BMVBS 2026

Background
Heparan sulphate (HS) is a major component of the endothelial glycocalyx, which supports vascular integrity and cardiac function. However, the specific role of endothelial HS in cardiac function and its involvement in ischemia–reperfusion (IR) injury remains unclear.

Methods
Transgenic mice with endothelial-specific Ext1 knockdown (KD), disrupting HS elongation, were generated. Ext1 deficiency was confirmed by RT-qPCR, and HS loss was confirmed by HS staining. Heart structure and function were evaluated via echocardiography. Myocardial oedema was measured using the wet-to-dry ratio of dissected heart tissues. Cardiac function in isolated hearts was examined through Langendorff perfusion, and infarct size was assessed with 2,3,5-triphenyltetrazolium chloride staining after an IR challenge (30mins global ischemia and 120mins reperfusion).

Results
Three weeks after tamoxifen injection, Ext1^f/f Cre^+/- mice exhibited significantly reduced Ext1 mRNA expression (p=0.01) and reduced HS immunoreactivity associated with cardiac microvessels (p=0.044) compared to Ext1^f/f Cre^-/- mice, consistent with endothelial HS deficiency. Echocardiography demonstrated a thicker left ventricular (LV) wall (p=0.036), a smaller LV volume (p=0.04), and LV mass trending upwards (p=0.06) in Ext1^f/f Cre^+/- mice, suggesting cardiac hypertrophy. Additionally, an elevated cardiac wet-to-dry ratio (p=0.003) in Ext1^f/f Cre^+/- mice demonstrates myocardial oedema formation. Other in vivo heart function indices were not significantly different between groups. However, Ext1^f/f Cre^+/- mouse hearts had significantly lower developed pressure (p=0.01) and reduced dP/dt(max) (p=0.005) on an ex vivo Langendorff preparation. Following IR injury, Ext1^f/f Cre^+/-  hearts exhibited worse functional recovery across several parameters (developed pressure, p=0.0001; Rate Pressure Product, p=0.025; dP/dt(max), p=0.0016; dP/dt(min), p=0.006) and larger infarct size (p=0.025).

Conclusion
Endothelial Ext1 KD–mediated HS deficiency worsens cardiac structure and function and substantially impairs recovery from IR injury. These findings indicate that endothelial HS is crucial for maintaining myocardial performance and limiting IR injury, suggesting endothelial HS preservation is a potential therapeutic approach.

Endothelial heterogeneity is well recognised across organs and vessel types, where endothelial cells (ECs) adopt specialised phenotypes adapted to local demands. However, we have shown that heterogeneity also persists within confluent endothelial monolayers under controlled in vitro conditions. Using medium-throughput screening combined with high-content spatial image analysis, we quantified single-cell variability in junction organisation, proliferative state, and signalling readouts, including NOTCH pathway activity. These measurements revealed multimodal, non-Gaussian, and spatially organised patterns, indicating structured phenotypic diversity within an apparently uniform population.

While statistical analysis clearly documents this variability, it does not explain how it arises. Observing multiple peaks or spatial clustering in the data tells us that heterogeneity is present, but not whether it reflects stable subtypes, random fluctuation, or dynamic switching between states. We therefore asked whether intrinsic, contact-dependent signalling could generate such diversity.

We focused on NOTCH signalling, a short-range cell–cell communication pathway known to produce nonlinear behaviours, including stable alternative states and oscillations, in developmental systems like neurogenesis and somitogenesis. We confirmed that heterogeneous nuclear HES1 expression is not restricted to culture but is also present in murine aortic endothelium, supporting physiological relevance of these observations.

To investigate the underlying mechanism, we developed a spatially explicit computational model of an endothelial monolayer using the Cellular Potts framework (CompuCell3D), allowing realistic cell geometries and contact areas to influence signalling strength. Within each simulated cell, established NOTCH regulatory interactions, lateral inhibition, lateral induction, cis interactions, and feedback, were modelled using ordinary differential equations, with key parameters constrained by quantitative single-cell measurements.

