**3. Biological Functions of C3G-Ms**

Only several C3G-Ms have shown potential biological function, although more than 20 kinds of C3G-Ms have been identified [8,30]. PCA and phloroglucinaldehyde (PGA) are considered as the major bioactive phenolic metabolites produced by phase Imetabolism, which undergo cleavage of the C ring of C3G. PCA can increase the antioxidant capacity of cells potentially by increasing the activity of antioxidant enzymes, such as catalase (CAT) in hypertensive rats or arthritis-model rats [31,32], superoxide dismutase (SOD) [33], and glutathione peroxidase (GPx) in mice or macrophages [33–36], and thus attenuate lipid peroxidation. Meanwhile, PCA has been reported to inhibit the production of inflammatory mediators, such as interleukin (IL)-6, tumor necrosis factorα (TNFα), IL-1β, and prostaglandin E2 (PGE2) [37–39], potentially by suppressing the activation of nuclear factor-κB (NF-κB) and extracellular signal-regulated kinase (ERK) [33,38] in murine BV2 microglia cells and colitis-model mice. PGA has also shown an inhibitory e ffect on inflammation potentially by modulating the production of IL-1β, IL-6, and IL-10 [40] in human whole blood cultures, although there are few reports about the molecular mechanisms. Our previous studies have revealed that both PCA and PGA are capable to down-regulate the MAPK pathway, especially suppress the activation of ERK, and PGA can directly bind to ERK1/2 [41] in murine macrophages.

Phase II metabolites of C3G, such as PCA-3-glucuronide (PCA-3-Gluc), PCA-4-glucuronide (PCA-4-Gluc), PCA-3-sulfate (PCA-3-Sulf), PCA-4-sulfate (PCA-4-Sulf), VA, VA-4-sulfate (VA-4-Sulf), isovanillic acid (IVA), IVA-3-sulfate (IVA-3-Sulf), and FA, are mostly derived from PCA and PGA [1,8]. VA and FA represent the bioactive phenolic metabolites based on recent studies. VA may suppress the generation of reactive oxygen species (ROS) [42] and lipid peroxidation [32], potentially by increasing the activity of antioxidant enzymes such as SOD, CAT, and GPx [43,44], as well as the level of antioxidants such as vitamin E [43,44], vitamin C [43,44], and glutathione (GSH) [45] in mice, hamster, and diabetic hypertensive rats. Additionally, VA can inhibit the production of pro-inflammatory cytokines such as TNFα, IL-6, IL-1β, and IL-33 by down-regulating caspase-1 and NF-κB pathways [45–47] in mice or mouse peritoneal macrophages and mast cells. FA has also been reported to attenuate both oxidative stress and inflammation potentially by suppressing the production of free radicals (ROS and NO in rats, rat intestinal mucosal IEC-6 cell, or murine macrophages) [48–50], enhancing Nrf2 expression and down-stream antioxidant enzymes (SOD and CAT in rats or swiss albino mice) [48,51], and inhibiting the activation of proinflammatory proteins (p38 and IκB in HUVEC cells) [52] and cytokines production, such as IL-18 in HUVEC cells [52], IL-1β in mice [53], IL-6 in obese rats [54], and TNFα in mice [53]. However, both VA and FA showed a limited e ffect on the activation of MAPK pathway and production of inflammatory cytokines, such as monocyte chemoattractant protein-1 (MCP-1) and TNFα in a high-fat diet-induced mouse model of nonalcoholic fatty liver disease [41]. Table 1 summarizes the biological functions of the main bioactive metabolites, including PCA, PGA, VA, and FA.




**Table 1.** *Cont.*

Notes: C3G-Ms, cyanidin-3-glucoside metabolites; CAT, catalase; COX-2, cyclooxygenase-2; ERK, extracellular signal-regulated kinase; FA, ferulic acid; GSH, glutathione; ICAM-1, intercellular adhesion molecule-1; LPS, lipopolysaccharide; MAPKs, mitogen-activated protein kinases; MCP-1, monocyte chemoattractant protein-1; MDA, malondialdehyde; NF-κB, nuclear factor-κB; NO, nitric oxide; PCA, protocatechuic acid; PGA, phloroglucinaldehyde; PGE2, prostaglandin E2; ROS, reactive oxygen species; T-AOC, total antioxidant capacity; VA, vanillic acid; VCAM-1, vascular cell adhesion molecule-1; SOD, superoxide dismutase; TNF-<sup>α</sup>, tumor necrosis factor-<sup>α</sup>.

#### **4. Crosstalk between Gut Microbiota and C3G&C3G-Ms**

Bacteria can use phenolic compounds as substrates to obtain energy [56,57] and to form fermentable metabolites which can exert bioactive functions similar to parent anthocyanins [58], and thus, gu<sup>t</sup> microbiota play an important role in the metabolism of anthocyanins and the secondary phenolic metabolites after the removal of anthocyanins' sugar moiety [59].

