1. Introduction
BBB injury associated with neurodegenerative and neuroinflammatory disorders is driven mainly by oxidative stress and inflammation induced by numerous inflammatory mediators that act both from the capillary lumen and the brain parenchyma. As in other barrier tissues, the mechanisms of the protective effect of SCFAs on BBB integrity are generally based on their antioxidant and anti-inflammatory actions mediated by inhibition of NF-κB and activation of Nrf2, a redox-sensitive transcription factor, important in counteracting the NF-κB-driven inflammatory response. Both nuclear factors compete for the same binding site in the nucleus and reciprocally influence each other’s expression and activity through a variety of mechanisms (for reviews, see
[1][2]).
2. SCFAs/HDAC/NF-κB
Numerous data provide evidence that inhibition of NF-κB transcriptional activity plays a central role in the anti-inflammatory effects of SCFAs, mainly butyrate, in various cellular types
[3][4][5][6][7] (
Figure 1). Excessive ROS production is a common trigger for the downstream pathways that mediate BBB leakage, rearrangement of the cytoskeleton and suppression of TJ proteins expression
[8][9][10]. Oxidative stress triggers nuclear translocation of NF-κB/p65, that promotes the transcription of a great diversity of pro-inflammatory genes, such as cytokines, chemokines, adhesion molecules, COX2 and iNOS enzymes, which affect all NVU components and disrupt the BBB integrity by numerous mechanisms, including suppression of the assembly and expression of TJ proteins
[11][12][13]. Inhibition of NF-κB activity restores the low permeability and upregulates the expression of TJPs, as was shown in different barrier tissues including the BBB
[14][15][16][17]. Overexpression of the NF-κB/p65 alone repressed claudin 5 promoter activity in mouse brain endothelial cells
[12]. SB was shown to inhibit NF-κB/p65 nuclear translocation alleviating inflammatory stimuli-induced damage of the barrier integrity in intestinal
[17][18][19], bovine ruminal
[20] and mammary
[21] epithelia. Inhibition of NF-κB activation under propionate was observed in the brain microvascular cell line hCMEC/D3
[22].
Figure 1. Mechanisms of SCFA effects on a brain endothelial cell (Part 1). Pathological stimuli activate NF-κB-mediated expression of multiple genes, including MMP-9 and NLRP3, and cause the degradation of TJ proteins. Activation of Nrf2 counteracts the NF-κB-driven inflammatory response. SCFAs suppress NF-κB and promote Nrf2 activation, leading to the recovery of TJPs. SCFAs can act after entry via transporters (MTC1, FAT/CD36) or via binding to receptors (e.g., GPR41). SCFAs inhibit HDAC, which results in an increase in histone acetylation, facilitating expression of genes, including Nrf2, that contribute to BBB integrity, and increase NF-κB acetylation, inhibiting its transcriptional activity. Under pathological stimuli, MLCK phosphorylates MLC2 to form actin stress fibers resulting in the endocytosis of transmembrane TJPs and loss of BBB integrity. SCFAs may promote the reassembly of TJs via suppression of the MLCK/MLC2 pathway. Dotted lines—putative signaling pathways. Blue lines—positive effects; red lines—negative effects. Symbols on the scheme—see in the list of abbreviations.
The underlying mechanisms of the SCFAs effect on inhibition of NF-κB are poorly understood. NF-κB activity is known to be modulated by its post-translational acetylation that depends on the balance between the HDACs and the histone acetyltransferase activities
[23][24]. Inhibition of HDACs can cause hyperacetylation of NF-κB/p65 leading to modulation of p65 binding to IκB and downregulation of NF-κB transcriptional activity
[25]. Such a mechanism of NF-κB signaling interruption and subsequent protection of the BBB has been demonstrated in a rat model of cerebral ischemia upon administration of valproate
[26], a broad HDAC inhibitor, or RGFP966, an HDAC3-specific inhibitor
[27], suggesting that SCFAs, being HDAC inhibitors, can suppress NF-κB by the same mechanism (
Figure 1). In the human colon adenocarcinoma cell line, Colo320DM, butyrate, propionate and acetate dose-dependently inhibited NF-κB reporter activity with the same rank of potency as HDAC inhibition (butyrate > propionate > acetate)
[7].