Systematic model exploration identified specific signalling regimes, particularly those involving productive cis-interactions, that reproduce the experimentally observed spatiotemporal HES1 heterogeneity through asynchronous, contact-dependent dynamics.

Together, these findings show how quantitative phenotyping coupled with mechanistic modelling can guide hypothesis generation and support a deeper understanding of endothelial heterogeneity.

Two opposing and concurrent clinical syndromes characterise the immune and inflammatory response to major traumatic injury: a pro-inflammatory systemic inflammatory response syndrome (SIRS) and a counteracting compensatory anti-inflammatory response syndrome (CARS). A timely and appropriate CARS response is critical for the restoration of immune homeostasis, whilst a prolonged and/or exaggerated CARS response leads to a state of profound systemic immunosuppression thar predisposes patients to poor clinical outcomes. Focussing on the anti-microbial functions of neutrophils and monocytes, our work investigates how quickly following major trauma innate immune dysregulation develops and the mechanisms that contribute to its development.

The RBC surface is covered by a complex glycocalyx (RBCGlx). In health, intravascular neutrophil hyperactivity is associated with vascular damage. We hypothesise that RBCGlx is damaged early in sepsis, allowing unregulated intravascular neutrophil activation leading to organ failure.

RBCGlx was quantified in patients with sepsis and controls using our novel peak to peak assay (n=30). Control and sepsis serum were added to Click-iT™ManNAz-labelled human glomerular endothelial cells (GEnC) and glycocalyx loss was quantified (blind) using confocal microscopy (n=30). Primary neutrophils were isolated from controls and incubated with control or glycocalyx depleted RBCs (neuraminidase, heparinase or chondroitinase treated). Neutrophil activation was quantified by PMA induced oxidative burst and lipopolysaccharide (LPS) induced interleukin-8 (IL8) production.

RBCGlx was significantly reduced in sepsis (0.06 +/- 0.03 v 0.10 +/- 0.01 micrometers p<0.0001). In vitro, exposure of ManNaz labelled human GEnC to sepsis serum reduced glycocalyx 1.5-fold (p<0.05).

5:1 control RBC:neutrophils inhibited neutrophil oxidative burst (mean difference=0.31, 95% CI: 0.15-0.46, adjusted p=0.0078, n=4) and 24-hour IL8 production (mean difference=0.20, 95% CI 0.12–0.29, adjusted p=0.00040, n=8). Neuraminidase removed sialic acid from RBCs which maintained their suppressive effect on IL8 (mean difference=0.21, 95% CI 0.09–0.34, adjusted p=0.0035, n=8), and ROS production (mean difference=0.26, 95% CI 0.14-0.38, p=0.0065, n=4). However, heparinase and chondroitinase removed RBC inhibition of neutrophil IL8 production (all adjusted p>0.05, n=8).

50:1 RBCs:neutrophils also suppressed neutrophil IL8 at 6 hours, but again neuraminidase did not remove this effect (mean difference=0.46, 95% CI 0.39–0.54, adjusted p=0.0014 and mean difference=0.44, 95% CI 0.087–0.79, adjusted p=0.033, n=3). However, neutrophil suppression was lost when RBC were treated with heparinase or chondroitinase (all adjusted p>0.05) suggesting RBC glycosaminoglycans are important for neutrophil inhibition.

RBCGlx is damaged in human sepsis, amplifying the sepsis cascade via intravascular neutrophil activation. RBCGlx glycosaminoglycan preservation could be a novel therapeutic target to limit sepsis progression.

Introduction
Myocardial infarction (MI) triggers a systemic inflammatory response that extends beyond the heart. The pulmonary microcirculation, with its dense capillary network and resident immune population, may be particularly susceptible to secondary injury. Neutrophils and neutrophil extracellular traps (NETs) are central mediators of lung injury in chronic obstructive pulmonary disease (COPD), yet their role in post-MI pulmonary damage remains poorly defined. We hypothesised that MI induces neutrophil- and NET-driven pulmonary microvascular injury comparable to that observed in COPD.

