Inflammatory Stroke Extracellular Vesicles Induce Macrophage Activation
Background and Purpose—Extracellular vesicles (EVs) are protein–lipid complexes released from cells, as well as actively exocytosed, as part of normal physiology, but also during pathological processes such as those occurring during a stroke. Our aim was to determine the inflammatory potential of stroke EVs.
Methods—EVs were quantified and analyzed in the sera of patients after an acute stroke (<24 hours; OXVASC [Oxford Vascular Study]). Isolated EV fractions were subjected to untargeted proteomic analysis by liquid chromatography mass-spectrometry/mass-spectrometry and then applied to macrophages in culture to investigate inflammatory gene expression.
Results—EV number, but not size, is significantly increased in stroke patients when compared to age-matched controls. Proteomic analysis reveals an overall increase in acute phase proteins, including C-reactive protein. EV fractions applied to monocyte-differentiated macrophage cultures induced inflammatory gene expression.
Conclusions—Together these data show that EVs from stroke patients are proinflammatory in nature and are capable of inducing inflammation in immune cells.
Inflammation plays a crucial role in the pathophysiology of stroke. Within the core of the infarct neurons rapidly depolarize and die, and these cells are phagocytosed by microglia and infiltrating circulating macrophages. Propagation of deleterious signals and processes from within the core of the infarct, to the circulation can result in the infiltration of systemic immune cells, significantly worsening infarct evolution.1
Many mechanisms are purported to communicate central nervous system (CNS) injury signals to the periphery,2 the most recent of which is the circulating extracellular vesicle (EV). EV is an umbrella term encompassing fragments blebbed from the membrane (microvesicles) and actively exocytosed vesicles (exosomes).
Although there is an increasing presence of EVs in the literature, the focus has to date been on quantification and association, rather than on investigation of structure and function. Their role as novel biomarkers in cancer,3 as well as in metabolic4 and cardiovascular disease,5 has begun to be explored, but there remains a paucity of knowledge on their specific role in CNS disease, in particular acute CNS injury. On this background, we aimed to determine the potential for EVs to communicate injury and inflammatory signals after a stroke.
Materials and Methods
Thirty-eight patient samples were selected from the OXVASC (Oxford Vascular Study) cohort—a population-based study of all acute vascular events in ≈92 000 residents of Oxfordshire.6 EVs were isolated using ultracentrifugation techniques and analyzed using nanosight tracking analysis, electron microscopy, and Western blotting for canonical EV markers (Tsg101, Alix, and CD9).7 Proteomics was performed using untargeted liquid chromatography mass-spectrometry/mass-spectrometry and pathway analysis (Ingenuity Pathway Analysis; Qiagen). To determine functionality, human THP-1 cells were treated with stroke and control EVs and mRNA expression levels of tnf, il-1β, cxcl-1, and ccl2 were studied by quantitative polymerase chain reaction. Detailed Methods are available in the online-only Data Supplement. All research was conducted according to the principles of the Declaration of Helsinki.
Acute Stroke Results in Increased Numbers of EVs that Associate With C-Reactive Protein and Neuron-Specific Enolase
Patients from the OXVASC study had a median age of 75.4 years (interquartile range, 68–80 years), had a minimum National Institute of Health Stroke Scale score of 5, and were age and sex matched with controls. Nanosight tracking analysis of EVs showed a standard distribution with an elevated number, but not size, in patients with stroke (Figure 1A). Characterization revealed a mixed population of vesicles expressing canonical vesicle markers (Figure 1B). Number of EVs increased in stroke compared to age-matched controls (Figure 1C). Correlation analysis showed a significant correlation between EV number and C-reactive protein (Figure 1D).
EVs From Stroke Patients Have an Inflammatory Proteomic Profile
Of the 381 proteins found, 75 were consistently and significantly different in stroke EVs, when compared to age-matched controls, and hierarchical cluster analysis revealed an overall increase in inflammatory proteins (Figure 2A; C-reactive protein is an example highlighted in Figure 2B and confirmed with Western blot in Figure 2C). Pathway analysis (Ingenuity; Qiagen, United Kingdom) performed on average-fold change showed an overall upregulation of proteins associated with the acute-phase response including C-reactive protein (Figure 2D).
