Introduction

Stroke is the most common cerebrovascular disease and the second leading cause of death behind heart disease and is a major cause of long-term disability worldwide [1]. Our understanding of the pathophysiological cascade following ischaemic injury to the brain has greatly improved over the past few decades. Cell therapy, as a new strategy addition to traditional surgery and thrombolytic therapy, has attracted increasing attention [2]. The therapeutic options for stroke are limited, especially after the acute phase. Cell therapies offer a wider therapeutic time window, may be available for a larger number of patients and allow combinations with other rehabilitative strategies.

The immune response to acute stroke is a major factor in cerebral ischaemia (CI) pathobiology and outcomes [3]. In addition to the significant increase in inflammatory levels in the brain lesion area, the immune status of other peripheral immune organs (PIOs, such as the bone marrow, thymus, cervical lymph nodes, intestine and spleen) also change to varying degrees following CI, especially in the spleen [4]. Over the past decade, the significant contribution of the spleen to ischaemic stroke has gained considerable attention in stroke research. At present, the spleen is becoming a potential target in the field of stroke therapy for various stem cell treatments represented by multipotent adult progenitor cells (MAPCs).

Two cell therapy strategies

Two distinct cell therapy strategies have emerged from clinical data and animal experiments (Fig. 1). The first is the nerve repair strategy, which uses different types of stem cells with the ability to differentiate into cells that make up nerve tissue and thus can replace damaged nerves to promote recovery during the later stages after stroke [5,6,7,8,9,10,11]. This strategy usually involves cell delivery to the injury site by intraparenchymal brain implantation and stereotaxic injection into unaffected deep brain structures adjacent to the injury site. The main problem with this strategy is that we should not only ensure the efficient delivery of cells to the injury site but also try to reduce the invasive damage caused by the mode of delivery. Moreover, evaluation of the extent to which cells survive over the long term, the differentiation fates of the surviving cells and whether survival results in functional engraftment is difficult. This strategy mainly includes intracerebral [12,13,14,15], intrathecal [16] and intranasal administration [17] (Fig. 2).

Fig. 1
figure 1

Two cell therapeutic strategies for stroke. Replacement of necrotic cells and immunomodulation. Therapeutic stem cells have traditionally been known to differentiate into cells that make up nerve tissue to replace necrotic cells, thereby promoting nerve regeneration and angiogenesis. Recent studies have shown that the immune regulatory capacity of stem cells provides a favourable environment for nerve and vascular regeneration

Full size image
Fig. 2
figure 2

The main routes of administration of stem cell therapy for stroke. Although many preclinical studies and clinical applications have been carried out, the most adequate administration route for stroke is unclear. Each administration route has advantages and disadvantages for clinical translation to stroke patients. a Intranasal, b intracerebral, c intrathecal, d intra-arterial, e intraperitoneal and f intravenous

Full size image

The second strategy is an immunoregulatory strategy (typically therapeutic cells are injected intravenously), which takes advantage of the release of trophic factors to promote endogenous stem cell (NSC/neural progenitor cell (NPC)) mobilisation and anti-apoptotic effects in addition to the anti-inflammatory and immunomodulatory effects encountered after systemic cell delivery. The mechanism of action appears to be reliant on “bystander” effects; these effects are likely to include immunomodulatory and anti-inflammatory effects mediated by the systemic release of trophic factors [18, 19], since neither animal nor human data have found any signs of actual engraftment of intravenously delivered cells in the brain [20,21,22]. In addition, many therapeutic stem cells have been found to migrate to PIOs, such as the spleen, following brain injury to play an immunoregulatory role, thus providing a good environment for nerve and vascular regeneration in vivo. This strategy mainly includes intra-arterial [23,24,25,26], intraperitoneal [27] and intravenous administration [28,29,30,31] (Fig. 2). Currently, systemic immunomodulatory therapy based on intravenous cells is attracting increasing attention [29].

