STEM CELLS Translational Medicine

State‐of‐the‐Art Stroke Treatment

Ischemic stroke treatment currently involves three concepts: The admission of stroke patients to stroke units, the application of thrombolytics, and the recanalization of the occluded vessel by endovascular clot removal 1-4. With the first stroke units being introduced in the 1990s, stroke management has turned from a purely observational field toward an evidence based therapeutic field. Controlled randomized studies not only demonstrated the utility of the thrombolytic recombinant tissue plasminogen activator to improve stroke outcome when administered intravenously within 4.5 hours after symptom onset 5, but more recently revealed the efficacy of endovascular recanalization therapy 1, 2. Despite this great success, the majority of patients receive none of the two aforementioned treatments, partially because of narrow time windows or because of significant complication risks. This justifies the need for additional treatments, which alleviate the long‐term consequences of a stroke.

Post‐Stroke Brain Repair

With strategies on brain protection having failed in clinics in the 1980s and 1990s, current preclinical research strongly focuses on promoting the regenerative capacities of the ischemic brain. The physiological basis of the latter is the persistence of endogenous neurogenesis in the adult mammalian brain within so called stem cell niches, namely the subventricular zone (SVZ) of the lateral ventricles 6-8 and the subgranular zone of the dentate gyrus 9, 10. Upon stroke, neural progenitor cells (NPCs) within the SVZ migrate toward the ischemic lesion site where they proliferate 11, 12. Yet, the stroke‐induced promotion of post‐stroke neurogenesis has restricted functional relevance, as new‐born cells show both low survival rates and poorly differentiate into mature neurons 13-15.

In order to use the endogenous regenerative potential of the ischemic brain, two different strategies to manipulate neurogenesis are under investigation: (a) enhancing the resistance of NPCs to delayed degeneration and (b) augmenting the number of NPCs in the ischemic brain tissue. The former can be achieved by the administration of antiapoptotic drugs 14, 16, the latter is thought to be accomplished by stimulating NPC proliferation or by transplantation of exogenous NPCs. Although transplantation of stem cells improves post‐stroke symptoms, grafted stem cells do not replace cells lost in injured tissue. Rather, grafted stem cells act in an indirect manner, very likely by releasing trophic and anti‐inflammatory factors that promote the survival, remodeling, and plasticity of the ischemic brain tissue 17-19.

Considering the paracrine nature of stem cell‐mediated beneficial effects, the choice of stem cell source might not be essential for achieving recovery‐promoting effects of cellular therapeutics. As a matter of fact, in addition to NPCs stem cells derived from various adult tissues have been found to promote restorative effects in the ischemic brain 18, 20-22. Especially due to their broad availability, their simple handling and their low side effects, bone marrow‐derived mesenchymal stem cells (MSCs) became an attractive cell source to treat ischemic stroke in a number of different preclinical models.

Transplantation of MSCs and NPCs After Stroke

Preclinical transplantation studies in a plethora of stroke models using MSCs or NPCs have shown beneficial effects (Table 1) in a large number of different readouts 23, 26-36, 43-45. NPCs, either administered intracerebrally or systemically, mediate neuroprotection and enhance neurological recovery via stimulation of endogenous angiogenesis and neurogenesis. The mechanisms involved in the process of NPC‐induced brain protection and brain regeneration greatly depend on both cell delivery routes and cell delivery timing 34, 35. For example, acute NPC transplantation reduced neuronal injury and infarct volume, while transplantation at later stages rather modifies post‐stroke brain regeneration and neuronal plasticity.

Likewise, the transplantation of MSCs, which have been administered systemically in the majority of studies, revealed promising effects in experimental stroke models. MSC transplantation was found to reduce neuronal injury and infarct volume, increase angiogenesis and neurogenesis, and improve neurological recovery. Although a majority of studies has been performed on BM‐derived MSCs, some studies imply the application of adipose‐derived MSCs which might appear to be an attractive cell type as well 24, 25, since the latter is easy to obtain. Due to their beneficial effects in the preclinical models, controlled randomized clinical trials (Table 1) using MSCs (and to a lesser extent NPCs as well) for stroke treatment have been started 38-40, 46. Although patient recruitment is so far low, which precluded more final conclusions from these studies, some studies reported beneficial outcomes after MSC transplantation. Of note, no clinically relevant side effects within the observation periods of maximal 5 years have been observed.

