Exploiting the well-known immunomodulatory potential of MSC, this study is the first, to our knowledge, to examine the influence of intravenously infused MSC on the activated kynurenine pathway (KP) and glutamate neurotransmission. Here, we report that peripherally administered HUCPVC regulate KP enzymes and metabolites in the LPS-activated CNS. Furthermore, these MSC were also found to exert a modulatory effect on the expression profile of glutamate receptor subunits and glutamate transporters. Thus, this study lends a novel approach to target aberrant signaling of KP and the subsequent glutamate excitotoxicity that are known to have diverse neuropathological consequences.
Inflammation-associated upregulation and activation of IDO (product of ido1 gene) in the brain is a critical step in the break down of tryptophan to kynurenine and initiation of the KP (Fig. 1A) [11, 43, 44]. Moreover, IDO activation is shown to be crucial for depressive-like behavior in mice treated with LPS for 24h [11, 32]. We have previously reported that HUCPVC modulate neuroinflammation and depressive behavior after 24h of LPS injection in mice [23]. In this study, we report that HUCPVC regulate IDO at the transcriptional level and the enzyme’s activation in the brain, as assessed by kynurenine:tryptophan ratio. This finding informs us of the possibility that the modulation of inflammation-associated depressive behavior by HUCPVC could be due to their ability to influence the catalytic function of IDO. Kynurenine synthesized by IDO can be a favorable substrate for either astrocytic enzyme KAT leading to the production of neuroprotective KYNA or microglial enzymes KMO and HAAO leading to the production of neurotoxic QUIN. Consistent with other studies [45, 46], we recorded an LPS-induced aberration in the brain mRNA levels of Kmo, Haao and Kat2, which is a predominant isomer of the KAT enzyme responsible for KYNA production [47]. Since the inadequate metabolism of QUIN by the enzyme QPRT contributes to increased neurotoxicity [48], we also tested the brain mRNA level of the enzyme, which was sharply downregulated by LPS. Thus, the recovery of an LPS-activated imbalance of the KP enzymes by HUCPVC treatment indicates the capability of MSC to influence the KP enzymatic machinery of glia governing the central production of KP metabolites.
KP metabolite levels in the brain have been intensely interrogated owing to their strong affiliation with many CNS disorders [1, 49]. In the current study, we evaluate the downstream immunomodulatory effect of intravenously administered HUCPVC on the neurotoxicity index, as assessed by the brain QUIN/KYNA ratio. Since the majority of brain kynurenine is peripherally derived during inflammation [13], plasma levels of the KP metabolites were also evaluated. Furthermore, to delineate the potential MSC-associated immunomodulatory effect, human foreskin-derived fibroblast cells (HS68), which have a relatively low level of immunomodulatory potential compared to MSC [50, 51], were independently injected into a group of animals. Our results indicate that compared to the brain, IDO activation in the plasma was not sufficient to increase the flux of circulating kynurenine. We hypothesize that this mild increase in kynurenine:tryptophan ratio, as shown in this study and by others [52], could merely be the statistical outcome of non-significant changes in tryptophan or kynurenine levels, which likely holds little metabolic or clinical significance. This lack of increase in plasma kynurenine may be due to its rapid clearance by the kidney and excretion of its metabolites in urine [53, 54]. Sufficient IDO-induced tryptophan oxidation is required to exceed the effect of renal processing and result in appreciable levels of plasma kynurenine. Moreover, kynurenine being a substrate to KMO, the inflammation-induced expression and/or activity of KMO can counter the effect of IDO. In addition, since circulating proinflammatory cytokines are shown to peak within 1–6 hours of LPS treatment [42, 55], arguably, these early timepoints could correspond to a relatively higher plasma IDO activity, as shown by Wirthgen et. al [56]. Thus, the low plasma levels of other KP metabolites in the current study could be the downstream effect of this lack of increase in kynurenine. Conversely, in the brain, we noted significant aberration of KP metabolite levels due to LPS. Significantly altered KP metabolites in the brain due to LPS were rescued back to basal levels by HUCPVC treatment; an effect not seen for the most part with fibroblast cells. These findings reveal significant and specific immunomodulatory effect of MSC on KP metabolism in response to LPS; thus maintaining homeostasis between two functionally contradictory branches of the pathway. Interestingly, and in accord with other studies [57], we found an increase in tryptophan by LPS. This increase, which could be due to LPS-induced lipolysis resulting in increased availability of albumin-free tryptophan to cross the blood-brain barrier (BBB) [58], was reversed by HUCPVC. Cytokine-stimulated IDO activation is also known to have a negative impact on serotonin (5HT) turnover [59]. Since the level of 5-hydroxy indole acetic acid (5HIAA), the metabolite of 5HT was not examined in this study, the unchanged 5HT level does not reflect its actual turnover and thus does not preclude the possibility of inflammation-afflicted modulation of serotonergic neurotransmission.
