Proanthocyanidins and Flavan-3-ols in the Prevention and Treatment of Periodontitis—Immunomodulatory Effects, Animal and Clinical Studies

This paper continues the systematic review on proanthocyanidins and flavan-3-ols in the prevention and treatment of periodontal disease and covers the immunomodulatory effects, and animal- and clinical studies, while the other part discussed the direct antibacterial properties. Inflammation as a major response of the periodontal tissues attacked by pathogenic microbes can significantly exacerbate the condition. However, the bidirectional activity of phytochemicals that simultaneously inhibit bacterial proliferation and proinflammatory signaling can provide a substantial alleviation of both cause and symptoms. The modulatory effects on various aspects of inflammatory and overall immune response are covered, including confirmed and postulated mechanisms of action, structure activity relationships and molecular targets. Further, the clinical relevance of flavan-3-ols and available outcomes from clinical studies is analyzed and discussed. Among the numerous natural sources of flavan-3-ols and proanthocyanidins the most promising are, similarly to antibacterial properties, constituents of various foods, such as fruits of Vaccinium species, tea leaves, grape seeds, and tannin-rich medicinal herbs. Despite a vast amount of in vitro and cell-based evidence of immunomodulatory there are still only a few animal and clinical studies. Most of the reports, regardless of the used model, indicated the efficiency of these phytochemicals from cranberries and other Vaccinium species and tea extracts (green or black). Other sources such as grape seeds and traditional medicinal plants, were seldom. In conclusion, the potential of flavan-3-ols and their derivatives in prevention and alleviation of periodontal disease is remarkable but clinical evidence is urgently needed for issuing credible dietary recommendation and complementary treatments.


Introduction
In the previous paper, we demonstrated that both free flavan-3-ols and oligomeric proanthocyanidins are very promising constituents for combating various bacteria involved in periodontitis pathogenesis [1]. Here, using the same systematic approach, we have selected and discussed the recent data on anti-inflammatory and immunomodulating activities of these compounds, including in vivo models and clinical studies.
According to the latest concept of periodontitis etiopathology, the development of the disease requires the co-existence of dental plaque that accumulates on the teeth surface and the patient's immune-inflammatory response [2]. Periodontal bacteria cause the mobilization of innate immune response (e.g., large phagocytes like macrophages, antigenpresenting dendritic cells (DCs), natural killer (NK) cells and neutrocytes) as well adaptive immunity mechanisms (T cells and B cells) that leads to the release of pro-inflammatory

Methods
Search strategy, as well as inclusion, exclusion criteria, and data organization are described in our previous review [1], in which the antimicrobial activity of proanthocyanidins and flavan-3-ols in the prevention and treatment of periodontitis is discussed. In brief, a systematic review in compliance to PRISMA guidelines was performed. An electronic database search was conducted using PubMed, Web of Science and Scopus (accessed 23 December 2020).
The search terms included all combinations of the following key words: 'periodontitis' OR 'periodontal diseases' OR 'gingivitis' OR 'gingival diseases' AND 'proanthocyanidins' OR 'condensed tannins' OR 'flavan-3-ols' OR 'catechin' OR 'epicatechin' AND 'anti-bacterial' OR 'antiadhesive' OR 'anti-inflammatory', respectively. All titles with abstracts were imported into a citation manager program "Mendeley" (Elsevier, London, UK), and all duplicates were removed. References of imported articles were also screened for other relevant studies. Two investigators (N.-H.I. and K.-R.P.) independently reviewed the titles and abstracts of the imported references to determine whether they met the inclusion and exclusion criteria (Figure 2). Finally, 31 studies were reviewed in [1] (antibacterial effects) and 50 studies are included in the present paper ( Figure 2).