Immunoregulation may be a better strategy

Further insight into the role of the two strategies has been provided by studies using cellular therapies in experimental models of brain ischaemia. All cells are more efficacious when administered systemically than when delivered via intracerebral administration [32,33,34,35,36,37], probably because intracerebral administration does not guarantee the extent to which cells can migrate from their implantation site in human subjects. Placing cells within the cystic space left as a long-term consequence of ischaemic damage in the absence of some type of bio-scaffold will be unlikely to promote cell adherence or persistence. Moreover, gliosis on the margins of the damaged region may impede cell migration or axonal outgrowth in the same manner as encountered after spinal cord injury.

The pathological progression of stroke is a complex systemic process, and changes in state occur in tissues besides intracranial tissues. Studies have shown that the immune response/regulation after stroke plays an important role in the pathological progression of stroke. The immune response is an important endogenous mechanism of post-stroke activation. Although the immune response following stroke, including cytokine production and inflammatory cell infiltration into damaged brain tissues, has been known for many years [38,39,40], the complexity of the mechanisms involved in post-stroke immune activation, inflammatory damage and tissue repair are unknown. In the future, immunomodulation will be an important potential therapeutic strategy for stroke. Moreover, finding the most appropriate therapeutic target for therapeutic cells may further improve the effectiveness of immunomodulatory treatment.

Stem cells and immunoregulation after stroke

At present, most cells used for immunoregulation therapy after stroke are various types of stem cells. However, animal experiments have shown that anti-inflammatory immune cells (such as regulatory T cells (Tregs), helper T (Th)-2 cells and regulatory B cells (Bregs)) can also alleviate brain damage [41,42,43,205]. Based on their experimental results, they put forward the hypothesis that administration during the acute phase (within 1.5 h) after ischaemic attack allowed the hAECs to migrate preferentially to the spleen and damaged brain; subsequently, cell apoptosis and inflammation were inhibited. Early brain infiltration of immune cells, aggravation of infarction and systemic immunosuppression were also alleviated [48].

HSCs injected intravenously 24 h after reperfusion were first detected at 24 h after injection in the spleen and later in the ischaemic brain parenchyma [35]. In addition, compared with that of the sham-operated control group, the immune environment after CI increased HSC migration to the spleen 72 h after reperfusion. In the absence of induction of an injury, the cells did not preferentially accumulate in the spleen [35]. HSC treatment reduced the infarct volumes, apoptotic neuronal cell death in the peri-infarct areas and immune cell (T cells and macrophages) infiltration into the ischaemic hemispheres. Moreover, HSC therapy decreased the TNF-α, IL-1β, CCR2 and CX3CR1 levels in the spleen after CI [35].

These experiments suggest that stem cell therapy works to some extent by regulating the immune response after stroke, especially at the spleen level, which may be crucial and is an important potential therapeutic target.

Important therapeutic targets

Many factors and regulatory pathways change significantly during the entire pathophysiological process of stroke. They may play different roles in different pathological stages after CI. Fully understanding their effects on stroke may identify new targets for development of novel therapeutic strategies.

IL-10

IL-10 is a multicellular, multifunctional cytokine that regulates cell growth and differentiation and participates in inflammatory and immune responses. It is mainly produced by cells such as Tregs, Th2 cells and Bregs and currently is recognised as an anti-inflammatory and immunosuppressive factor. IL-10 is an important component of the endogenous repair mechanism after stroke. The IL-10 levels in the brain and spleen increase after stroke [90, 155, 156]. The use of exogenous IL-10 for anti-inflammatory therapeutic approaches has been shown to provide neuroprotection during ischaemic stroke [206]. However, an excessive IL-10 response can contribute to immunosuppression after CI, which worsens outcomes. Additionally, sex differences may exist in the role of IL-10 in stroke recovery [207].