Methods
Human lung tissue from MI and COPD patients was analysed using immunostaining for neutrophils, NET-associated markers, and endothelial activation. Circulating inflammatory mediators were quantified by multiplex assay. In parallel, mice underwent myocardial ischaemia-reperfusion injury (IRI) to model MI, or intranasal elastase and rhinovirus administration to model COPD exacerbations. Lungs were harvested for collagenase digestion and flow cytometry or processed for immunofluorescence analysis.

Results
MI significantly increased pulmonary neutrophil infiltration and NET formation in both human and mouse lung tissue. Endothelial activation markers and oxidative damage were upregulated, consistent with microvascular injury. Notably, neutrophil-derived and NET-associated S100A8/A9 expression in post-MI lungs reached levels comparable to those observed in COPD. Myocardial IRI disrupted the pulmonary immune microenvironment, with expansion of activated neutrophil subsets and enhanced inflammatory signalling.

Conclusions
These findings identify the lung as a major remote target of MI-induced inflammation. MI upregulates neutrophil- and NET-associated S100A8/A9 to levels comparable to COPD, highlighting a shared inflammatory axis between myocardial injury and chronic lung disease. Neutrophil-driven pulmonary microvascular injury may represent an under-recognised contributor to post-MI morbidity.

Artificial intelligence tools have become increasingly visible across the research landscape. Many researchers will have encountered large language models in some capacity - whether through direct use, institutional discussions, or the growing presence of AI-assisted tools in publishing and literature search. However, the tools available in 2026 have moved considerably beyond the conversational chatbot format, and the pace of development has made it difficult for researchers to keep track of what has changed.

Modern AI systems can connect directly to a researcher's existing tools - cloud storage, email, reference managers - through open protocols such as the Model Context Protocol (MCP), removing the need to manually transfer information between platforms. Features like persistent memory allow these tools to retain context about a researcher's field, terminology, and ongoing projects across sessions, building a working relationship that reduces the repetitive setup which has historically limited their usefulness. Perhaps most significantly, agentic capabilities now allow AI to decompose complex tasks into subtasks and execute them with minimal oversight - from screening thousands of papers against defined criteria and returning structured summaries, to running statistical analyses on uploaded datasets and producing publication-ready figures. For researchers without a programming background, these code execution environments lower a barrier that has long separated them from computational methods.

Given this rapidly shifting landscape, there is a need for researchers to understand not just what AI can do, but how to use it effectively and responsibly. Emerging evidence suggests that while AI adoption substantially boosts individual productivity and citation rates, it may collectively narrow the diversity of scientific inquiry and reduce cross-disciplinary collaboration. These are important considerations as our community begins to integrate these tools into daily research practice. Practical, low-barrier strategies for getting started - alongside an honest appraisal of the limitations - are presented, with a focus on what is most relevant to vascular and microvascular researchers.

The Avon Longitudinal Study of Parents and Children (ALSPAC) cohort is a large population study conducted on pregnant women, their partners (G0) and children (G1) within Avon, UK. As part of the follow ups conducted on these individuals, proxy measures for the endothelial glycocalyx were taken at the approximate 30-year timepoint for G1 participants, together with G0. The aim of this study was to investigate the relationship between glycocalyx measures and sociodemographic factors within a large population dataset.

The GlycoCheck^TM system uses sidestream dark field imaging of sublingual microvessels, along with custom software, to measure the perfused boundary region (PBR) which is inversely related to glycocalyx depth.

GlycoCheck^TM was applied to 6,266 individuals in the @30 clinic, making this the largest population cohort with PBR measurements to date.