Influence on Matrix Metalloproteinases (MMPs)
It was already proved that matrix metalloproteinases (MMPs) play a very important role in the process of periodontal connective tissue destruction. MMPs are a calciumdependent zinc-containing endopeptidases, responsible for the tissue remodeling and degradation of the extracellular matrix (ECM), including collagens, elastin, gelatin, matrix glycoproteins, and proteoglycan [6]. Major cell types that can be found in the periodontium, like fibroblasts, neutrophils, and macrophages, release these proteolytic enzymes [7]. These proteolytic enzymes are secreted as latent proenzymes (except membrane type (MT)-MMPs) and they must be later activated by tissue, plasma or bacterial proteinases extracellularly or at the cell surface. Under normal condition, MMPs play an important role in the healing of the wounds, the process of angiogenesis, and remodeling of the gingival tissue [8]. However, in periodontitis host cells are threatened by the periopathogens and their products such lipopolysaccharides (LPS) of Gram-negative bacteria. As a result, an increased production of MMPs can be observed that influences the degradation of periodontal ligaments, the loss of gingival collagen, and the resorption of alveolar bone, leading to destruction of periodontal tissues [9]. Increased activity of the enzymes-collagenases (matrix metalloproteinases 1 and 8) and gelatinases (matrix metalloproteinases 2 and 9) have been described in in the gingival crevicular fluid and in the inflamed gingival tissues of patients with periodontitis [10]. MMPs have the ability to activate and inactivate or even antagonize the biological functions of cytokines and chemokines. Through the influence on the cytokines and chemokines MMPs can regulate the inflammatory process by promoting or suppressing it. On the other hand, when the inflammatory cells are stimulated by cytokines and chemokines a production of MMPs may be induced [11]. Many studies proved that proanthocyanidins could inhibit activity of matrix metalloproteinases (Table 1). All the tested cells were exposed for 24 h to different cranberry concentrates (25,50, and 100 µg/mL). After 0, 3 and 7 days cell viability assay was performed.
After 24 h exposure, the HGF, SAOS-2, and macrophages viability was not reduced by the cranberry concentrates; Expression of proinflammatory IL-8 and IL-6 was downregulated by 50 µg/mL and 100 µg/mL PAs, but expression of anti-inflammatory IL-10 was upregulated at 100 µg/mL. No influence on expression of IL-1β was seen. Exposed LPS-stimulated macrophages to PAs significantly decreased M1 polarization and increased M2 polarization.
HPDLFs used in the study were treated by the tumor necrosis factor-alpha (TNF-α), PA, or their combination. The mineralization markers and osteogenic differentiation markers and associated markers were detected by quantitative real-time polymerase chain reaction (qRT-PCR), alizarin red S staining, and alkaline phosphatase (ALP) activity assay.
PA (0.1, 1, 10 µg/mL) significantly upregulated expression of osteogenesis-related genes and proteins and ALP activity in HPDLF compared with the control group in the non-inflammatory environment. PA (1 µg/mL) has significantly reversed the inhibition of osteogenesis-related gene and protein expression, ALP activity, and mineralization caused by TNF-α. PA could regulate osteogenesis of HPDLFs by suppressing the NF-κB signaling pathway.
S-G cells were incubated with IL-1β in the presence or absence of NDM or inhibitors of NF-κB-NBD or AP-1-SP600125. The IL-6 levels were measured by ELISA kit. Effects of NDM on IL-1β-activated NF-κB and AP-1 and phosphorylated intermediates in both pathways were measured by ELISAs.
Production of IL-6 was increased in IL-1β stimulated S-G cells. NDM was stronger inhibitor of IL-6 production in IL-1β stimulated S-G cells than either NBD peptide or SP600125 alone and was similar to the combination of these inhibitors. IL-1β stimulated NF-κB and AP-1 activation was inhibited by NDM. However, NDM did not affect IL-1β-stimulated levels of phosphorylated intermediates in the NF-κB pathway (IκBα) or the AP-1 pathway (c-Jun, ERK1/2).
S-G and normal human gingival fibroblasts were incubated with NDM, IL-17, or NDM+IL-17. IL-6 and IL-8 in culture supernatants were measured by ELISA.
In both cell lines, IL-17 has been found to significantly stimulate the production of IL-6 and IL-8. Non cytotoxic levels of NDM (5-50 µg/mL) inhibited constitutive IL-6 and IL-8 production as well their IL-17-stimulated cytokine production by epithelial cells and fibroblasts.
This study investigated the influence of the AC-PAs on osteoclast formation and bone resorption. The degree of osteoclast formation was assessed by quantification of TRAP-positive stained multinucleated cells, while the secretion of IL-8 and MMP-2, MMP-9 was measured with ELISA. Bone resorption was evaluated using a human bone plate coupled with an immunoassay that detected the release of collagen helical peptides. MTT assay was used to measure the cytotoxic effect of AC-PAs on osteoclastic cells.
AC-PAs at 10 µg/mL, 25 µg/mL, and 50 µg/mL caused a 38%, 84%, and 95% inhibition of RANKL-dependent osteoclast differentiation, respectively. AC-PAs increased the secretion of IL-8 and inhibited the secretion of both MMP-2 and MMP-9 in a dose-dependent manner. AC-PAs significantly decreasing the release of collagen helical peptides suggested that can prevent bone resorption. AC-PAs did not exhibit any toxic effect on osteoclastic cells from 10 µg/mL to 100 µg/mL. KB cells were pretreated with MF (10 µg/mL and 100 µg/mL) and infected with P. gingivalis. The cytokine gene expression was monitored using RT-qPCR and IL-6 level using ELISA.