Stroke EVs Activate Macrophages
EVs applied to monocyte-differentiated macrophages resulted in an overall increase in tumor necrosis factor (TNF) mRNA expression, which was different in cells treated with EVs (Figure 3A). This was reflected in an increase in TNF mRNA expression in cells treated with EVs from stroke patients compared to both EV controls and untreated controls. Interleukin-1β mRNA expression was changed after EV treatment (Figure 3B), resulting in an increase in interleukin-1β expression in stroke EV-treated cells compared to both untreated controls and age-matched EV controls. Neutrophil chemoattractant CXCL-1 showed the most profound increase in mRNA expression, a 400-fold increase in stroke-EV–treated cells compared to untreated controls (Figure 3C). Levels of CXCL-1 mRNA were increased in stroke-EV-treated cells when compared to control EVs and when compared to untreated controls. Monocyte chemoattractant CCL-2 mRNA was also altered by treatment with EVs (Figure 3D). Post hoc testing revealed that CCL-2 mRNA increased after cells were challenged with stroke EVs compared to when they were challenged with control EVs and compared to untreated cells.
In this brief report, we have demonstrated, for the first time, that EVs released into the circulation after a stroke contain proinflammatory proteins and are capable of inducing an inflammatory response. Stroke patients show increased levels of circulating EVs, which are a mixed population of large and small vesicles expressing canonical vesicle markers and inflammatory proteins. The mechanisms of communication between the CNS and the systemic immune system after stroke remain poorly understood. Insights such as these into the fundamental mechanisms of this communication pathway are vital to further early interventional studies in stroke.
Studies on the EV population in stroke patients are still relatively sparse and focus largely on their thrombogenic activity.8,9 Kuriyama et al10 have suggested that their potential as either a point of intervention or as a biomarker is limited; however, this is likely to be because of the field still being in its infancy. Here, we have used dynamic light scattering to demonstrate a significant increase in EV numbers post stroke, which is not correlated with stroke severity (Supplementary I in the online-only Data Supplement), and have also demonstrated the presence of standard vesicle markers. To our knowledge, this is the first stroke article on EVs to present data in this manner since the publication of the ISEV guidelines (International Society for Extracellular Vesicles) on EVs.7
An increased number of EVs does not necessarily indicate that the vesicles are physicochemically different between the control and the stroke patients, or indeed whether the vesicles are functional. It was important, therefore, for us to establish a functional role for what has previously been considered cellular debris. To date, studies have found elevated numbers of platelet-derived microparticles in the acute phase of cerebral infarction and suggested that their prothrombotic nature may result in infarct development without shedding further light on the precise molecular mechanisms.4,9 Our data here suggest that the profile of EVs from stroke patients is different from that of age-matched controls. Considering the average age of stroke patients, and the association of age with an altered inflammatory profile,11 these data suggest that the changes in the EV population reflect a disease-specific process. However, it should be noted these patients were not disease-matched and comorbidities in the elderly may confound the interpretation of the data and remain the focus of ongoing work. Despite this, the proteomics data here show that there is a significant acute phase profile in the EVs from stroke patients, and when these were applied to human immune cells, they caused a significant increase in cytokine and chemokine expression when compared to EVs from control patients. It is likely that EVs in stroke patients reflect the proinflammatory nature of the stroke and as such can act as a mechanism for signaling CNS injury to the periphery. Although the mechanisms of this may be fundamental to many disease processes, including other ischemic insults such as myocardial infarction, the origins and functions of the EV population are likely to be unique, and the purpose of this remains the focus of our ongoing research.
In conclusion, these data demonstrate a change in the fundamental components of circulating EVs after a stroke and a significant functional role for them in activating immune cells. Although this research is preliminary, it provides a promising candidate route for CNS to immune system communication after brain injury and paves the way for significant advances in the fields of both neuroimmunology and EV research.
We would like to acknowledge the assistance of Dr Errin Johnson of the Dunn School Bioimaging Facility at the University of Oxford. Mass spectrometry analysis was performed in the TDI MS Laboratory led by Benedikt M. Kessler.
A.A. Neuhaus received Medical Research Council funding as part of the Radcliffe Department of Medicine Scholarship Programme. The other authors report no conflicts.
↵* Drs Couch and Buchan share senior authorship.
Guest Editor for this article was Miguel Perez Pinzon, PhD.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.117.017236/-/DC1.
- Received March 3, 2017.
- Revision received April 4, 2017.
- Accepted April 14, 2017.
- © 2017 American Heart Association, Inc.
- Maestrini I,
- Strbian D,
- Gautier S,
- Haapaniemi E,
- Moulin S,
- Sairanen T,
- et al
- Loyer X,
- Vion AC,
- Tedgui A,
- Boulanger CM
- Rothwell PM,
- Coull AJ,
- Giles MF,
- Howard SC,
- Silver LE,
- Bull LM,
- et al
- Lötvall J,
- Hill AF,
- Hochberg F,
- Buzás EI,
- Di Vizio D,
- Gardiner C,
- et al
- Cherian P,
- Hankey GJ,
- Eikelboom JW,
- Thom J,
- Baker RI,
- McQuillan A,
- et al