At the mechanistic level, it was initially proposed that applied MSCs enter the damaged tissue and replace lost cell types. However, in preclinical stroke as well as in other disease models, MSCs are hardly detected in affected tissues 47-52. Most of systemically applied MSCs get trapped within the lungs 53, 54. Due to these observations, the initial idea that MSCs can replace cells in affected tissues or directly interact with target cells became challenged, and the hypothesis emerged that MSCs effectively act in a paracrine rather than a cellular manner 55.

Structure of Extracellular Vesicles and Biological Properties

Whereas early studies proposing a paracrine mode of action of administered MSCs claimed that soluble factors, such as growth factors or cytokines, mediate the stem cells' beneficial therapeutic effects 47; more recent data qualified extracellular vesicles (EVs) as the critical agents 56. Indeed, MSC‐derived EVs (MSC‐EVs) mediating therapeutic activities have been documented in a variety of different preclinical models and in a GvHD patient as well 49, 56-59.

EVs are released by almost all cell types and are detected as membrane‐surrounded vesicles in all body fluids 60. According to their origin, different EV types can be discriminated 61. Exosomes are derivatives of the late endosomal compartment and have diameters of 70–150 nm. They correspond to intraluminal vesicles (ILVs) that are formed by the inward budding of the limiting membrane of sorting and late endosomes. The ILV containing endosomes are called multivesicular bodies (MVBs) or multivesicular endosomes. At the example of maturing reticulocytes, it has been shown that MVBs can fuse with the plasma membrane and release their ILVs as exosomes into the extracellular compartment 62-64. In contrast, microvesicles (MVs), which have diameters of 100–1,000 nm, are formed as bud offs of the plasma membrane; together with apoptotic bodies which have said sizes of 500 nm to several micrometers, exosomes and MVs form the most prominent EV subtypes 65.

EVs contain specific molecular signatures reflecting their cell of origin 60, 66, 67. Apart from lipids and proteins, metabolites and nucleic acids are recovered in prepared EV fractions 68-70. A proportion of EVs might contain molecules that cells cannot metabolize, which are released into the extracellular environment for further processing. Other EVs seem to be assembled in a tailored manner to act as intercellular communication vehicles mediating complex signal exchanges between cells within and between different organs 60, 61, 71.

Preclinical Studies Using EVs in Animal Models Unrelated to Ischemia

In recent years, EVs have made a tremendous progress in biomedical research. At first, EVs were considered as debris. In 1996, however, Raposo and colleagues showed that B cells release MHC‐II containing EVs which can activate T cells 72. Yet, until the finding that EVs contain RNAs, in 2006 and 2007 68, 70, EV research was sparse. Thereafter, the EV field started to grow exponentially. Positive therapeutic effects of MSC‐EVs were reported for the first time in 2009; the group of Giovanni Cammussi described EV‐mediated therapeutic activities in a kidney failure model 59. In 2010, the group of Sai Kiang Lim and Dominque de Kleijn discovered cardioprotective activities in their MSC‐EV fractions 49. We were the first group who applied MSC‐EVs to a human patient in an individual treatment attempt. We applied an allogeneic MSC‐EV fraction to a steroid refractory graft‐versus‐host disease patient, who failed to react on several second side strategies. Remarkably, the clinical symptoms declined significantly during and after the 2‐week MSC‐EV therapy, without revealing any side effects 57. Meanwhile, EVs have been applied to several preclinical diseases models unrelated to ischemia, with some of them mentioned in Table 2.