10 and 100 µg/mL of MF significantly decreased (upregulated by P. gingivalis) gene expression for IL-1β, IL-8 and TNF-α, but not IL-6 compare to control cells (not exposed to tested extract). However, preincubation of the KB cells with MF before exposure to P. gingivalis resulted in significant lower concentration of IL-6 in the cells than in MF-untreated control group.  Gingival crevicular fluid (GCF) collected from periodontitis patients; purified collagenase; in vitro study.
Ethyl acetate fraction and 6 isolated catechins were tested for their ability to inhibit purified collagenase activities using collagenase of Clostridium histolyticum and supernatant of Porphyromonas gingivalis as well collagenase activity in GCF. Collagenase activity was determined using a commercially available kit.
The complete inhibition of the collagenase activity was achieved by using a 100 µg/mL of ethyl acetate fraction from tea and 50 µg/mL ECG, EGCG. Other catechins, without the gallate residue had no effect on collagenase.
Makimura et al., 1993 [46] Particularly, the n-butanol fraction from the Ulmus macrocarpa Hance bark, defined as elm extract (contain 20% of procyanidins) and the mixture of procyanidin oligomers (composed of 3 to 12 flavan-3-ol monomers, with average molecular weight of 1518) isolated from elm extract in range 100-1000 µg/mL exhibited inhibitory effects on the MMPs, present in gingival crevicular fluid (GCF) of adult patients with periodontal disease (mainly, MMP-8 and MMP-9) and on the pro and active forms of MMP-2 (from the conditioned media of cultured periodontal ligament (PDL) cells treated with a periodontopathogen, Treponema lecithinolyticum) [10]. The inhibition of enzyme activity by procyanidin oligomers was more effective than by the elm extract, with IC50 values 25 and 33 µg/mL for GCF collagenases (mostly MMP-8 and MMP-9) and MMP-2, respectively. Moreover, elm extract and procyanidin oligomers inhibited proteolytic enzymes of two periopathogens, T. denticola, and P. gingivalis responsible for degradation of the interstitial and basement membrane collagens as well as activating of matrix metalloproteinases e.g., MMP-8, MMP-9 or MMP-1, MMP-3, and MMP-9 [10].
In opposition to the above results are results obtained by Lombarto et al. who showed [23] that AC-PAs and EGCG, individually or in combination, had no effect on the regulation of MMP (1, 2, 3, 7, 8, 9, 10, 12, and 13) and tissue inhibitors of metalloproteinases -TIMP (1, 2, 3, and 4) secretion but inhibited the secretion of several cytokines in the (3D) co-culture model of gingival epithelial cells and fibroblasts stimulated with A. actinomycetemcomitans LPS (Table 1).

Influence on Bone Tissue Resorption
Yun et al. [44] reported an inhibitory effect of EGCG (20 µM) on the MMP-9 gene expression in osteoblasts and on the formation of osteoclasts, which suggested that EGCG may prevent the alveolar bone resorption that occurs in periodontal diseases leading to teeth loss. Importantly, in the periodontal disease an enhanced osteoclastogenesis can occur due to the presence of the of inflammatory cytokines that stimulates osteoclast proliferation or promotes the differentiation of progenitor cells. Mature osteoclasts that derive from hematopoietic monocyte/macrophage precursors under the action of RANKL (receptor activator of NF-κB ligand) and M-CSF (macrophage colony-stimulating factor) mediate the destruction of the alveolar bone by attaching to the bone surface and promoting mineral dissolution. The demineralized bone matrix is later degraded by proteases such as cathepsin K and metalloproteinases (MMPs) [32]. Tanabe et al. [32] showed that AC-PAs have influence the osteoclast formation and bone resorption activity. In a range of 10-100 µg/mL, AC-PAs inhibited RANKL-dependent osteoclast differentiation, as well as secretion of both MMP-2 and MMP-9 but secretion of IL-8 was increased. IL-8 from normal human bone marrow stromal cells inhibits the bone resorbing activity of osteoclasts [47].
