Background
Stroke is the fourth most common cause of death and a leading cause of disability in the United States alone. It is increasingly accepted that human stroke does not just create a single organ injury but induces a complex interaction between the central nervous system (CNS) and peripheral immune system. Experimental stroke in mice induces a biphasic response in the peripheral immune system characterized by an initial activation phase (6 to 22 hours) [
1] followed by an immunosuppressive phase (96 hours) which is accompanied by a pronounced atrophy of the spleen and thymus [
2]. Peripheral immune cells home to the brain, transmigrate over the activated endothelium and invade the damaged brain in a timed fashion [
3]. The developing infarct is exacerbated by the influx of inflammatory cells and the time course and degree of accumulation of multiple inflammatory cell types in the brain has been extensively studied [
4‐
6]. T cells may be major offenders in mediating the post-stroke inflammatory response, contributing to increased brain damage. When activated, T cells produce cytokines that initiate an inflammatory cascade involving recruitment of other inflammatory cells to sites of injury [
7]. T cells are observed in the brain within hours of the ischemic insult [
4,
8] and T cell-deficient animals have reduced infarct size after stroke [
9].
The activation of T cells via the various antigen-presenting cells (APCs) forms an integral part of the inflammatory process. This process gains importance especially since the inflammatory cells in ischemic stroke include CD4
+ and CD8
+ T lymphocytes, microglia, and macrophages [
4]. T cell activation is a complex and multistep phenomenon and the B7 family of co-stimulatory molecules play a pivotal role in optimal activation of T cells [
10]. Programmed death-1 (PD-1), a member of the B7-CD28 family, is a co-inhibitory receptor expressed by a variety of activated immune cells, including CD4
+ and CD8
+ T cells, natural killer T (NKT) cells, B cells, monocytes, and some dendritic cell (DC) subsets [
11]. PD-L1 and PD-L2 are the two known ligands for PD-1 [
12] and they have different expression patterns. While PD-L1 is found constitutively expressed on murine lymphoid cells, such as T cells, B cells, macrophages, DCs, mesenchymal stem cells, and bone marrow-derived mast cells [
13], it is also expressed by non-hematopoietic cells [
12,
14‐
18]. Expression of PD-L2 on the other hand is restricted to macrophages, DCs, bone marrow-derived mast cells, and peritoneal B1 cells [
12,
13,
19‐
21].
Our previous work demonstrated an increased expression of PD-1 on brain macrophages and microglial cells and PD-L1/2 on B cells from the spleen, blood, and CNS in mice after induction of experimental stroke by middle cerebral artery occlusion (MCAO). Thus, when MCAO was compared in PD-1 knockout (
-/-) versus wild-type (WT) male mice after 96 hours of reperfusion, cortical, striatal, and total infarct volumes were significantly larger in PD-1
-/- versus WT with substantial recruitment of inflammatory cells from the periphery into the CNS [
22], clearly implicating the role of the PD-1/PD-1 ligand (PD-L) pathway in limiting infarct volume in MCAO. Thus, the aim of the present study was to extend our previous studies and investigate the role of the PD-Ls, PD-L1 and PD-L2, in modulating severity of ischemic brain injury and associated CNS inflammation. For this purpose, PD-L1
-/- and PD-L2
-/- male mice, on a C57BL/6 background, were subjected to 60 minutes of MCAO followed by 96 hours of reperfusion and infarct volumes and immunological parameters were compared to similarly treated WT mice. Our results clearly demonstrate that PD-L1
-/- and PD-L2
-/- mice had lower total infarct volumes compared to WT mice. The immune parameters matched the stroke outcome in that the PD-L1
-/- and to a lesser extent PD-L2
-/- had reduced levels of proinflammatory activated microglia and/or infiltrating monocytes and CD4
+ T cells in the ischemic hemispheres. There was a reduction in ischemia-related splenic atrophy accompanied by lower activation status of splenic T cells and monocytes in the absence of the PD-L1, suggesting a pathogenic rather than a regulatory role for both PD-Ls. Also, suppressor T cells (IL-10-producing CD8
+CD122
+ T cells) trafficked to the brain in PD-L1
-/- mice possibly acting as key contributors of immunoregulation. Thus, these results clearly establish a role for PD-L1 and to a lesser extent PD-L2 in increasing infarct volumes in experimental stroke, making them potential targets for future immunotherapy.
