Anti-inflammatory effects in intestinal epithelial cells
Given growing understanding of the importance of peripheral inflammation, gut-health and the gut-brain axis on cognitive health, the effects of sage extract were assessed in human small intestinal epithelial cells as a model of the intestinal barrier. Figures 1 and 2 summarise the levels of cytokines and chemokines released into apical and basolateral cell media respectively by human small intestinal epithelial cells incubated with LPS and sage extract for 24 hours. Compared to the control group, the LPS treatment caused a significant increase in the release of all eight of the protein markers tested in the apical media (Figure 1), and all but CRP and VCAM-1 in the basolateral media (Figure 2); which supports successful induction of pro-inflammatory response by the LPS treatment.
Within the apical media, the 5µg/ml and 20µg/ml doses of the sage extract showed significant (p<0.01) reductions in the levels of CRP after stimulation with LPS; returning it close to the basal level of release (44% and 33% reductions respectively; Figure 1A). The sage treatment also showed a dose-response attenuation of the induction of IL-8 levels in the apical media after LPS stimulation; reducing the release of IL-8 by 20% and 34% for the 5µg/ml and 20µg/ml extract doses respectively (Figure 1E). A significant decrease (22%) in the basal level of IL-8 release was shown for the 20 µg/ml extract dose treatment (Figure 1E), and both sage extract doses caused a significant reduction in the release of IL-6 under LPS induced conditions (16% and 17% respectively; Figure 1C). There was also attenuation of LPS-induced release for SAA, VCAM-1, and ICAM-1 (31%, 20%, and 23% reductions respectively) in response to the 5µg/ml sage extract treatment (Figures 1F, 1G and 1H).
Cytokine and chemokine levels were much lower in the basolateral media than the apical media (Figures 1 and 2). Within the basolateral media, the 5µg/ml sage extract dose showed a significant decrease of basal (69%) and LPS-stimulated levels of TNF-α (35%; Figure 2B). Conversely, there were significant increases in the basal levels of IL-6 for both sage extract doses in the basolateral media, as well as after LPS stimulation for the 20µg/ml sage extract treatment (Figure 2C). Similarly, SAA and IL-8 basal levels (p<0.001) were significantly increased by 20µg/ml treatment dose (Figures 2E and F), and the basal levels of ICAM-1 were increased in a dose responsive manner (p<0.05 and p<0.001; Figure 2G).
The role of cytokines in maintaining and modulating gut barrier function is complex (48, 49) and difficult to model; especially in an in vitro setting. However, the reduction of the apical release of cytokines in response to the sage treatment in the intestinal epithelial cell model, would translate to reduced levels within the intestinal lumen. Given that increased levels of inflammation markers within the intestinal lumen is linked to inflammation (50, 51), a reduction of such markers by the sage extract could contribute towards anti-inflammatory effects and a reduction in peripheral inflammation.
Anti-inflammatory effects in human brain microvascular endothelial cells
As well as reducing peripheral inflammation, we were interested in investigating the potential anti-inflammatory effects of the sage extract in reducing neuroinflammation. Integrity of the BBB plays an important role in maintaining homeostasis in the brain microenvironment. Inflammation can have negative effects on the function of the BBB, with IL-17 in particular linked to disruption of the BBB (17, 18). The release of cytokine and chemokines were assessed in human brain microvascular endothelial cells as a model of the BBB. Figure 3 shows the levels of cytokine and chemokine markers released into the media by human brain microvascular endothelial cells incubated with IL-17A, and sage extract (20µg/ml) or IL-17 antibody treatments for 24 hours. Consistent with induction of inflammation, compared to the control treatment the IL-17A treatment significantly increased the release of multiple inflammatory markers including MCP-1, SAA, ICAM-1, IL-6, and IL-8 (Figure 3). Although IL-17A treatment did not significantly increase the levels of CRP, both the sage extract and IL-17 antibody caused a significant reduction in the level of CRP in the IL-17A induced conditions (both 37%; Figure 3B). IL-17 antibody also caused an attenuation of the increase in release of MCP-1 (25%), SAA (32%), IL-6 (39%), and IL-8 (6%), but no effects were shown on VCAM-1 or ICAM-1, and TNF-α showed a significant increase in response to the antibody treatment (Figure 3).