Huang et al. [13] reported that proanthocyanidins (PA) may have an influence on the bone regeneration in the host inflammatory microenvironment by suppressing NF-κB signaling pathway and therefore may be a potential inducer of bone regeneration. In this study an effect of proanthocyanidins on osteogenic differentiation of human periodontal ligament fibroblasts (PDLFs) with or without TNF-α stimulation was tested and the biological mechanism was explored. The assumption was that periodontal ligament fibroblasts are capable of differentiating into osteoblasts, but pro-inflammatory cytokines like TNF-α inhibit this process. Osteogenic differentiation and mineralization associated markers were detected by qRT-PCR, alizarin red S staining, and alkaline phosphatase (ALP) activity assay. In result, proanthocyanidins in low concentration (0.1 µg/mL, 1 µg/mL, 10 µg/mL) significantly upregulated expression of osteogenesis-related genes and proteins and ALP activity in PDLFs compared with the control. However, proanthocyanidins at higher concentrations of 30 µg/mL and 50 µg/mL significantly suppressed the alkaline phosphatase activity of PDLFs. For the rest assay, authors used only lower concentration of PA (0.1 µg/mL, 1 µg/mL, 10 µg/mL). Proanthocyanidins in concentration of 1 µg/mL significantly reversed inhibition of osteogenesis related gene and protein expression, alkaline phosphatase activity, and mineralization caused by TNF-α. The authors also suggested that proanthocyanidins may reverse TNF-α inhibited osteogenic differentiation via NF-κB signaling pathway. These authors used commercial proanthocyanidins claimed to possess an untypical for proanthocyanidins structure ( Figure 3) and with molecular weight = 594.52. The supplier's website states that proanthocyanidins have been isolated from grapes (the fruits of Vitis vinifera L.).

Influence on Cytokines
The overproduction and secretion of inflammatory cytokines by resident and immune cells modulate the progression and severity of periodontitis. Increase of such proinflammatory cytokines as: IL-1α, IL-1β, TNF-α, IL-6, and IL-17 were shown in patients with acute or chronic periodontitis [48]. More specifically, TNFα can be found at high levels in gingival crevicular fluid (GCF) and in diseased periodontal tissues, where it is positively correlated with MMPs and RANKL expression. Human and animal studies confirmed that TNF-α plays a central role in inflammatory reaction, alveolar bone resorption, and the loss of connective tissue attachment. Moreover, TNF-α up-regulates the production of other proinflammatory innate immunity cytokines, such as IL-1β and IL-6 associated with inflammatory cell migration and osteoclastogenesis [48]. IL-1β plays an important role in the pathogenesis of periodontitis also by regulation of the IL-6 production in a variety of cell types, including fibroblasts and epithelial cells [26]. Similar to bacterial LPS, cellular response to cytokines or chemokines (e.g., IL-1β) can be mediated via signaling cascades, including NF-κB and MAPK/AP-1 pathways, which lead to gene expression of certain proteins (for example IL-6). There are more and more studies proving that proanthocyanidins and flavan-3-ols inhibit the secretion of cytokines by influencing NF-κB and MAPK/AP-1 activation (Table 1) [13,14,[18][19][20]22,26,28,38].
Many studies have shown inhibition of production and/or secretion of inflammatory cytokines by proanthocyanidins. Bodet et al. [42] demonstrated that proanthocyanidinenriched cranberry fraction at concentrations 25-50 µg/mL, significantly inhibited the IL-6, IL-8, and PGE 2 production by gingival fibroblasts stimulated with the Aggregatibacter actinomycetemcomitans lipopolysaccharide (LPS). The most spectacular inhibitory effect was seen towards IL-8 that belongs to chemokines (CXCL8) also known as neutrophil chemotactic factor. It directs the migration of polymorphonuclear leukocytes, monocytes, and macrophages to the infection site. Increased level of IL-8 was observed in the gingival crevicular fluid of inflamed periodontal sites [42]. PGE 2 is another proinflammatory molecule involved in destructive process in periodontal disease. It is secreted in response to pro-inflammatory cytokines, periodontopathogens and LPS. The cranberry fraction significantly inhibited PGE2-response even at low tested concentration-25 µg/mL, and reduced COX-2 protein expression-the enzyme involved in PGE2 production. Moreover, cranberry fraction influence on the phosphorylation and expression of various intracellular proteins (Jun, Fos, MKK3, MKK6, Rac1, and Mnk1) which are implicated in cytokine production. Bodet et al. concluded that cranberry fraction may act especially via a downregulation of AP-1 activity [42]. Feldman and Grenier [31] showed an inhibitory effect of 25 or 50 µg/mL of A-type cranberry proanthocyanidins (APA) on TNF-α, IL-6, and IL-8 secretion in a macrophage model. The 50 µg/mL concentration of APA reduced the LPSinduced secretion of TNF-α, IL-6 and IL-8, by about 50%, but had not influence on IL-1β. A significant reduction in IL-1β secretion was seen when ACPA was used together with licochalcone A (chalcone, not proanthocyanidin). Further studies on proanthocyanidins, in the predominant amount on A-type cranberry proanthocyanidins, prove their influence on the secretion and production of interleukins, as well as provided explanation of molecular mechanisms responsible for this activity [12,17,23,[26][27][28][29]37] (Table 1).