The therapeutic benefit of EVs has been analyzed in various disease conditions, including inflammatory processes and cancer models. Similar to stem cells derived from different tissues, stem cell derived EVs exert multiple effects on different target cells. Similar to stem cells derived from different tissue sources inducing a variety of actions in biological tissues, EVs depending on their stem cell source have multiple effects on target cells, which may show overlaps, but also differences between cell sources. The latter is vital in understanding the different beneficial effects that EVs can yield. As such, EVs from a certain cell might show beneficial effects in a variety of malignant diseases, such as hepatocellular carcinoma, gastric cancer, or brain tumor, but not be equivalent in their cellular actions. Although a direct comparison between these studies is not eligible due to different study designs, EVs might either have a direct impact on tumor formation or enhance sensitivity to chemotherapy 75, 84, 86, 96. Similar evidence for overlapping effects of EVs came from studies in inflammatory/infectious conditions, such as arthritis, hepatitis C, HIV, and sepsis 74, 87, 93, 98, 99. One has to stress that several observations are still limited to in vitro research only. Particularly important from the authors’ point of view, EVs have successfully been used in preclinical neurodegenerative disease models, such as amyotrophic lateral sclerosis and Parkinson's disease, as well as in myasthenia gravis where EVs were found to modulate inflammatory responses and cell survival 73, 82, 88. Further evidence for a role of EVs in modulating inflammatory responses and tissue regeneration was found in animal models of traumatic brain injury and skin wounds 91, 92.

Preclinical Studies Using EVs in Animal Models Associated with Ischemia

More recent studies identified the therapeutic efficacy of EVs in experimental conditions mimicking peripheral limb, heart or brain ischemia, that is, in models of peripheral occlusive artery disease, myocardial infarction and stroke (Table 3). For myocardial ischemia, the therapeutic efficacy of EVs has been shown in a large number of in vitro and in vivo studies 49, 101-111. Thus, EVs from various cell sources including MSCs and embryonic stem cells, promoted cellular survival, reduction of infarct size, and stimulated myocardial remodeling and angiogenesis. Of note, these EV actions were associated with functional recovery evaluated by ejection fraction.

To the best of the authors' knowledge, six different studies have examined effects of EVs in ischemic stroke models, most in rats and one in mice 112-114, 116-118. In the first rat study, Chopp and colleagues 113 intravenously applied MSC‐EVs in a model of transient intraluminal middle cerebral artery occlusion. EVs were administered via tail vein injection at 24 hours post‐stroke. The authors observed a significant reduction of brain injury and neurological impairment that was associated with enhanced post‐ischemic neurogenesis. In the hitherto only mouse study, we studied effects of MSC‐derived EVs in transient intraluminal middle cerebral artery occlusion. Using the polyethylene glycol (PEG) method EVs were enriched from MSC conditioned media. MSCs were raised from BM samples of two healthy bone marrow donors; as serum supplement 10% human platelet lysate was used 119, 120. MSC‐EVs were administered at days 1, 3, and 5 post‐stroke. The treatment enhanced neurological recovery and increased endogenous neurogenesis and angiogenesis, at the same time reversing stroke‐induced peripheral immunosuppression. In a head‐to‐head comparison, the therapeutic potential of MSC‐EVs was comparable to that of the transplanted MSCs from which the MSC‐EVs were derived 112.

A more recent rat study examined the effects of MSCs combined with MSC‐EVs 114, demonstrating that combined MSC and MSC‐EV delivery was superior in terms of brain protection and neurological recovery when compared with MSC transplantation or EV injection only. These studies raised the question of how therapeutic effects of EVs may be boosted by loading naïve EVs with biologically active molecules such as noncoding RNAs, which by means of EVs may safely be transported to target tissues 121. In rats exposed to transient middle cerebral artery occlusion, increased neural plasticity and neurological recovery were noted after delivery of EVs obtained from miR‐133b overexpressing MSCs when compared with EVs obtained from naïve MSCs 117. In vitro experiments using oxygen‐glucose‐deprivation suggested that the enhanced action of miR‐133b containing EVs may be due to stimulation of secondary EV release from astrocytes 117. In another study, EVs harvested from MSCs transfected with a miR‐17‐92 cluster plasmid induced better neurological recovery when compared with EVs derived from naïve MSCs 116. These observations stress the heterogeneity of EV actions depending on the loading of EVs with survival and plasticity promoting molecules.

Clinical Studies Using EVs in Humans

Despite an increasing body of evidence demonstrating that EVs might serve as biomarkers for stroke outcome 122, there is currently no study in which EVs (and especially MSC‐EVs) have therapeutically been administered to human stroke patients. According to the promising data obtained in a variety of different animal models and the very promising result of the individual treatment attempt of a GvHD patient with MSC‐EVs, a number of groups now try to translate EVs into the clinics. As EVs are novel biological agents and MSC‐EVs are not considered as Advanced Therapy Medicinal Products (ATMP), they provide a new class of biologicals, for whose production no concrete rules have been defined by the FDA or any other national regulatory agency, yet. To this end, experts in the field have summarized in an International Society of Extracellular Vesicles (ISEV) position paper the different therapeutic EV‐application fields, discussed their regulatory status and recommended requirements to be fulfilled to translate EVs as therapeutic agents into the clinics 56.