3. SCFAs/GPR43/β-Arrestin-2/NF-κB
The GPR43 receptor engages a signaling pathway mediated by β-arrestin-2 which directly interacts with IκB, the NF-κB inhibitor, and blocks IκB phosphorylation/degradation, thereby suppressing NF-κB downstream signaling
[28][29]. GPR43/β-arrestin-2-mediated suppression of NF-κB signaling pathway has been shown to be a mechanism of the anti-inflammatory effect of SCFAs in macrophages
[4] and microglia
[3]. The expression of GPR43 in brain microvessels has not yet been confirmed, so this mechanism may be rather important for the indirect effect of SCFAs on the BBB through the activation of GPR43 in microglia (see below).
4. SCFAs/NF-κB/NLRP3 Inflammasome
Activation of NF-κB induces NOD-, LRR- and the pyrin domain-containing protein 3 (NLRP3) inflammasome that causes the amplification of the inflammatory response, mainly by induction of IL-1β and caspase-1 expression
[30]. The NLRP3 inflammasome has been identified as a mediator of BBB disruption in sepsis-associated encephalopathy
[31]. In brain microvessel epithelial cells treated with LPS, downregulation of NF-κB-mediated NLRP3 activation was shown to restore TJ protein expression and cell monolayer permeability
[31]. SCFA-induced suppression of the NLRP3 inflammasome, resulting in the restoration of the barrier function and TJ protein expression, has been shown in the intestinal epithelium
[19][32][33].
5. SCFAs/NF-κB/MMP-9
NF-κB activation leads to upregulation of MMP-9, a member of the zinc-dependent endopeptidase family
[15][26][34]. In brain tissues, MMP-9 is a critically important contributor to BBB damage (
Figure 1). Under the action of pro-inflammatory cytokines, MMP-9 can be activated and secreted by recruited neutrophils
[35], pericytes, microglia, and brain microvascular endothelial cells resulting in the proteolytic damage of the extracellular matrix components, degradation of the basement membrane and TJ proteins
[15][16]. Both in vivo and in vitro data indicate that down-regulation of MMP-9 expression or activity in endothelial cells restores BBB disruption and elevates TJ protein expression
[15][16][36][37]. In brain microvascular endothelial cells, IL-1β-induced MMP-9 expression occurs via complex signaling pathways including ROS-triggered c-Src-mediated transactivation of the EGF receptor and subsequent upregulation of MAP-kinases (e.g., ERK1/2, p38, and JNK1/2), resulting in turn in activation of NF-κB and MMP-9 expression
[38].
Downregulation of the NF-κB/MMP-9 signaling pathway by SCFAs was demonstrated in a rat focal cerebral ischemic model
[26]. I.p.-injected butyrate significantly reduced nuclear translocation of NF-κB, strongly inhibited MMP-9 protein expression and activity followed by the restoration of protein levels of claudin-5 and ZO-1 in cortex and striatum
[26]. The inhibition of the NF-κB/MMP-9 pathway by butyrate was also observed in other cell types, such as IL-1β-inflamed chondrocytes
[39].
6. SCFA/Keap-1/Nrf2 Signaling Pathway
SCFA-induced suppression of excessive ROS production and oxidative stress observed in endothelial and other cell types is realized via numerous mechanisms, affecting both ROS producing and expression/activity of ROS eliminating enzymes
[40][41][42][43][44]. Ubiquitous defense networks against oxidative stress and inflammation involve the Keap-1 (Kelch-like ECH-associated protein)/Nrf2/ARE (antioxidant response element) signaling pathway, which promotes the expression of multiple antioxidant genes containing ARE in their promoter region (for reviews, see
[1][2][45]). In different cell types, the anti-oxidative effect of SCFAs depends on activation of the Nrf2 defense pathway
[40][43][44][45][46][47]. In microglia and mammary epithelial cells, butyrate-induced Nrf2 activation and oxidative stress inhibition are mediated by GPR109A
[40][46].
Although Nrf2 activation is regulated at multiple levels via various signaling pathways, the excessive ROS production is a well-known Nrf2 activator and trigger for transcription of Nrf2-target antioxidant genes
[48]. In the BBB, Nrf2 augments oxidative stress and preserves integrity by increasing TJ and AJ protein expression under different pathological conditions
[49][50][51][52][53] (
Figure 1). Silencing Nrf2 in hCMEC/D3 cells abrogated the expression of claudin-5 and VE-cadherin leading to an increase in permeability of the monolayers and to the decline of TEER
[53]. Pharmacological activation of Nrf2 signaling post-brain injury significantly restored the loss of TJ proteins and prevented BBB disruption
[49].