Adoptive transfer of IL-10-secreting cells is widely used for the treatment of experimental stroke. However, after the earlier use of Tregs [41], Bregs are attracting increasing attention. Adoptive reinfusion of spleen-derived Bregs can improve neurological injury and motor dysfunction after CI in experimental rats [34, 140]. Elevated Breg and Treg levels in the spleen at different time points after CI have been found in some studies investigating the preconditioning protection of I/R [208, 209]. In addition to their anti-inflammatory effects, Bregs can also promote the activation and proliferation of other anti-inflammatory immune cells and have a cascade amplification effects on the repair mechanism after CI [147, 210] (Fig. 8). Therefore, we have proposed that Bregs may represent an important potential strategy to rapidly start the endogenous protective mechanism during the early stage of stroke and increase the Breg levels in the body, especially in the spleen.

Fig. 8
figure 8

Effects of Breg cells on other immune cells after stroke. Through IL-10, IL-35 and TGF-β production, Bregs can inhibit the differentiation of pro-inflammatory lymphocytes, such as TNF-α-secreting monocytes, IL-12-secreting dendritic cells, Th17 cells, Th1 cells, IL-17+ γδT cells and cytotoxic CD8+ T cells. Bregs can also induce the differentiation of immunosuppressive T cells, such as Foxp3+ Tregs and T regulatory 1 (Tr1) cells and contribute to the maintenance of iNKTs. Therefore, Bregs result in immune regulation at sites of inflammation, such as the CNS

Full size image

Interferon

IFN-γ is the only member of the type II IFNs and is produced only by activated T cells, NKs and natural killer T cells (NKTs). IFN-γ is a marker cytokine of Th1 cells that can activate APCs and promote the differentiation of Th1 cells by upregulating related transcription factors. Deletion of the IFN-γ gene has been shown to reduce brain damage after CI [137]. In the first 3 days after I/R, intraventricular administration of an IFN-γ neutralising antibody can protect the brain from CI injury [140]. IFN-γ participates in the Th1 inflammatory response by activating monocytes, microglia and macrophages. Since activation of microglia/macrophages is part of the cause of delayed cell injury following ischaemic injury, IFN-γ may play a role in the splenic response by aggravating the inflammation associated with ischaemic injury. Seifert et al. previously proposed that the spleen contributed to stroke-induced neurodegeneration through IFN-γ signalling [119]. IFN-γ is increased early in the spleen but later in the brain following I/R. Splenectomy reduces the IFN-γ level in the infarct after MCAO. The protective effect of splenectomy was eliminated by systemic recombinant IFN-γ accompanied by an increase in IFN-γ expression in the brain post-pMCAO [119]. Moreover, IP-10 has been shown to be a key factor in stroke-induced neurodegenerative diseases [154]. NKs participate in CI and promote neuronal necrosis through IFN-γ. IP-10 expands the damage of the BBB induced by NKs through CXCR3 [149]. Furthermore, hUCB therapy inhibits the proliferation of splenic T cells after MCAO by increasing IL-10 and IFN-γ production [32].

IFN-β is a type I IFN that binds to the IFN-α/β receptor. IFN-β exerts anti-inflammatory effects, and systemic administration of recombinant IFN-β has been used for the treatment of multiple sclerosis, which is a neuroinflammatory condition of the CNS [211]. Therefore, IFN-β has been considered to be able to decrease neuronal death and promote functional recovery after CI by limiting inflammation. In agreement with this view, IFN-β has been put forward as a candidate drug for the treatment of stroke. Several animal experiments have shown that systemic administration of recombinant IFN-β at different time points before and after MCAO results in neuroprotection [212,213,214]. This outcome may be related to endogenous IFN-β signalling not only to reduce CNS inflammation but also to reduce autoreactive T cell proliferation via inhibiting the antigen presenting capacity of astrocytes and microglia. Inácio’s research also showed that endogenous IFN-β signalling could alleviate local inflammation and regulate peripheral immune cells, thereby contributing positively to stroke outcomes [215].

Cholinergic anti-inflammatory pathway

As mentioned above, VNS causes prominent attenuation of the systemic inflammatory response evoked by CI in experimental animals. This effect is mediated by ACh stimulation of acetylcholine receptors on splenic macrophages. Therefore, the circuit is known as the “cholinergic anti-inflammatory pathway”, which casts the spleen as the major effector [216]. α7nAChR is considered to be an important target for alleviating the release of pro-inflammatory cytokines from macrophages and DCs.