The nexus between activated cerebral KP metabolism and NMDAR activation, leading to enhanced glutamate function, is associated with many neuropathological conditions [60]. The observed HUCPVC-induced modulation of KP metabolites in this study, notably that of QUIN, an endogenous NMDAR agonist, provoked further investigation into the expression profile of the obligatory subunits of NMDAR, that mediate the receptor’s activity, and whose upregulation is implicated in various brain pathologies, including inflammation-related depressive phenotype [61]. In this study, we report a significant increase in the transcript levels of the subunits, NR2A and NR2B, in response to LPS, which resonates with similar observations by others [62, 63]. The demonstrated ability of HUCPVC to modulate these subunits may suggest a novel and relatively safer therapeutic alternative to the NMDAR subunit-targeting antidepressants that are shown to have psychoactive side effects and cardiovascular toxicity [64]. Glutamatergic circuitry is also negatively impacted by the perturbed transport mechanism responsible for the clearance of synaptic glutamate, leading to glutamate excitotoxicity [28]. Our study, for the first time, illustrates the potential of MSC, specifically HUCPVC, in regaining the LPS-induced decline in the astrocytic glutamate transporter EAAT2. Since the levels of glutamate (Glu) and glutamine (Gln) in the blood and brain also reflect glutamate excitotoxicity [65], we tested these metabolites in the plasma and whole-brain homogenate by LCMS. However, levels of Glu and Gln showed no significant changes between the control and LPS treatment groups. Since Glu and Gln cross BBB [66] and are expressed differentially in different brain regions [67], the lack of modulation in Glu and Gln levels reported in this study could be the consequence of the limitations of the methodology, which is unable to distinguish between the source of the metabolites in the plasma (central vs peripheral) or to delineate the region-specific expression of Glu and Gln in the brain. We further extended the scope of this study to investigate the immunomodulatory effect of systemically infused MSC on the neuroinflammation-associated synaptic imbalance implicated in neurodegenerative and psychiatric illnesses, including depression [68]. The effect of peripherally administered HUCPVC on LPS-induced dysregulation of synaptic markers such as drebrin, synaptophysin, PSD95, and NMDAR regulatory subunit NR2B, was tested in a synaptosomal isolate. Drebrin, an actin-binding protein in dendritic spines and one of the key players in the NMDAR-dependent synaptic neurotransmission, has been shown to be negatively regulated by the neuroinflammatory cascade associated with neurodegenerative diseases and psychiatric disorders [40, 69]. The HUCPVC-mediated restoration of LPS-induced downregulation of drebrin expression suggests a novel potential application of MSC in restoration of synaptic loss. Furthermore, we report the upregulation of NR2B expression in response to LPS. This corroborates the previous findings that the proinflammatory cytokines in the brain facilitate the activation of the NMDAR subunit [70]. Pharmacological suppression of NR2B has been achieved using various selective antagonists of the NMDAR subunit in various brain pathologies [71]. However, these pharmacological agents cause undesirable side effects including neurotoxicity and hypertension [72]. Thus, our findings support the possibility that MSC may be a potentially safer and more efficient therapeutic alternative to target NR2B.
There has been a considerable lack of understanding of the basis of systemically administered MSC’s ability to provide neuroprotection in many diseases and injury models. In our previous study using an LPS-induced mouse model of neuroinflammation and depression, we have demonstrated phagocytosis-driven immunomodulation by peripherally infused HUCPVCs [23]. The resulting systemic innate immune alteration from pro- to anti-inflammatory phenotype was plausibly correlated to the mitigation of LPS-induced neuroinflammatory response and depressive symptoms. In the current study, modulation of LPS-activated microglia via the kynurenine pathway in the CNS by HUCPVC may be another mechanism for peripheral immune modulation by the MSCs. Considering our previous findings that peripherally infused MSC fail to cross BBB [22], an alternate mechanism of MSC neuroprotective potential is its paracrine action, by which secreted factors are shown to mediate neuroprotection [73–75]. Contextually, studies showing the interplay between the kynurenine pathway and MSC support the involvement of IDO in the immunosuppressive effect of MSC [76]. Moreover, KYNA is also shown to regulate the expression of TNF-stimulated gene 6 (TSG-6), a paracrine factor, and promote TSG-6-mediated immunosuppressive and anti-inflammatory effects of MSC [77]. However, the modulation of a broad spectrum of activated kynurenine metabolites and the downstream signaling consequences vis-à-vis the glutamatergic system by MSC, as shown in this study, represents a novel finding.