Preliminary results suggest that there are significant relationships between PBR, age and body mass index (BMI), across various stratifications including vessel size, gender, and ALSPAC generations (G0, G1). In line with current literature, we see a significant inverse relationship between PBR and BMI (R²=0.0014, p<0.05), suggesting that the higher the BMI, the thicker the glycocalyx, similar to the previous studies into BMI and PBR.¹ This finding is further evidenced when stratifying by vessel size, with the PBR of vessels sized 4-9μm (R²=0.0025, p<0.05), 10-19μm (R²=0.0014, p<0.05), and 20-25μm (R²=0.0057, p<0.005) all showing a negative association with PBR. When stratified by generational cohort we also see a positive relationship between age and PBR in G1 (R²=0.022, p<0.005).

In conclusion, we have observed PBR and sociodemographic relationships replicated in large population cohorts that were previously only modelled in smaller observational studies.

The endothelial glycocalyx is composed primarily of glycosaminoglycans (GAGs) and is a known modulator of vascular permeability and renal function, with known disruption in diseases such as diabetes and sepsis. Here we aim to characterise how the composition of this glycocalyx varies both within a healthy patient cohort and as a consequence of chronic kidney disease. To do this, we have utilized time of flight secondary ion mass spectrometry (ToF-SIMS), a form of imaging mass spectrometry that enables concurrent spatial and compositional GAG analysis. We have previously shown that the technique can discriminate between nanogram quantities of individual GAGs, developed a library of ions discriminatory of GAGs in-situ and their detection in renal tissue.  This study has used ToF-SIMS for the analysis of individual glomeruli from deceased organ donor tissue (N=20) and patients with chronic kidney disease (CKD)(N=10). Renal GAG composition was found to vary significantly with sex, age, type of death & pregnancy status in the organ donor population and as a result of CKD. This work suggests that population demographics can influence an individual’s matrix, particularly in the kidney and that this should be considered when targeting this complex environment as a disease treatment or potential biomarker. Future work will expand the considered population in both scale and diversity, including the analysis of a range of renal diseases (including diabetes and Alport syndrome) to determine how specific the observed changes in CKD are and whether GAGs could serve as a disease specific biomarker. To enable more direct translation to clinic, urinary and serum GAG content will also be assessed to provide more accessible tissue sources and determine if this variation if tissue specific.

Lung microvasculature is a common site for the seeding of metastatic breast cancer. This organotropism is created by the primary tumour, which remodels the lung niche into an immunosuppressive and pro-inflammatory space before tumour cell colonisation. While platelets are typically involved in thrombosis and haemostasis, they can also promote haematological metastasis through direct interactions with circulating tumour cells. However, the role of platelets within a "pre-metastatic" lung microenvironment remains understudied.

To model pre-metastatic lungs, 4T1 breast cancer cells were injected into the mammary fat pads of BALB/cJ mice, which develop pre-metastatic and then metastatic lung disease after 10 and 21 days, respectively. Increased neutrophil and macrophage populations were detected within lungs before metastatic outgrowth, implying the generation of a pre-metastatic niche, while intravital two-photon microscopy demonstrated that platelets where also sequestered within lungs capillary beds prior to established metastases.

To test if platelets are active contributors to the pre-metastatic niche, 4T1 mammary tumours were generated in mice with normal (Thpo^+/+) or low (Thpo^-/-) platelet counts. While low platelet counts had no effect on mammary tumour growth, platelet deficiency led to fewer neutrophils and macrophages within pre-metastatic lungs, with an increased proportion of CD4⁺ lymphocytes.

Our data show that platelets are sequestered within pre-metastatic lung microvasculature where they recruit pro-inflammatory myeloid-derived suppressor cells known to support subsequent intravascular metastasis. Ongoing experiments involve establishing a microfluidic model of tumour-induced thromboinflammation to mechanistically investigate 1) how tumours activate distant endothelial cells, 2) how platelets are tethered to activated endothelium, and 3) how platelets recruit myeloid cells to inflamed lung vasculature. Future work will also test if anti-platelet medications can mitigate tumour-driven thromboinflammation and the generation of a (pre)metastatic niche.