Noradrenergic neurons provide innervation to all primary and secondary lymphoid tissues, including the bone marrow, spleen and lymph nodes [217, 218]. Most immune cells express one or more ARs, but β2-AR is the most widely distributed and mediates most of the effects of sympathetic nerves on immune function [219,220,221,222]. Sympathetic nerves affect the innate and adaptive immune responses through stimulating β2-AR. Most evidence shows that β2-AR activation has immunosuppressive effects on monocytes and macrophages [221, 222]. Stimulation of β2-AR-naïve CD4+ T cells (Th0) results in their differentiation into Th1 cells, which enhance cellular immunity, or Th2 cells, which decrease cellular immunity [221, 222]. Norepinephrine (NE) can also produce β2-mediated anti-inflammatory effects by releasing ACh from cholinergic spleen cells [221, 222]. Blocking adrenocortical receptors has been shown to inhibit the splenic response after pMCAO and reduce injury [123]. The spleen receives noradrenergic innervation from the postganglionic sympathetic neurons [223]. Splenic T cells are the source of ACh and may be a link between NE and splenic macrophage suppression [224]. β2-AR on T cells is essential for the anti-inflammatory effect of VNS [225, 226]. Splenectomy eliminates the role of VNS in increasing plasma NE, which supports the conclusion that plasma NE is released from the spleen into circulation during VNS [227]. The spleen plays a central role in the pathophysiology of the hyperinflammatory state triggered by threatening conditions, such as stroke, and splenic macrophages are the dominant source of pro-inflammatory cytokines [226].

Targeted therapy for α7nAChR on microglia and macrophages after stroke has long been considered an important potential strategy. α7nAChR expression on activated microglia and infiltrated macrophages after CI plays an important role in the pathological process of stroke. Nicotine therapy has been shown to significantly attenuate the increase in microglia and inflammatory cytokines (i.e., TNF-α and IL-1β) induced by CI through mediation of α7nAChR [228]. In addition, α7 agonists can reduce the infarct volume and functional deficits in different animal models of stroke [229, 230]. In contrast, blockade of α7nAChRs with a selective antagonist increases the infarct volume, which suggests some degree of α7nAChR stimulation by the endogenous agonist [230]. The endogenous choline released from damaged brain tissue may fulfil this role. The use of an allosteric modulator of α7nAChRs 6 h after tMCAO can reduce the infarct volume and improve neurological performance [231, 232]. In addition, regulation of α7nAChRs has also been shown to be associated with changes in M1 and M2 microglia/macrophages after CI [229].

Multipotent adult progenitor cells

MAPCs are a unique type of adult adhesion cells (MSC-like) that extend understanding of how intravenous cell therapy participates in and regulates the peripheral immune system after stroke. MAPCs can be isolated from bone marrow and other tissues [233], are well characterised and can be easily distinguished from bone marrow mononuclear cells (BMMNCs) and MSCs based on their size [234], transcriptome [235], microRNA profile [236], differentiation capability [237] and secretome [238].

Intravenous injection of MAPCs has been shown to have beneficial effects on other CNS injuries, such as TBI [239]. MAPC therapy attenuates activated microglial/macrophage responses, preserves the BBB, reduces cerebral oedema and improves spatial learning after TBI [118, 239, 240], and MSPC treatment within 24 h after injury can achieve better outcomes [241]. After TBI, intravenous MAPCs also tend to migrate to the spleen [234].