Galarraga-Vinueza et al. [12] revealed that cranberry concentrate from capsules (Uriach-Aquilea OTC, Barcelona, Spain) containing 130 mg A-type PAs significantly decreased M1 polarization and increased M2 polarization in LPS-stimulated macrophages. M1 phenotype of macrophage are activated by bacteria sub-products like lipopolysaccharide and are associated with the secretion of pro-inflammatory cytokines (such as IL-1β, IL-6, IL-8), whereas a M2 phenotype of macrophages are activated by alternative ways and are associated with the secretion of anti-inflammatory cytokines (such as IL-10) and growth factors which enhance tissue regeneration. Galarraga-Vinueza et al. [12] confirmed the effect of A-type PAs (50 and 100 µg/mL) on cytokine expression-proinflammatory cytokines: IL-8 and IL-6 were significantly downregulated in LPS-stimulated macrophages and A-type PAs, whereas an anti-inflammatory IL-10 was upregulated. No influence on expression of IL-1ß was seen. Lagha et al. [17] showed that fraction of proanthocyanidins (PAs) from cranberries at a concentration of 15.625-125 µg/mL markedly reduced cytotoxicity of leukotoxin on macrophages and significantly reduced (by about 80-90% at 15.625 and more than 98% at 125 µg/mL) release of caspase-1, IL-1β, and IL-18 from LtxA-induced macrophages. Leukotoxin (LtxA), released by A. actinomycetemcomitans is an important virulence factor playing a critical role in the pathogenic process of localized aggressive periodontitis (LAP). LtxA affects immune cells by activates pyroptosis of monocytes and macrophages and inducing the release of pro-inflammatory cytokines. Pyroptosis is the inflammatory form of programmed cell death, involves the activation of caspase-1, which in turn coverts of pro-IL-1β and pro-IL-18 to the biologically active forms. In macrophages, pyroptosis leads to the formation of pores in the plasma membrane which allows secretion of IL-1β and IL-18, cytokines known as damage-associated molecular patterns (DAMPs) and contribute to the progression of periodontitis by increasing cell migration and osteoclastogenesis [49,50]. Moreover, PAs reduced the expression of CIAS and P2X7 genes (increase by LtxA, in macrophages) by about 30-45%, similarly for a range 15.625-125 µg/mL [17]. This is important because the P2X7 receptor activation and CIAS activation leads to the rapid formation of membrane pores and to the release of IL-1β and IL-18. Moreover, the cranberry proanthocyanidins blocked the binding of LtxA to macrophages and reduced ROS and superoxide production in LtxA-induced macrophages.
Cranberry proanthocyanidins (PAs) can differently affect interleukins secretion/production, depending on a cell type. In lipopolysaccharide-stimulated normal human gingival fibroblast, cranberry non-dialyzable material (NDM) rich in proanthocyanidins decreased level of IL-6, what is consistent with other studies, but NDM significantly increased IL-6 in lipopolysaccharide-stimulated human gingival fibroblast cells from a patient suffering from aggressive form periodontitis (AgP fibroblasts) [28]. This increasing level of IL-6 occurred only in the presence of LPS; NDM alone did not show a significant increase in production of IL-6. Simultaneously, NDM inhibited NF-κB activity (increased by LPS treatment) in AgP fibroblasts hinting at the involvement of other molecular mechanisms of IL-6 regulation in these cells.