Discussion
Ischemic stroke induces neurological deficits in almost one-third of patients, leading to increased mortality and long-term functional disability [
33,
34]. Although underlying mechanisms have not been completely unraveled, that ischemia evokes inflammatory responses has been characterized. This process depends in part on T cell activation, in which the B7 family of co-stimulatory molecules plays a pivotal role. T cells are localized in close vicinity to blood vessels in the infarct boundary as early as 24 hours after experimental focal cerebral ischemia in rodents [
8,
35]. This early (24-hour) appearance of T cell infiltration into the brain after MCAO [
8,
35] may indicate that recruitment of activated cells is antigen-nonspecific, perhaps generated by sympathetic signaling from the brain to the periphery. Alternatively, leakage of brain antigens across a compromised blood brain barrier could initiate a peripheral immune response. Activated T cells have the capacity to infiltrate the brain, and could contribute to expansion of the ischemic penumbra, an area that already contains infiltrating neutrophils after 24 hours. In our previous studies in severe combined immunodeficiency (SCID) mice examining the hypothesis that activated peripheral T cells and B cells would home to the injured brain and alter the evolving infarct, we found that both cortical and total hemispheric infarct volumes were strikingly reduced at 22 hours after MCAO [
36], thus indicating a damaging effect of T cells and B cells on early evolving ischemic brain injury. However, subsequent studies have convincingly demonstrated the regulatory role played by the B cells in our model of ischemic stroke [
3,
9]. Thus, this shifted the focus on further discerning the role of pathogenic T cells. The outcome of effector T cells is decided by the interaction of co-stimulatory molecules on T cells and APCs. Studies in the transient MCAO model [
37] point to the non-essentiality of the accessory TCR signaling via the PD-1/PD-L1 or CD28-B7 pathway in early stroke progression. However, since the end-point of these studies was as early as 22 hours post-stroke, it was necessary to investigate the role of the molecules that comprise the co-stimulatory pathway in T cell activation beyond the 22-hour post-MCAO window. Moreover, it is known that inflammation plays an important role in the pathogenesis of ischemic stroke, with stroke patients with systemic inflammation exhibiting clinically poorer outcomes [
38,
39]. Hence, in our earlier studies, we determined the role of the receptor, PD-1, known to be a part of a co-inhibitory pathway in T cell stimulation using PD-1
-/- mice which were subjected to 60 minutes of MCAO followed by 96 hours of reperfusion. These studies demonstrated a critical role for PD-1 in limiting functional and histological damage after MCAO [
7]. However the role of the two PD-Ls was not known. Therefore, to complete the picture, it was necessary to investigate the nature of interactions between PD-1 and both its ligands for understanding the susceptibility, pathogenic mechanisms, and protection afforded after ischemic stroke. Hence, the same time point of reperfusion, as that followed for our earlier studies in PD-1
-/- mice, was selected as a starting point of discerning the effects of interaction between PD-1 and its ligands in ischemic stroke.