A hallmark of severe COVID-19 is an increase in immune cell infiltration into sites of infection. This is a process highly regulated by endothelial cells which co-ordinate chemokine presentation and adhesion signalling. Trans-endothelial migration of immune cells is separated into distinct phases with the firm adhesion stage being mediated by endothelial ICAM-1 and immune cell adhesion receptors in the high-affinity conformation; for example, LFA-1 present on T-lymphocytes. Direct SARS-CoV-2 protein interactions with host cells may influence transendothelial migration of T-lymphocytes: the SARS-CoV-2 ORF7a accessory protein can interact with T-lymphocyte LFA-1 and we hypothesise this to promote T-lymphocyte-endothelial firm adhesion contributing to COVID-19 pathogenesis.

T-lymphocytes were isolated from healthy donors’ whole blood through negative selection removing cell populations expressing removing cells expressing CD14, CD15, CD16, CD19, CD34, CD36, CD56, CD123, and CD235a. T-lymphocyte purity was checked, and T-lymphocytes were subsequently activated overnight (10ng/mL PMA, 1ug/mL Ionomycin). T-lymphocytes were labelled with Calcein-AM before added to a  HUVEC monolayer and incubated with SARS-CoV-2, backbone pseudoviruses, pseudoviruses-ORF7a or recombinant ORF7a. After removing non-bound T-lymphocytes, adherence of T-lymphocytes to endothelial cells were quantified through fluorescence intensity measurements taken by a FLUOstar spectrophotometer (485nm/525nm excitation/emission).

Co-treatment of endothelial cells and T-lymphocytes with SARS-CoV-2 increased T-lymphocyte firm adhesion by 61.8% ± 11.9% (n=4, SEM), co-treatment with pseudoviruses-ORF7a increased firm adhesion by 16.0% ± 4.1% (n=5, SEM) and co-treatment with recombinant ORF7a increased firm adhesion by 47.3% ± 9.4% (n=5, SEM). These data suggest ORF7a to promote T-lymphocyte-endothelial cell binding which can happen both with recombinant protein and when ORF7a is incorporated into virions. This increased binding mediated by ORF7a may contribute to the inflammatory environment observed during severe COVID-19.

Introduction: The lung microvasculature is a unique endothelial network that supports gas exchange and regulates immune cell trafficking. The airways are continuously exposed to external threats. Immune-vascular crosstalk is therefore central to immune training, enabling effective defence while preserving barrier integrity. Recent single-cell spatial transcriptomics analyses from our laboratory identified two pro-inflammatory transitional endothelial states in human lung (Joulia et al., 2025). However, the molecular regulators of this organisation and the contribution of these spatial states on local immune cell remain unclear.

Methods: We integrated single-cell spatial transcriptomics with quantitative neighbourhood analysis based on Delaunay triangulation to map endothelial–immune interactions. Neighbourhoods were based on the number of interactions endothelial cells present with their close partners. After defining local cellular ecosystems, we investigated local inflammatory gene expression and their influence on immune cells using ligand-receptor based analysis. 

Results: Unsupervised analysis revealed 4 neighbourhoods within non-diseased human lung with one niche strongly enriched in pro-inflammatory endothelial cells. This specific endothelial state is characterised by increased Atypical Chemokine Receptor 1 (ACKR1), known to retain chemokines locally and interleukin 33 (IL-33), critical regulator of mucosal immune response. This cellular ecosystem was able to influence resident immune cells such as mast cells, expressing IL-33 receptor, and macrophages, consistent with heightened immune readiness.

Conclusions: In summary, we believe that the lung microvasculature can specialise specific areas by modifying endothelial cell phenotype to shape immune response. Defining these niches provides a framework for understanding immune–vascular crosstalk and establishes a baseline for future disease-focused studies.

