelial cells, was demonstrated and was also shown to overlap with areas of major pericellular proteolysis. Moreover, the roles of TM and Plg in neoangiogenesis were demonstrated in vivo through the skin wound-healing model. In conclusion, we propose that TM is a novel Plg receptor that regulates uPA/uPA receptor-mediated Plg activation and pericellular proteolysis within lipid rafts at the leading edge of migrating cells during angiogenesis.

Human plasminogen kringle 1-5 inhibits angiogenesis and induces thrombomodulin degradation in a protein kinase A-dependent manner.

Kringle 1-5 (K1-5), an endogenous proteolytic fragment of human plasminogen (Plg), is an angiostatin-related protein that inhibits angiogenesis. Many angiostatin-related proteins have been identified, but the detailed molecular mechanisms underlying their antiangiogenic effects remain unclear. Thrombomodulin (TM) is a transmembrane glycoprotein that plays a major role in the anticoagulation process in endothelial cells. Previously, we demonstrated that recombinant TM could interact with Plg to enhance Plg activation. In the present study, we investigated the interaction between TM and K1-5, and their functions in endothelial cells. We found that K1-5 colocalized with TM and directly interacted with TM through the TM lectin-like domain. After K1-5 interacted with TM, it induced TM internalization and degradation. In addition, the K1-5-induced TM internalization and degradation in proteasomes after ubiquitin modification were dependent on protein kinase A (PKA). Moreover, a PKA-specific inhibitor reversed the effects of K1-5 on cell migration and tube formation. Consistent with these findings, TM overexpression resulted in increased cell migration; moreover, K1-5 inhibited the increase of TM-mediated cell migration in a PKA-dependent manner. We determined that TM acts as a K1-5 receptor and that K1-5 induces TM internalization, ubiquitination, and degradation through the PKA pathway, by which K1-5 may inhibit endothelial cell migration and tube formation.

Thrombomodulin regulates keratinocyte differentiation and promotes wound healing.

The membrane glycoprotein thrombomodulin (TM) has been implicated in keratinocyte differentiation and wound healing, but its specific function remains undetermined. The epidermis-specific TM knockout mice were generated to investigate the function of TM in these biological processes. Primary cultured keratinocytes obtained from TM(lox/lox); K5-Cre mice, in which TM expression was abrogated, underwent abnormal differentiation in response to calcium induction. Poor epidermal differentiation, as evidenced by downregulation of the terminal differentiation markers loricrin and filaggrin, was observed in TM(lox/lox); K5-Cre mice. Silencing TM expression in human epithelial cells impaired calcium-induced extracellular signal-regulated kinase pathway activation and subsequent keratinocyte differentiation. Compared with wild-type mice, the cell spreading area and wound closure rate were lower in keratinocytes from TM(lox/lox); K5-Cre mice. In addition, the lower density of neovascularization and smaller area of hyperproliferative epithelium contributed to slower wound healing in TM(lox/lox); K5-Cre mice than in wild-type mice. Local administration of recombinant TM (rTM) accelerated healing rates in the TM-null skin. These data suggest that TM has a critical role in skin differentiation and wound healing. Furthermore, rTM may hold therapeutic potential for the treatment of nonhealing chronic wounds.