In our current study, we hypothesized a critical role for the PD-Ls in exacerbating ischemic stroke. However, as clearly demonstrated in Figures
1 and
2, the loss of PD-L1 and to a lesser extent PD-L2 resulted in better stroke outcomes and decreased infiltration of immune cells in the ischemic hemisphere of PD-L1
-/- mice. The PD-1/PD-L pathway is recognized to control peripheral T cell tolerance in several ways. This pathway can limit the initial phase of activation and expansion of self-reactive T cells, and restrict self-reactive T cell effector function. The PD-1/PD-L interactions also play a role in inhibiting expansion of naive self-reactive T cells and/or their differentiation into effector T cells [
11]. Thus, PD-L1 and PD-L2 can signal bidirectionally by engaging PD-1 on T cells and by delivering signals into PD-L-expressing cells. Although PD-L1 has been consistently shown to be inhibitory in secondary phases of immune reactions [
40,
41], its role in primary immune reactions is less clear. Initial reports have attributed both stimulatory and inhibitory properties under conditions of primary immune reactions [
22]. However, a number of recent studies suggest a ‘proautoimmune’ role for PD-L1 rather than a suppressive function. For example, transgenic over-expression of PD-L1 on pancreatic beta cells enhanced autoimmunity instead of suppressing it [
42]. As a result of PD-L1 over-expression in beta cells, CD8
+ T cell proliferation was enhanced and immunological tolerance was broken, as mice developed spontaneous diabetes [
42]. In yet another study [
43], an unexpected beneficial effect from PD-L1
-/- DCs was demonstrated where intracerebral microinjections resulted in amelioration of subsequent experimental autoimmune encephalomyelitis (EAE) [
43]. Furthermore, this treatment was accompanied by amplified neuroantigen-specific CD8
+ Treg recruitment into the CNS, suggesting that the lack of DC-derived PD-L1 allows for the development (or recruitment) of regulatory CD8
+ T cells (CD122
+) in the CNS. Thus, by ‘inhibiting the inhibitors’, DC/APC expression of PD-L1 at the site of inflammation leads to exacerbated autoimmunity [
44]. Our finding in this current study match this aforementioned scenario, since as demonstrated in Figure
5A,B, there was a loss of the suppressor CD8
+ subpopulation from the periphery post-MCAO in the PD-L1
-/- mice, only to find its accumulation in the ischemic half of the brains.
A number of lines of evidence have suggested a receptor for PD-L1 (B7-H1) or PD-L2 (B7-DC), aside from PD-1. CD80 (B7-1) has recently been identified as a binding partner for PD-L1 [
11]. Surface plasmon resonance studies demonstrate specific and unique interaction between PD-L1 and CD80, with an affinity (approximately 1.7 μM) intermediate between the affinities of CD80 for CD28 (4 μM) and CTLA-4 (0.2 μM), and PD-L1 for PD-1 (0.5 μM). CD86 (B7-2) does not bind to PD-L1 or PD-L2, and PD-L2 does not bind to CD80. CD80/PD-L1 interactions can induce an inhibitory signal into T cells. Ligation of PD-L1 on CD4 T cells by CD80, or ligation of CD80 on CD4 T cells by PD-L1, delivers a functionally significant inhibitory signal. CD80 acts specifically through PD-L1 on the T cell in the absence of CD28 and CTLA-4. Thus, PD-L1 can exert an inhibitory effect on T cells either through CD80 or PD-1. Because both PD-L1 and CD80 are expressed on T cells, B cells, DCs, and macrophages, there is the potential for bidirectional interactions between CD80 and PD-L1 on these cell types. In addition, PD-L1 on nonhematopoietic cells may interact with CD80 as well as PD-1 on T cells to regulate cells. In this scenario, PD-L1 binds CD80 to compete off CD28 binding due to its higher affinity interaction. Not mutually exclusive, PD-L1 might also be required as a survival signal for brain-derived CD8
+ Tregs. Thus, as demonstrated in Figure
6A,B,C in our present study, an increased expression of CD80 on APCs in the WT and PD-L2
-/- mice might suggest an overriding CD80/CD28 interaction leading to T cell activation. Conversely, low CD80 expression in PD-L1
-/- mice and the increase in the total PD-L2 expression in the PD-L1
-/- mice as compared to the WT mice makes a plausible case for a PD-1/PD-L2 co-inhibitory interaction in the absence of PD-L1.