In addition to the same effect on CRP under IL-17A induced conditions, the sage extract also showed a similar effect as the IL-17 antibody treatment on TNF-α levels, all be it in the opposite direction to CRP with both treatments instead showing an increase in release of TNF-α (Figures 3B and H). The sage extract also reduced the release of MCP-1 (16%); although this was only significant under basal conditions and not under IL-17A induced conditions as for IL-17 antibody (Figures 3A). However, conversely, the sage extract showed no effect on SAA, and caused slight increases to IL-6 and IL-8 levels in human brain microvascular endothelial cells, whereas the IL-17 antibody treatment had shown a reduction in these markers (Figures 3C, F and G).
Similar effects of the treatments on CRP, TNF-α, and to some extent MCP-1, suggest that the sage extract is attenuating the IL-17A induction, in a similar way to the IL-17 antibody treatment. However, the opposite effect of the sage and IL-17 antibody treatments on levels of IL-6 and IL-8, suggest that there are also differences in the mechanisms that are responsible. The action of independent mechanisms between the treatments is further supported by the sage extract showing significant reduction in the release of VCAM-1 under both basal and IL-17A-induced conditions (22% and 28% reductions respectively), whereas the IL-17 antibody treatment had no effect on this marker (Figure 3D).
The induction of ROS formation by IL-17A has been proposed as one mechanism leading to disruption of the BBB caused by this cytokine (17). Sage has previously been shown to have antioxidant activity; linked to high levels of phenolic and other active compounds present in the plant (23). Therefore, we also tested the anti-oxidant activity of the sage extract in human brain microvascular endothelial cells. Figure 4 shows the level of ROS production from Human brain microvascular endothelial cells incubated with sage for 24 hours. Treatment of the cells with H2O2 significantly increased ROS production after 24 hours compared to the control conditions, and the lowest doses of sage extract (1µg/ml and 5µg/ml) showed no significant effect on H2O2-induced ROS production (Figure 4). However, the 20µg/ml and 50µg/ml sage extract doses significantly reduced H2O2-induced ROS production by 31% and 19%, respectively, supporting anti-oxidant effects for the botanical extract that may contribute to reducing inflammation and maintaining BBB integrity (Figure 4).
BioMAP® Diversity PLUS®
To further investigate the anti-inflammatory effects of the sage across a broader set of cellular contexts, as well as seek new insights into potential mechanisms of action, the extract was assessed using the BioMAP® Diversity PLUS® Panel. This systems biology approach provides 148 biomarker readouts across 12 different primary cell-based disease models, advanced analytics, and a comprehensive reference database for insights on mechanism of action, efficacy and safety. The panel is typically applied in pharmaceutical development, but has been used by the U.S. Environmental Protection Agency, and has previously been shown to successfully work for plant extracts (43). Figure 5 summarises the results of profiling sage extract in BioMAP® Diversity PLUS®.
The extract showed dose-dependent effects on multiple inflammation and immunomodulatory markers across multiple cell systems; biomarker annotations included reductions in ICAM-1, VCAM-1, IL-17 and IL-10 among other markers (Figure 5). In the case of VCAM-1 this was reduced across three different cellular disease models; together with the data from the intestinal epithelial cells and vascular endothelial cells, as well as our previous data (32), this offers strong support for effects on this vascular injury marker. Also consistent with our previous results was the reduction in ICAM-1 in the HDF3CGF panel, and mild increases in IL-8 across the KF3CT (keratinocytes and dermal fibrobalsts), MyoF (lung fibroblasts) and /Mphg panels (Macrophages and venular endothelial cells; Figure 5).
The greatest number of annotated activities of the sage extract were shown in the HDF3CGF cell system; a model of wound healing and matrix/tissue remodelling, comprised of dermal fibroblasts stimulated with TNFα, IL-1β, Interferon-gamma (IFNɣ), Epidermal Growth Factor (EGF), basic Fibroblast Growth Factor (bFGF), and Platelet-Derived Growth Factor-BB (PDGF-BB). In addition to the reduction of inflammation markers in this panel, there was also a reduction of Plasminogen activator inhibitor-I (PAI-I), EGF receptor (EGFR), Collagen-I (Col-I) and Col-III; suggesting that the sage extract may also modulate tissue remodelling in response to perturbation (Figure 5). This cell panel, as well as the 3C (Venular endothelial cells stimulated with TNFα, IL-1β, and IFNɣ), and SAg (Venular endothelial cells and peripheral blood mononuclear cells stimulated with TCR ligands) also showed antiproliferative effects (Figure 5), which may relate to the tissue remodelling effects, but also to potential anti-cancer activities proposed for sage previously (23).