Background

Galectin-9 (Gal-9), a glycan-binding lectin in the galectin family with immunomodulatory functions, is known to drive vascular inflammation. Endothelium secretes Gal-9, and its upregulation during inflammatory conditions has been implicated in a range of diseases, including thrombotic disorders. Our recent data indicate that Gal-9 drives platelet activation through GPVI and CLEC-2 receptors. However, the functional impact of Gal-9-mediated platelet activation in venous thrombosis remains unknown.

 Aims

To investigate the functional impact of Gal-9-mediated platelet activation and how it contributes to the development of venous thrombosis.

 Methods

We measured plasma levels of Gal-9 in a cohort of patients with deep vein thrombosis (DVT) and in healthy controls. Using a whole blood microfluidic model, we assessed platelet-mediated thrombus formation over immobilised Gal-9 or endothelial cells overexpressing Gal-9. Co-culture of Gal-9-stimulated human platelets with neutrophils was performed to evaluate neutrophil extracellular traps (NETs) formation (MPO and CitH3 immunofluoroscence staining). Thrombus development was further investigated in Gal-9 knockout (KO) mouse models following inferior vena cava (IVC) stenosis.

 Results

DVT patients at baseline (n=60) had significantly higher plasma Gal-9 levels than healthy controls (n=30) (p<0.001), and these levels markedly decreased at follow-up (n=59) (p=0.001). Immobilised Gal-9 in combination with collagen or vWF significantly accelerated platelet activation and thrombus formation at venous shear, and this effect was markedly reduced by a Gal-9 inhibiting antibody. Moreover, whole blood perfusion over endothelial cells with high Gal-9 expression following stimulation with TNF-α and IFN-γ led to significantly increased platelet adhesion and fibrin deposition. In addition, Gal-9 stimulated platelets induced NETosis. In vivo, Gal-9 KO mice showed significantly reduced thrombus weight and length and exhibited lower thrombus prevalence compared to WT mice(p<0.0001) following IVC stenosis.

 Conclusion

Gal-9 may represent a novel therapeutic target to modulate platelet activation to protect from venous thrombosis.

Introduction: Respiratory diseases remain a leading cause of global morbidity, often characterised by the loss of endothelial barrier integrity. The bitter taste receptor T2R14 is expressed in the lung microvascular endothelium, where its activation has been shown to disrupt barrier function. However, the therapeutic potential of T2R14 antagonism in preventing or reversing lung injury remains unexplored. This study investigates the protective effects of a novel T2R14 antagonist, LW129, against inflammatory and growth factor-induced endothelial dysfunction.

Research Question: To investigate the role of the novel T2R14 antagonist, LW129, in protecting against inflammatory and growth factor-induced endothelial dysfunction within primary human pulmonary microvascular endothelial cells (HPMECs).

Methods: HPMECs were utilised to assess barrier integrity. Endothelial permeability was quantified using Trans-endothelial Electrical Resistance (TEER) and FITC-Dextran Transwell permeability assays. Cells were challenged with LPS or VEGF in the presence or absence of LW129 to evaluate both barrier protection and the promotion of barrier recovery.

Results: LW129 demonstrated a significant protective effect against both LPS and VEGF-induced barrier injury when administered concurrently. Specifically, TEER measurements remained significantly higher, and FITC-Dextran flux was reduced compared to controls. Furthermore, LW129 was found to accelerate the recovery of barrier function following LPS-induced injury, suggesting a role in promoting endothelial stability.

Conclusion: These findings demonstrate that T2R14 antagonism via LW129 effectively preserves pulmonary endothelial barrier integrity against both inflammatory and growth-factor mediated insults. By preventing paracellular leakage and enhancing barrier recovery, T2R14 represents a promising novel therapeutic target for pathologies defined by acute vascular dysfunction and pulmonary oedema such as ARDS.

Acknowledgments: Prof Nicholas Pugh, Dr Clett Erridge, Dr Mareike Posner, Dr Joseph Bird, Cambridge Philosophical Society