However, one also needs to keep in mind that the broad distribution of PD-L1 expression on hematopoietic and parenchymal cells suggests one or both cell types may be central to limiting autoimmune disease during multiple phases of the immune responses. Due to its broad expression pattern in lymphoid and non-lymphoid organs, the PD-L1/PD-1 pathway has been suggested to play a crucial role for the maintenance of immune tolerance [
45,
46]. In fact, the expression of PD-L1 on vascular endothelial cells has led to the hypothesis that PD-L1 on endothelial cells may regulate the activation of T cells that contact the vessel wall, the extravasation of T cells into tissue, and/or limit detrimental consequences of immunopathology [
12]. Therefore, further studies pertaining to the role of PD-L1 on the endothelial cells become necessary in determining the exact role of PD-L1 in ischemic stroke.
The present study has some limitations with regards to outcome evaluations. First, we used a neurological deficit scoring system to evaluate each animal. While this scoring system is sensitive enough to confirm ischemia and consequent ischemic injury in each animal, it may not be sensitive enough for evaluating subtle changes in functional outcomes due to experimental manipulations or treatments. Future follow-up studies will use more sophisticated and sensitive behavioral measures to evaluate functional outcomes at different time points following ischemic insult in PD-L1
-/-, PD-L2
-/-, and WT mice. Second, individuals performing infarct volume and immunological assessments were not blinded to genotype. This potentially could have led to subjective bias in assessing results. However, this seems unlikely given that our observed results suggested a pathogenic rather than the expected regulatory role for both PD-Ls. Lastly, differences in striatal infarct volumes between PD-L1
-/- and PD-L2
-/- mice could be due in part to altered glial proliferative responses as suggested by earlier studies [
47]. Evaluation of glial proliferation and the number of microglia and astrocytes surrounding the infarct will be done in future studies.
In summary, the current study conclusively demonstrates for the first time that PD-L1-/- and PD-L2-/- mice have smaller total infarct volume compared to WT mice. Improved stroke outcome was accompanied by significant reduction in the brain infiltrating cells as well as their proinflammatory states in the PD-L1-/- and to a lesser extent PD-L2-/- mice as compared to the WT mice. Our study also demonstrated a novel connection between the ischemic lesion in the brain and evolving inflammatory changes in distant peripheral immune cell populations. Reduction in infarct volumes promoted reduction in ischemia-related splenic atrophy accompanied by lower activated states of the splenic T cells and monocytes in the absence of the PD-L1, suggesting a pathogenic rather than a regulatory role for both PD-Ls. Suppressor T cells (IL-10-producing CD8+CD122+ T cells) trafficked to the brain in PD-L1-/- mice and their presence in the right MCAO-inflicted hemisphere of the PD-L1-/- mice implicates them as possible key contributors in controlling further adverse effects of ischemia. There was increased expression of CD80 on splenic APCs in WT and PD-L2-/- mice, suggesting an overriding interaction leading to T cell activation. Conversely, low CD80 expression by the APCs in PD-L1-/- mice suggests suppression possibly via PD-1/PD-L2 interactions. Our novel observations are the first to implicate PD-L1 in enhancing the severity of experimental stroke. These results suggest that agents (for example antibodies) that can target and neutralize PD-L1/2 may have therapeutic potential for treatment of human stroke.
Competing interests
The authors declare that they have no competing financial interests.
Authors’ contributions
SB performed and interpreted the immunology experiments, carried out statistical analyses, prepared graphics, and wrote the manuscript. YC performed the MCAO procedures, carried out statistical analyses, prepared the graphics, and wrote the methods and results for infarct volume data. AAV critiqued and edited the manuscript. SJM directed study design and data analysis of the MCAO experiments and edited the manuscript. HO directed the overall study, designed and supervised the immunological studies and data analysis, and edited the manuscript. All authors read and approved the final version of the manuscript.