In addition to offering independent support for anti-inflammatory effects, and suggesting potential tissue remodelling and antiproliferative effects for the sage extract, the BioMAP® Diversity PLUS® Panel also provides the possibility to map a test agent’s response profile against a database of known agents to suggest potential mechanistic similarity with any of these agents. At the top dose tested the sage extract showed above statistically significant (r ≥ 0.7) threshold hits to antimicrobial and antifungal treatments, which is perhaps not surprising given traditional application in this area (23) (See Additional file 2). However, perhaps more interesting was the above threshold connection shown with caffeic acid at the 10µg/ml sage treatment dose, indicating that the sage extract may share some mechanistic similarity with this phenolic acid known to have anti-oxidant, anti-inflammatory and antineoplastic properties (52) (See Additional file 2).
Caffeic acid is synthesized broadly across plant species, including sage. The sage extract assessed in the current study is relatively rich in rosmarinic acid (4.2% w/w), which is an ester of caffeic acid and 3,4-dihydroxyphenyllactic acid. The extract also contains total phenolic levels of approximately 15.5% (w/w), which is consistent with the possibility that these compounds contribute towards the anti-inflammatory effects observed for the sage extract.
The sage extract assessed has previously been demonstrated to provide cognitive performance benefits in older adults as well as younger adults; which appears to be at least partially due to cholinergic properties through the inhibition of acetylcholinesterase (AChE) (26, 27). It has been proposed that this activity was due to monoterpenes including 1,8-cineole and α-pinene present in high levels in Salvia essential oils. These compounds have both been shown to inhibit AChE activity (both with IC50 estimates of approximately 0.67mM) (29, 35). However, phenolic acids provide alternative candidates to modulate this activity; with caffeic acid (IC50 estimates of 23.3µM (37)), and rosmarinic acid (IC50 > 300µM (36)) having reported inhibition of AChE activity in vitro. Such activity shared between the sage extract and caffeic acid might contribute towards the functional similarity suggested between these entities by the BioMAP Similarity Search analysis.
In addition to cholinergic activities, caffeic acid and rosmarinic acid have also been shown to potentially modulate the metabolism of monoamine neurotransmitters through inhibition of Monoamine Oxidase A (MAO-A; rosmarinic acid IC50 50.1µM, caffeic acid IC50 138.5µM (36)), MAO-B (rosmarinic acid IC50 184.6µM, caffeic acid IC50 247.7µM (36)), and Catechol-O-methyl transferase (COMT; rosmarinic acid IC50 26.7µM, caffeic acid IC50 89.9µM (36)). This is perhaps unsurprising given the structural similarity between caffeic acid and monoamines. Consistent with this, a previous analysis of the sage extract on the BioMAP® Diversity PLUS® Panel, run across a slightly different concentration range, showed a below threshold connection to two COMT inhibitor agents among the top three hits: Entacapone and Phenazopyridine (Additional file 3). The other agent in the top three was SR1001, a retinoic acid-related orphan receptor-α (RORα)/RORγ ligand, which could offer a potential mechanism for the IL-17-related effects of sage (53). However, it must be reiterated that these database hits were below the level of significance (r < 0.7), so they cannot be considered as sharing mechanistically relevant similarity based on this analysis alone.
Given the significant connection to caffeic acid, and previously demonstrated activities of this compound (36), the ability of the sage extract to inhibit COMT and MOA activity was assessed in vitro. Table 1 summarises enzyme inhibitory activity for the sage extract. Assay interference meant that is what not possible to assess the effect on MAO-A, but the sage extract did show inhibition of both COMT and MAO-B, with IC50 values of 31.2µg/ml and 84.2µg/ml respectively (Table 1; also see Additional file 4). Indeed, comparison with the AChE inhibition activity for the sage extract (IC50 1794.0µg/ml) suggested that the extract had greater potency for inhibition of metabolism of monoamine neurotransmitters than acetylcholine (Table 1).
Table 1
Inhibition of enzyme activity by sage extract
Enzyme | IC50 (µg/ml) |
COMT | 31.2 |
MAO-B | 84.2 |
AChE | 1794.0 |