Current Limitations and Benefits of EV‐Based Treatment Paradigms

Despite their different origin and their different proposed sizes, EV subtypes could not discriminated during isolation until now. Thus, the ISEV agreed in 2014 to name fractions proposed to contain exosomes, MVs, apoptotic bodies and/or other EV types appropriately as EV fractions 123. Since EV fractions contain a heterogeneous mix of different EV types, care has to be taken, of how EVs are purified and characterized. As such, the application of differential centrifugation (i.e., ultracentrifugation) is hampered by a low EV output due to restricted sample volumes in comparison to other techniques like size exclusion chromatography 124. In this sense, the recently identified observation of low density lipoprotein contamination after EV enrichment might pose a problem for the evaluation of past and future work when dealing with mechanistic approaches 125. On the contrary, for pure therapeutic applications, contaminations might be tolerated. Despite a plethora of different enrichment techniques available, ultracentrifugation, however, remains to be the gold standard for EV enrichment, albeit other techniques such as PEG isolation provide some advantages (own unpublished observation). Consequently, the ISEV has released consensus recommendations on EV purification and characterization 123. Still, several studies do not follow these recommendations, making it difficult to compare research outcomes. To increase the reliability of the data and to promote standardization in the field the EV‐TRACK consortium was formed which defined several criteria to score EV‐based studies that will hopefully be followed in the future 126.

Furthermore, caution has to be taken when interpreting studies from both the stem cell and the EV field. Comorbidities and comedications, for instance, might modulate experimental outcomes. As such recommendations—especially from the cardiologic field—have been made in order to overcome typical pitfalls of cell‐based therapies 127-129. The latter emphasize the necessity of selecting the appropriate cell type or components of the secretome depending on the endpoint chosen and the definition of the application mode, including the amount of applications, the application timing and the delivery routes, to name but a few.

As EVs lack nuclei they cannot self‐replicate and thus in contrast to cells do not contain any endogenous tumorigenic potential. In addition, EVs are easier to handle and, due to their small size, they can be sterilized by filtration 56. Thus, EV‐based therapeutics provide several advantages over cellular therapeutics, resulting in a competition between several research groups to produce MSC‐EVs for the clinical setting. There are several challenges connected to this issue. On the one hand, large volumes have to be processed under good medical practice compliant conditions to obtain sufficient material to treat a patient. Then, as MSCs provide a heterogeneous cell entity, MSC‐EV fractions may show varying therapeutic activities as well. Indeed, the authors detected significant differences in the cytokine profile of independent MSC‐EV preparations during their own research activities 57.

Conclusion

The application of stem cell derived EVs, especially that of MSC‐EVs, offers a great opportunity for adjuvant stroke treatment. For now, EVs appear to be safe in mammals and potentially also in man, thus avoiding putative side effects that are inherent to stem cell transplantation such as malignant stem cell transformation. Besides, tissue engineering techniques allow the usage of EVs as potent carriers for bioactive molecules, which may be used for overcoming tissue barriers such as the blood‐brain barrier for targeting distinct cell populations 56. Yet, fundamental questions as to their exact mode of action and their optimal enrichment, characterization, and storage have to be answered to optimize them for the clinical setting 56.

Acknowledgments

For collaboration and general support, we thank our colleagues from the institute, as well as all collaboration partners. For the financial support of our EV research, we are grateful to the Medical Faculties of the University of Duisburg‐Essen (IFORES) and the University of Goettingen, the Volkswagen Foundation, the Stem Cell Network North Rhine Westphalia, and the LeitmarktAgentur.NRW and the European Union (European Regional Development Fund 2014‐2020), as well as to the European Union for the EU COST program ME‐HaD (BM1202).

Author Contributions

T.R.D., M.B., D.M.H., and B.G.: manuscript writing, final approval of the manuscript.

Disclosure of Potential Conflicts of Interest

The authors indicated no potential conflicts of interest.