MAPCs have become a new hotspot of cell therapy for stroke after BMSCs, hUCBs and HSCs. MAPCs can significantly inhibit the inflammatory reaction of injured brain tissue [242, 243] and improve the motor function and neurological outcomes of experimental stroke animals [242, 243]. Moreover, MAPCs exhibit more robust tissue sparing and mitigation of glial activation than MSCs regardless of whether intravenous or intraparenchymal administration is used [244, 245]. Similarly, recent animal data support a role for intravenous MAPCs in regulation of the peripheral immune system through specific interactions between MAPCs and splenocytes, thereby promoting stroke recovery and inhibiting splenic atrophy in the 24 h after CI. Similar to results obtained from testing hUCBs, stroke can also accelerate the migration of splenocyte MAPCs to the spleen and inhibit splenocyte apoptosis. In addition, MAPCs decreased the levels of CD3+, CD4+ and CD8+ cells in the spleen and increase Tregs after tMCAO compared with the levels in the vehicle-treated group [246]. In terms of inflammatory factors, IL-1β and TNF-α were significantly lower in the splenic cells of rats with I/R after MAPC treatment than those of the saline-treated animals, whereas the IL-10 level was higher [246].

MAPCs not only have a good protective effect on acute CI but also maintain a stable and sustained therapeutic effect on chronic stroke. Compared with those of the saline-treated group, the MAPC treatment group exhibited advantages in locomotor and neurological outcomes that persisted for more than 28 days. In addition, the results of a comparative trial of MAPCs for stroke in splenectomized and sham splenectomized mice also demonstrated that MAPCs could inhibit expansion of the infarct volume through a non-spleen-mediated mechanism, but the MAPC-induced IL-10 production after tMCAO was still dependent on an intact spleen. Moreover, no significant difference was found in the improvement of neurological outcomes between the splenectomy group treated with MAPCs 24 h after tMCAO and the splenectomy group treated with saline [246].

Preclinical animal data support the benefits of intravenous MAPC therapy for stroke. A phase I/II clinical trial is under way to test the safety and efficacy of MultiStem (the MAPC clinical-grade product) for treatment of patients with acute ischaemic stroke. Previous in vitro studies have shown that MACPs inhibit CD8+ T cells, which are harmful to stroke [247]. The MASTERS trial (MultiStem in Acute Stroke Treatment to Enhance Recovery Study) was conducted in 33 clinical centres in the USA and UK from October 2011 through December 2015. First, MultiStem treatment was proven to be safe. No infusion-related allergic reaction was observed in the MultiStem and placebo-treated groups, and no cases showed neurological worsening. MultiStem treatment can reduce the risk of life-threatening adverse events or death and secondary infections in stroke patients. Furthermore, MultiStem treatment also greatly shortened the time in the intensive care unit and the overall hospitalisation time compared with those of the placebo-treated patients. Compared with those of the placebo, MultiStem treatment significantly reduced the biomarkers of post-stroke inflammation (circulating CD3+ T cells and inflammatory factors). Importantly, MultiStem treatment significantly improved the chance for an excellent outcome at 1 year of onset [248, 250].

Currently, the phase III MASTERS-2 trial aims to expand our knowledge and understanding of treatment of ischaemic stroke patients with MultiStem and plans to start recruiting patients in 2018. Another MultiStem clinical study named TREASURE (Treatment evaluation of acute stroke for using in regenerative cell elements) was officially launched in 31 medical centres in Japan in 2017. The research project recruited 220 patients with acute ischaemic stroke, including speech or motor deficits, as defined by a National Institution of Health Stroke Scale (NIHSS) score of 8–20 at baseline. TREASURE is a randomised, double-blind, placebo-controlled, multicentre phase 2/3 trial to evaluate the efficacy and safety of intravenous administration of MultiStem® compared with those of a placebo in patients with ischaemic stroke [249].

Discussion

Intravenous cell therapy can modulate the acute and adverse contributions of the peripheral immune system after ischaemic stroke [250]. A single intravenous administration of cells 24 to 36 h after stroke onset that can mitigate and rebalance the immune response to the initial focal ischaemic injury is sufficient for nerve repair and improvement of long-term outcomes [29].

As mentioned earlier, the activation, migration and participation of peripheral immune cells, and perhaps most importantly immune cells from the spleen, are critical steps in the pathophysiological progression after stroke. Therefore, we believe that targeted inhibition of peripheral blood immune cell (especially splenic cell) activation, reduction of inflammatory cytokine production and inhibition of their entry into the brain parenchyma through the BBB are key steps in attenuating the expansion of pro-inflammatory microglial activation, neuronal die-back and tissue loss.