PCA has already been proven as the gu<sup>t</sup> microbiota metabolite of C3G [60], as *Lactobacillus* and *Bifidobacterium* have the maximum ability to produce the β-glucosidase so that anthocyanins are transformed to PCA [61]. *Lactobacillus* and *Bifidobacterium* are also observed to produce p-coumaric acid and FA under di fferent carbon sources [57,62], while *Bacillus subtilis* and *Actinomycetes* are involved in the bioconversion of VA to guaiacol [63].

On the other hand, anthocyanins are capable of modulating the growth of special intestinal bacteria [24] and increasing microbial abundances [64]. Anthocyanins have been reported to increase the relative abundance of beneficial bacteria such as *Bifidobacterium* and *Akkermansia*, which are believed to be closely related to anti-inflammatory e ffects [24,65]. Monofloral honey from *Prunella Vulgaris,*rich in PCA, VA, and FA, showed protective e ffects against dextran sulfate sodium-induced ulcerative colitis in rats potentially through restoring the relative abundance of *Lactobacillus* [66]. Our previous studies also found that the *Lonicera caerulea* L. berry rich in C3G could attenuate inflammation potentially through the modulation of gu<sup>t</sup> microbiota, especially the ratio of *Firmicutes* to *Bacteroidetes* in a mouse model of experimental non-alcoholic fatty liver disease [67]. Nevertheless, another study revealed that propolis rich in PCA, VA, and FA could suppress intestinal inflammation in a rat model of dextran sulfate sodium-induced colitis potentially by reducing the population of *Bacteroides* spp [68]. This may be because of the inhibitory and lethal e ffects on pathogenic bacteria by anthocyanins and their metabolites. PCA has been reported to inhibit the growth of *E. coli*, *P. aeruginosa*, and *S. aureus* [69]. VA can decrease the cucumber rhizosphere total bacterial *Pseudomonas* and *Bacillus spp.* community by changing their compositions [70]. FA is identified as highly e ffective against the growth of *Botrytis cinerea* isolated from grape [71]. Table 2 shows the microbial species that can biotransform C3G&C3G-Ms and the bacteriostasis e ffects of C3G-Ms.

The mechanisms underlying the anti-microbial e ffect of anthocyanins are not clear yet. Ajiboye et al. have pointed out that PCA may induce oxidative stress in gram-negative bacteria [69], that is, PCA can combine with O2 to form •O2<sup>−</sup>, which attacks the polyunsaturated fatty acid components of the membrane to cause lipid peroxidation, and attacks the thiol group of protein to cause protein oxidation. To be more precise, •O2<sup>−</sup> can be continually produced by autoxidation of PCA and semiquinone oxidation through the inhibition of NADH-quinone oxidoreductase (NQR) and succinate-quinone

oxidoreductase (SQR). Although SOD converts •O2<sup>−</sup> to H2O2, which can be finally changed to H2O and O2 by catalase, excessive H2O2 produces •OH during the Fenton reaction (Fe<sup>2</sup>+→Fe3<sup>+</sup>), and •OH attacks the base of DNA and results in DNA breakage. In addition, the suppression of NQR and SQR may lead to ATP depletion. Finally, bacterial death could be induced by lipid peroxidation, protein oxidation, DNA breakage, and ATP depletion. (Figure 2).


**Table 2.** Crosstalk between C3G&C3G-Ms and microorganism.

Given these, interactions between C3G&C3G-Ms and gu<sup>t</sup> microbiota can improve the bioavailability of C3G. C3G&C3G-Ms potentially ameliorate micro-ecological dysbiosis by inhibiting gram-negative bacteria. But it is worth noting that a few studies have demonstrated that the over-consumption of polyphenols had significant negative effects on reproduction and pregnancy [72–74]. Although it is inexplicit whether there is a correlation with the changes of gu<sup>t</sup> microbiota composition, the negative effects of polyphenols-mediated modulation of gu<sup>t</sup> microbiota should be focused on.

**Figure 2.** Potential mechanisms underlying the lethal effect of PCA on gram-negative bacteria. Autoxidation of PCA and semiquinone oxidation through the inhibition of NADH-quinone oxidoreductase (NQR) and succinate-quinone oxidoreductase (SQR) can cause ATP depletion and produce •O2<sup>−</sup>, which attacks the polyunsaturated fatty acid components of the membrane to cause lipid peroxidation and attacks the thiol group of protein to cause protein oxidation. Although SOD converts •O2- to H2O2, which can be finally changed to H2O and O2 by catalase, excessive H2O2 produces •OH during the Fenton reaction (Fe<sup>2</sup>+→Fe3<sup>+</sup>), and •OH attacks DNA bases to cause DNA fragmentation. Ultimately, lipid peroxidation, protein oxidation, DNA fragmentation, and ATP depletion induce bacterial death. PCA, protocatechuic acid; SOD, superoxide dismutase.

#### **5. The Potential Mechanisms of C3G&C3G-Ms against Intestinal Injury**

Multiple studies have shown that C3G&C3G-Ms have an essential role in intestinal health [55,75,76]. The potential mechanisms of C3G&C3G-Ms against intestinal injury are considered as they act in a synergistic manner between the antioxidant, anti-inflammatory, and anti-apoptosis function.