Influence of proanthocyanidins and flavan-3-ols from other source than cranberries on the secretion and production of interleukins was also demonstrated in several studies (Table 1). Jekabsone et al. [15] reported that Pelargonium sidoides root extract (PSRE) and especially proanthocyanidin fraction from PSRE (PACN) exhibits a strong, antibacterial properties (against Aggregatibacter actinomycetemcomitans), anti-inflammatory and gingival tissue protecting properties under periodontitis-mimicking conditions. The cells (gingival fibroblast, bone marrow-derived macrophages (BMDM) or human peripheral blood mononuclear cells (PBMCs) were stimulated using lipopolysaccharide (and IFNγ for BMDM) and treated with 50 µg/mL and 100 µg/mL of PSRE or PACN. The extracts protected human gingival fibroblast from A. actinomycetemcomitans infection, decreased lipopolysaccharide-induced release of IL-8 and prostaglandin E2 from gingival fibroblasts and IL-6 from leukocytes, blocked expression IL-1β, iNOS and COX-2 but not TNF-α.
Stronger anti-inflammatory activity of proanthocyanidin fraction (PACN) than root extract (PSRE) was associated with higher amounts of prodelphinidins. The study also reported that PSRE and PACN (100 µg/mL) blocked the surface presentation of CD80 and CD86 (surface markers of proinflammatory M1 phenotype) in LPS + IFNγ-treated macrophages, whereas PACN was characterized by stronger activity. These results indicate that both PACN and PSRE are potent in preventing macrophage conversion to proinflammatory M1 phenotype under exposure to LPS.
Low concentrations (7.9-62.5 µg/mL) extracts from a black and green tea as well as their flavan-3-ols (epigallocatechin-3-gallate, theaflavins) have influence on production and secretion proinflammatory cytokines. They reduce the epithelial gingival barrier dysfunction caused by TNF-α and modulate the hosts inflammatory response. They inhibited the activation of NF-κB and caspase-1 as well as reduced IL-1β secretion by macrophages (at 62.5 µg/mL by more than 94%, except black tea-64.5%), and secretion IL-8 (only black tea required higher concentration than 62.5 µg/mL for more than 70% inhibition) by human oral epithelial cells stimulated with recombinant TNF-α [16]. The green tea extract showed higher activity than black tea extract. Other studies have confirmed the inhibitory effect of tea derived flavan-3-ols on the secretion of pro-inflammatory cytokines from LPS stimulated macrophages as well from cytokines-stimulated gingival cells (Table 1) [14,23,24,36]. Ben Lagha et al. [20] presented consistent results in which they proved inhibitory effect theaflavins (TFs) from black tea on the secretion of pro-inflammatory cytokines from Porphyromonas gingivalis treated macrophages and on the activation of the NF-κB signaling pathway (Table 1). Lombardo Bedran et al. [24,25] in studies on green and black tea and their main galloylated flavan-3-ols revealed the ability of these compounds to induce human beta-defensin (hBD) secretion in gingival epithelial cells. HBDs are antimicrobial peptides secreted by gingival epithelium in response to periopathogens. HBDs interact with the bacterial cell membrane, and lead to pore formation and finally to the lysis of major periopathogens. Evidence indicated that level of hBDs is higher in healthy gingival tissues than in periodontal gingival tissues and that some periopathogens like P. gingivalis, are capable to down-regulate hBD expression by epithelial cells and/or to inactivate hBDs by the means of the proteolytic cleavage [25]. Both green and black teas and their galloylated flavan-3-ols stimulated secretion of hBDs and increased expression of the hBD gene in gingival epithelial cells as well as prevented the degradation of hBD1 and hBD2 by periopatogen P. gingivalis. Again, the tested non-galloylated flavan-3-ols-theaflavins failed to induce secretion of significant amounts of hBDs by the epithelial cells.
In addition to the inhibitory effect of EGCG on innate immune response (e.g., IL-1, IL-6, TNF-α), an influence on adaptive immunity mechanisms (Th1, Th2, Th17, and Tregs) was demonstrated. Hosokawa et al. [30] showed influence of EGCG on Th2-type chemokines, such as CCL11 production. The EGCG in range 3.125-50 µg/mL decreased CCL11 production in IL-1β/IL-4 or TNF-α/IL-4-stimulated human gingival fibroblasts (HGFs) in a concentration dependent manner (almost total reduction at 50 µg/mL). Moreover, they demonstrated that ERK and JNK activations, related to CCL11 production in HGFs, are inhibited by EGCG treatment. The same group demonstrated an inhibitory effect of EGCG and ECG on CXC chemokine ligand 10 (CXCL10 production (about 60% inhibition by 50 µg/mL) in human gingival fibroblasts (HGFs) stimulated oncostatin M (OSM)-cytokine belonging to the interleukin family [35]. CXCL10 is a Th1-type chemokine which plays a key role in the recruitment of Th1 cells, and thus in the development of periodontal disease. It is supposed that EGCG and ECG suppressed production of CXCL10 through the inhibition of phosphorylation of signal transduction molecules like JNK, Akt and STAT3 phosphorylation as well as by suppressed OSMRβ expression in stimulated HGFs [35]. Influence on adaptive immunity mechanisms was also proved for cranberry AC-PAs [37]. AC-PAs significantly decreased the secretion of C-C motif ligand 5 (CCL5) from P. gingivalis-stimulated oral epithelial cells (100 µg/mL of AC-PAs reduced secretion of CCL5 and also IL-8 by more than 80%). The CCL5 chemokine (has significant chemotactic activity for Th1 cells as well basophiles, eosinophiles, monocytes [37].