Simply inhibiting the participation of splenic components of the peripheral immune system in post-stroke pathophysiological processes allows tissue sparing but is not sufficient to enable neurological and locomotor benefits [246]. As mentioned above, splenectomy fails to provide long-term protection against ischaemic stroke, and delayed splenectomy neither reduces brain tissue loss nor alleviates sensorimotor and cognitive impairment [121]. These results suggest that the spleen may be a double-edged sword that plays completely opposite roles during different pathological stages of stroke. In the early stage of stroke, the spleen is mainly characterised by inflammation and harmfulness but then gradually transforms into an anti-inflammatory and protective phenotype. Therefore, inhibiting the inflammatory immune response of the spleen in the early stage of stroke and accelerating the initiation of its anti-inflammatory mechanism have become the keys for the use of stem cells in the treatment of acute stroke. We also summarise that many stem cells, including MAPCs, can simultaneously inhibit potentially harmful aspects of the innate immune system response to stroke while speeding up beneficial aspects or reparative responses, which confirms our viewpoint (Fig. 9).

Fig. 9
figure 9

Stem cell therapy enhances recovery after stroke. In the untreated scenario, ischaemic stroke leads to activation of the peripheral immune system. During this process, the spleen atrophies, lymphocytes undergo apoptosis in the spleen, the inflammatory factor levels in the spleen increase, and inflammatory cells are released from the spleen into circulation. The antigen presentation of DCs is enhanced, and the levels of various chemokines are elevated. These pro-inflammatory mediators contribute to M1 microglia-mediated destruction of the BBB and CNS inflammation. Infiltration of leukocytes further aggravates inflammatory necrosis of neurons. Intravenous administration of stem cells reverses splenic atrophy and converts macrophages from the pro-inflammatory M1 phenotype to the anti-inflammatory M2 phenotype and Th cells from the pro-inflammatory Th1 phenotype to the anti-inflammatory Th2 phenotype. The inflammatory cytokine and cell levels in the spleen decrease, and anti-inflammatory cytokines and cells begin to be produced and released into circulation, which ultimately lead to less BBB restructuring and CNS inflammation and provide a favourable environment for nerve regeneration and angiogenesis

Full size image

Through a number of preclinical and clinical studies, we have obtained a certain understanding of the biological distribution of stem cells after injection for the treatment of stroke and their effects on many immune cell subsets and cytokines in the CNS and PIOs. However, little is known about the effects of stem cell therapy on the immune regulation mediated by the ANS and SNS. Many trials will provide an opportunity to better evaluate and examine hypothetical mechanisms, and we believe that stem cell therapy provides benefits for stroke; however, further preclinical and clinical studies are needed to advance a more comprehensive understanding. Evaluating the safety and efficacy of various stem cell therapies at multiple doses and at different time points may also yield new information.

Conclusion

In summary, we can draw the following conclusions. First, the splenic response after stroke is critical for pathological damage and tissue repair. The key of immunomodulatory therapy for stroke may be to inhibit splenic inflammation at the early stage of pathology and accelerate the initiation of its anti-inflammatory/repair mechanism. Second, during the course of stem cell therapy for stroke, allowing more stem cells to migrate to the spleen may inhibit the splenic inflammatory response and accelerate the initiation of its anti-inflammatory/repair mechanism, which will have a better therapeutic effect than increasing the number of stem cells delivered to the injured brain tissue. Finally, inflammatory factors, such as TNF-α and IL-1β, produced after stroke can activate inflammatory pathways, such as NF-κB, in stem cells, thereby enabling stem cells to acquire stronger immune regulation potential. Therefore, stroke-like stimuli can be applied to stem cells for treatment before injection, which may lead to a better therapeutic effect. In the future, we believe that the spleen will become a potential target of various stem cell therapies for stroke represented by MAPC treatment. The research results will move closer to clinical application and ultimately benefit the majority of stroke patients.