Nrf2-mediated attenuation of BBB disruption in rats following subarachnoid hemorrhage was observed after administration of mitoquinone (MitoQ)
[54], a mitochondria-targeted antioxidant that exerts protective effects in many diseases associated with oxidative stress
[55]. Although mitochondrial ROS can inactivate Keap1, promoting nuclear translocation of Nrf2
[56], a mitochondrial ROS scavenger also inhibits Nrf2 degradation and subsequently upregulates the antioxidant genes, thus protecting barrier integrity, as was demonstrated in various barrier-forming cells
[54][57][58][59]. The precise mechanisms of the MitoQ effect on Nrf2 activity are still poorly understood.
SCFAs, being HDAC inhibitors, can activate Nrf2 via alterations of histone acetylation state
[60][61]. Such a mechanism of the SCFA anti-oxidative effect has been convincingly demonstrated in different cells
[46][60][62]. In high glucose-treated aortic endothelial cells, SB inhibited HDAC activity and increased the occupancy of the transcription factor aryl hydrocarbon receptor and transcriptional adaptor P300 on the Nrf2 gene promoter, elevating Nrf2 mRNA/protein expression and alleviating oxidative stress and inflammation
[60]. The interaction between SB-induced HDAC inhibition and Nrf2-associated anti-oxidative capacity has also been studied in bovine mammary epithelial cells subjected to H
2O
2-driven oxidative stress. SB promoted Nrf2 nuclear accumulation and H3K9/14 acetylation through the AMPK signaling pathway. Chromatin immunoprecipitation assays detected that SB enhanced acetylation of histones associated with anti-oxidative genes such as Nrf2, HO-1, GCLC, GCLM, SOD1 and NQO1, leading to the increase in their transcription and oxidative stress alleviation
[46]. The contribution of HDAC to Nrf2 activity was also found to be important in the restoration of the BBB integrity. In a mouse model of T2DM, HDAC3 inhibition by its specific inhibitor RGFP966 upregulated miR-200a, thereby reducing the Keap1–Nrf2 interaction and promoting Nrf2 activation, which in turn significantly ameliorated the BBB permeability and TJ protein downregulation associated with T2DM
[62]. These data indicate that epigenetic modification of Nrf2 and downstream genes may be a potential mechanism of SCFA action in the restoration of the BBB integrity under oxidative stress.
The involvement of both NF-κB- and Nrf2-driven pathways in the effect of propionate at physiological concentration (1 μM) was shown in the hCMEC/D3 cell line, a human BBB in vitro model
[22]. Transcriptomic analysis revealed two particular clusters of pathways regulated by propionate treatment: those involved in the non-specific inflammatory response to microbial products, including NF-κB and Toll-like receptor signaling pathways and those involved in the response to oxidative stress
[22]. Propionate restrained TLR4 activation by inhibiting the expression of the accessory protein CD14 mRNA following the reduction in protein abundance on the cell surface, thereby disrupting TLR4 signaling. In addition, exposure of hCMEC/D3 cells to propionate resulted in the upregulation of a number of antioxidant genes observed in transcriptomic analysis, and these changes occur downstream of the transcription factor Nrf2. Propionate induced translocation of Nrf2 into the nucleus and diminished the level of intracellular ROS in parallel with the alleviation of the LPS-evoked marked disruption in the intracellular localization of occludin, claudin-5 and ZO-1
[22]. These data indicate that the protective effect of propionate on BBB integrity is based on the involvement of anti-inflammatory and anti-oxidative defense mechanisms.
7. SCFAs/HDAC/FoxO1/Claudin-5
HDACs are known to be activators of FoxO1, a transcription factor involved in the negative regulation of the expression of claudin-5, the most important sealing component of the TJs. Loss of HDAC activity prevents the nuclear accumulation of FoxO1, resulting in suppression of FoxO1 activity and removal of transcriptional repression of claudin-5. Pharmacological inhibition of HDAC1 activity by entinostat, a class I HDAC inhibitor, rescued claudin-5 expression in the BBB
[63]. Such a mechanism can mediate SCFAs-induced enhancement of claudin-5 expression in the BBB.
8. SCFAs / HDAC / PPAR γ
Inhibition of HDAC3 may promote acetylation of peroxisome proliferator-activated receptor γ (PPARγ) thereby increasing its transcriptional activity independently of ligand
[64][65] (
Figure 2). In models of cerebral ischemia, both in vitro and in vivo, PPARs have been shown to contribute to the protection of the integrity of the BBB, which is partially attributed to modulating TJ protein expression
[66][67][68]. In human brain microvascular endothelial cells subjected to oxygen-glucose deprivation in the presence of a specific HDAC3 inhibitor and a PPARγ antagonist, it was shown that HDAC3 inhibitor-induced reduction in increased endothelial permeability is at least partly mediated by promotion of PPARγ DNA binding activity
[66]. The effect of SCFAs (a mixture of acetate, propionate and butyrate in 3:1:1 ratio) on the prevention of high-fructose diet-induced intestinal epithelial barrier impairment in mice was mimicked by a selective PPARγ agonist
[69].