In addition to the above, there is a couple of other well studied sources of proanthocyanidins with proven anti-inflammatory activities linked with periodontitis.
Proanthocanydin-enriched extract from Myrothamnus flabellifolia, plants traditionally used for treatment of gingival inflammation and periodontitis in South Africa, decreased gene expression of IL-1β, IL-8 and TNF-α, and level of IL-6 in KB cells, pre-incubated with MF (10 µg/mL and 100 µg/mL) and infected with Porphyromonas gingivalis [33].

Influence on Reactive Oxygen Species (ROS)
ROS and reactive nitrogen species (RNS) production by immune cells stimulated by periopathogens is an important factor in pathogenesis of periodontitis [43]. Their overproduction can lead to oxidative damage to healthy gingival tissue, periodontal ligament, and alveolar bone. Study of Houde et al. [43] showed that the stimulation of macrophages with lipopolysaccharide from Aggregatibacter. actinomycetemcomitans and Fusobacterium nucleatum induces increased NO and ROS release. However, macrophages pretreated with non-cytotoxic concentrations of grape seed extract (GSE) containing 95% oligomeric proanthocyanidins significantly inhibited free radical generation by inhibiting the production of the proinflammatory mediators NO and ROS and by modulating iNOS protein expression. Significant decrease of ROS (e.g., superoxide) production by macrophages (exposed to LtxA) was also observed in the presence of cranberry PAs [17]. The addition of 125 µg/mL of PAs reduced ROS and superoxide production by 92.2% and 72.7%, respectively. These outcomes are supported by the results from animal studies [51], [52], further discussed below.

In Vivo Studies Reporting Influence Proanthocyanidins or Flavan-3-ols on Periodontitis in Animal Models
Among the reviewed studies, eleven reported the influence of proanthocyanidins or flavan-3-ols on periodontitis in animal models. Three of them used grape seed proanthocyanidin extract [52][53][54]; four used the flavan-3-ols [14,39,55,56]; two used green tea extract [57,58]; finally, one study used cranberry (Vaccinium macrocarpon) juice [29] and one unverified commercial proanthocyanidin (PA) [51]. Toker et al. [53] presented results which indicated that grape seed proanthocyanidin extract (GSPE) can substantially decrease periodontal tissue inflammatory process and alveolar bone loss by decreasing MMP-8 and hypoxia-inducible factor 1-alpha (HIF-1-α) levels and increase osteoblast activity in diabetic rats with experimentally induced periodontal disease (Table 2). Giving 100 mg/kg and 200 mg/kg doses of grape seed proanthocyanidin extract (GSPE) administered by oral gavage to rats with induced diabetes and periodontitis significantly decreased alveolar bone loss, inflammatory cell numbers, MMP-8 and HIF-1-α levels compared to rats with diabetes + periodontitis but without GSPE. Moreover, the osteoblast number increased significantly in the GSPE groups compared to the periodontitis and diabetes + periodontitis groups. Oral administration of commercial grape seed proanthocyanidins (PC) [52] to rats with experimentally induced periodontitis (EP) revealed that PC enhanced the host resistance and inhibited the oxidative stress. In serum, proanthocyanidins (PC) significantly decreased reactive oxygen species, lipid peroxides, lysosomal enzymes, acute phase proteins and they increased antioxidant levels. Histopathological evidence of experimentally induced periodontitis without PC showed cellular infiltration of inflammatory cells whereas the groups treated with proanthocyanidins demonstrated only scattered inflammatory cells. Similar anti-inflammatory effect of grape seed extract (GSE) was observed in Ozden et al. study [54].
Cai et al. [55] study indicated that EGCG alleviates P. gingivalis induced periodontitis in the animal model. The mice orally inoculated with P. gingivalis in PBS, received sterile food and drunk water with 0.02% epigallocatechin-3-gallate solution in the period from 8 weeks to 15 weeks. It was found that epigallocatechin-3-gallate has significantly reduced alveolar bone resorption as well as decreased the high expressions (caused by P. gingivalis infection) of inflammatory cytokines and other mediators both in serum and in gingival tissue (details in the Table 2) what is consistent with a previous study of Lee y et al. [39], in which epigallocatechin-3-gallate showed a suppressing effect one the progression of periodontitis, by diminishing Cyr61 expression (a potential osteolytic mediator) in osteoblast cells and, subsequently, macrophage chemotaxis into the lesions. Cho et al. [56] observed decreased IL-6 and TNF expression in the tissue of rats orally fed EGCG compared to the group without EGCG. Downregulation of TNF-α and IL-6 expression by EGCG led to a decrease of the number of osteoclast number as well as decrease in their activity, which finally has resulted lower bone loss. They also noticed reduced collagen destruction in EGCG group. Similar results achieved Lee at al. [14] studying catechin, another of major flavan-3-ols in green tea. They showed that catechin has reduced the level of bone loss in mouse animal model with a P. gingivalis induced periodontitis.     Addition of catechin-rich extract resulted in a significant decrease in gingival index (more marked for 0.8 mg/g than 0.4 mg/g) and decrease the odor from the mouth, but significant only for diet with 0.8 mg/g extract. The experimental diet also caused a significant decrease in the percentage of Porphyromonas sp.
(stronger for 0.8 mg/g than 0.4 mg/g). Isogai et al., 2008 [58] In turn, Polak et al. [29] showed that cranberry non-dialyzable material (NDM) consumption by mice infected by P. gingivalis and F. nucleatum has lowered the alveolar bone loss compared to the mice with infection but without NDM treatment. Moreover, in subcutaneous chamber model of inflammation, NDM alone has been shown to reduce TNF-α levels induced by the mixed infection. In vivo studies were supported by in vitro study (Table 1).

Clinical Studies
Until this moment, only three studies have been published pertaining to the use of proanthocyanidins or flavan-3-ols in periodontal disease in humans [59][60][61]. Two of them relate to the use of local delivery drug therapy with green tea extracts. Specifically, a thermo-reversible sustained-release system with incorporated green tea extract and hydroxypropylcellulose strips containing green tea catechin were used. Local therapeutic systems turned out to be effective in reducing periodontal pockets and inflammation [60,61] (details in Table 3). However, the weakness of Hirsawa et al. study [61] was the limited number of subjects in experimental group, as only 6 patients were treated. Díaz Sánchez et al. [59] were the only one to design study using pills rich in oligomeric proanthocyanidins. In this clinical study, 10 of 20 healthy volunteers with an induced gingivitis took the experimental treatment with oligomeric proanthocyanidins supplement administered orally as a dissoluble pill. According to Diaz Sánchez et al., the supplement caused improvement in the periodontal tissues condition during the period of treatment [59]. Although this study does not refer to periodontitis but to reversible gingivitis, the positive effect of the use oligomeric proanthocyanidins draws attention and encourages further clinical research. The Silness and Löe gingival index was higher in the control group than in the experimental group. The bleeding was lower in the experimental group compared to the control group. In contrast to the above results, the amount of dental plaque was slightly higher (33%) in the experimental group versus in the control group. No significant differences between the study group and the control group were seen in brightness of the gingiva. Statistically significant differences in level of IL-6 were found at the baseline between the experimental group and the control group and in the subsequent visits. However, in experimental group level of IL-6 was lower.

Conclusions
Among the numerous in vitro studies (36) on the immunomodulatory effect of proanthocyanidins or flavan-3-ols on the host cells, most concern the tea leaves extract and its compounds-catechins with presence of the galloyl moiety as the most active, as well as of A-type proanthocyanidins from fruits of Vaccinium species. Other sources of proanthocyanidins such as grape seeds and traditional medicinal plants, were seldom. The in vitro studies proved their immunomodulatory activity, among others by influencing on immune cell regulation, proinflammatory cytokines synthesis and gene expression as well as by radical scavenging and inhibition of certain enzymes. They modulate NF-κB (nuclear factor kappa-light-chain-enhancer of activated B cells) and mitogen-activated protein kinase (MAPK) pathways. Despite these promising results there is still much less studies using animal models (11) and only a few clinical studies (3). In conclusion, the potential of flavan-3-ols and their derivatives in prevention and alleviation of periodontitis is remarkable but clinical evidence is urgently needed for issuing credible dietary recommendation and complementary treatments.