Figure 2. Mechanisms of SCFA effects on a brain endothelial cell (Part 2). Inhibition of HDAC by SCFAs leads to hyperacetylation of PPARγ and its translocation into the nucleus potentiating transcription of TJPs. Activation of the Wnt signaling pathway promotes translocation of β-catenin into the nucleus and upregulation of transcription of ZO-1, occludin and claudin-5. The crosstalk between SCFA signaling and the canonical Wnt/β-catenin pathway could be presumed. SCFAs can interact with microglia by binding to G-protein receptors or by inhibiting HDAC, suppressing their activation. Dotted lines—putative signaling pathways. Symbols on the scheme—see in the list of abbreviations.
9. SCFAs/Myosin Light Chain Kinase
BBB “opening” depends on the endothelial actin–myosin cytoskeleton which regulates junction assembly and function. Non-muscle myosin II binds to actin, which is linked via adapter proteins to TJ proteins providing a driving power for their rearrangement. Phosphorylation of the myosin light chain (MLC) by MLCK results in the contraction of the TJPs-associated actin filaments, leading to the relocalization of the TJ proteins and intercellular barrier opening
[70][71] (
Figure 1). Pharmacological inhibition of MLCK prevented an increase in BBB permeability induced by pathological stimuli, as was demonstrated both in vivo
[72][73] and in vitro
[73]. MLCK deficiency attenuates endothelial barrier dysfunction
[74]. Using primary brain microvessel endothelial cells isolated from mice null for MLCK, Beard and coauthors found that MLCK mediates IL-1β-induced claudin 5 repression and barrier dysfunction acting in a manner that promotes the nuclear translocation of β-catenin, resulting in the repression of Cldn5 gene expression
[75]. Thus, myosin light chain kinase (MLCK) is now considered a crucial regulator of endothelial permeability and a potential therapeutic target in the treatment of tissue barrier dysfunction.
A link between SCFAs and MLCK was studied in the intestinal epithelium. In Caco-2 cell monolayers, SB had no effect on the expression level of several components of the TJ complex, but promoted the reassembly of TJ proteins through AMPK-dependent suppression of MLCK activity and subsequent decrease in the phosphorylation of MLC2
[76]. Data obtained on fish intestine subjected to infection showed that SB dietary supplementation maintained the intestinal epithelium integrity decreasing the level of MLCK mRNA in the intestine
[47] indicating that the protective effect of SB is at least partly mediated by MLCK inhibition.
Given the critical importance of an MLCK-dependent mechanism in BBB function and dysregulation, the study of the potential involvement of SCFAs, mainly butyrate, in MLCK regulation in the BBB is a promising direction.
10. SCFAs/Wnt/β-Catenin
Both canonical and non-canonical Wnt pathways are involved in the regulation of BBB integrity (
Figure 2). The canonical Wnt/β-catenin signaling is critically required for the formation of BBB-specific barrier properties in endothelial cells during brain angiogenesis
[77] and is essential for the maintenance of the BBB integrity in the mature brain via transcriptional regulation of the expression of TJ proteins
[78]. In the canonical Wnt signaling pathway, Wnt inhibits the degradation of β-catenin through its binding to Frizzled receptors, promoting its translocation into the nucleus and upregulation of genes critical for the maintenance of the BBB, such as claudin-1 and -3
[78][79]. The noncanonical β-catenin-independent Wnt pathways, which include Wnt/calcium and Wnt/planar cell polarity (PCP) pathways, contribute to the regulation of TJ complex stability and endothelial cell polarity in the BBB
[80].
Despite the critical importance of the Wnt signaling pathway for the integrity of the BBB, researchers found only one work linking the protective effect of SCFAs in the BBB with its activation
[81]. In aged mice subjected to anesthesia/surgery, SB or Lactobacillus bacteria mixture improved postoperative cognitive function, reduced BBB permeability and increased the expression of occludin, claudin-5 and ZO-1. β-Catenin protein expression in the hippocampus was significantly enhanced under both treatments, suggesting crosstalk between SCFA signaling and the canonical Wnt/β-catenin pathway
[81].