**1. Introduction**

Periodontitis is an infectious inflammatory disease resulting in periodontal pocket formation, progressive bone reduction and teeth loss in many industrialized countries [1,2]. Common treatment strategies include systemic use of antibiotics and local synthetic antiseptic substances, both leading to undesirable side e ffects and increased resistance of bacteria [3]. In consequence, prolonged and/or repeatable treatment is risky, ine fficient and fails to stop disease remission and further progression. In fact, as a response to the extensive use of drugs, bacteria have developed a new mechanism

to skip and counteract antibiotics activity: resistant polysaccharide envelope, more e fficient e fflux pumps, intracellular modifications and genetic mutations are some of the pathways exploited by bacteria to withstand drugs e ffect [4]. However, it is important to consider that not all body-resident bacteria are pathogens: commensal strain present in the microbiota play a pivotal role in preserving homeostasis in the skin and mucosal physiological systems of the human body [5,6]. The use of very strong chemicals such as chlorhexidine [7] can be exploited only for short periods to prevent severe side e ffects that can occur after prolonged exposure [8]. It follows that an ideal new antibacterial compound should be able to affect bacteria metabolism by a di fferent mechanism than those exploited by antibiotics but at the same time would be harmless to the healthy cells and commensal bacteria. In this light, multicomponent plant-derived antibacterial substances like proanthocyanidins (PACN) make a promising alternative and adjunctive therapy candidates for periodontitis treatment because of a lower risk of resistance development and side e ffects [9].

PACN are condensed tannins constructed form flavan-3-ol units [10]. The compounds possess a range of biological activities including anti-inflammatory and antibacterial [11]. The capacity of PACN to suppress inflammation is related to both strong antioxidant and metalloproteinase (MMP) inhibiting properties [12,13], whereas antibacterial e fficiency is achieved due to prevention of bacterial adhesion and biofilm formation [14]. The chemical nature of PACN in crude extracts varies depending on plant species used. *Pelargonium sidoides* DC, a medicinal plant native to South Africa, is one of the most PACN-enriched plants. Medicinal raw materials—roots of the plant—are used in the treatment of infectious and inflammatory disorders, and *P. sidoides* root extracts (PSREs) possess the same properties with enhanced e fficiency [15–18]. PSREs mediate their pharmacological e ffects via two classes of compounds, namely oxygenated coumarins and prodelphinidins that belong to the PACN group [18]. The common properties of these compounds isolated from various sources sugges<sup>t</sup> the significant part of the activities of PSREs might be assigned to PACN. Indeed, we have recently shown that namely prodelphinidin fraction from PSRE more e fficiently suppress periodontal pathogens *Porphyromonas gingivalis* compared to PSRE itself [19]. Moreover, the activity appeared to be strain selective: reducing the viability of the pathogens while preserving the metabolic activity of the beneficial oral commensal *Streptococcus salivarius*.

Based on these promising results, in the present study we decided to extend the examination of antibacterial e fficiency of PACN to other broad-range pathogens and commensals: two commercial drug-resistant *Staphylococcus aureus* and *Aggregatibacter actinomycetemcomitans* strains, a clinical isolate pathogen *Staphylococcus epidermidis* strain and a commensal *Escherichia coli* strain. Next, after verifying extract cytocompatibility towards gingival fibroblasts, a "race for the surface" model of bacteria-cells co-culture [20] was carried out to verify the extract ability to reduce bacteria proliferation while preserving cells viability in the same microenvironment where cells and bacteria compete for the same surface. Finally, we have made an extensive investigation on PACN activity in bacterial lipopolysaccharide (LPS)-mediated inflammation, including measurement of secretion of inflammatory cytokines and other mediators, inflammatory gene expression and viability of gingival fibroblasts, macrophages and blood leukocytes.

#### **2. Materials and Methods**

#### *2.1. Pelargonium sidoides Root Extract and Proanthocyanidin Fraction*

The *P. sidoides* root extract (PSRE) was purchased from Frutarom Switzerland Ltd. Rutiwisstrasse 7 CH-8820 Wadenswil (batch no. 0410100). Proanthocyanidins (PACN) from PSRE were purified as described by Hellström and co-authors [21] with some modifications [19]. Briefly, 4 g of PSRE was dissolved in 200 mL of 50% methanol, the solution was centrifuged at 2000× *g* for 20 min and filtered through 0.45 μm nylon filters. The solution was purified by gel adsorption over Sephadex LH-20. The proanthocyanidins were released from the gel with 70% aqueous acetone (500 mL) and concentrated

under vacuum at 35 ◦C. The aqueous aliquot was freeze-dried. The freeze dried PACN preparation yielded in 1.37 ± 0.07 g and comprised about 34.25% of the loaded PSRE.

#### *2.2. Bacterial Strains and Growth Conditions*

Commercially available strains *Staphylococcus aureus* (*S. aureus*, pathogen, ATCC 43300), *Aggregatibacter actinomycetemcomitans* (*A. actinomycetemcomitans*, pathogen, ATCC 33384) and *Escherichia coli* (*E. coli*, non-pathogen, ATCC BAA-1427) were purchased from the American Type Culture Collection (ATCC, MA, USA) and cultivated following the manufacturer's instructions. A clinical isolate of *Staphylococcus epidermidis* (*S. epidermidis*, pathogen) was collected at the Clinical Microbiology Unit at the Novara Maggiore Hospital (Novara, Italy). The clinical isolate was obtained after patient's informed consent in full accordance with the Declaration of Helsinki. Clinical strain was cultivated in Luria-Bertani medium (LB, Sigma-Aldrich, Milan, Italy) at 37 ◦C. For experiments, a single colony from each strain was collected and inoculated in 9 mL of LB broth at 37 ◦C overnight (18 h). After incubation, a new fresh LB tube diluted 1:10 was prepared and incubated at 37 ◦C for 3 h to achieve the logarithmic growth phase. Finally, broth cultures were diluted in LB broth until the optical density was 0.001 at 600 nm, corresponding to a final concentration of 1 × 10<sup>5</sup> cells/mL.

#### *2.3. Antibacterial E*ffi*ciency Evaluation*

To test antibacterial activity, PSRE and PACN were used at the following concentrations: 10, 30, 50, 70, 80, 90 and 100 μg/mL. PSRE and PACN powders were mixed and diluted directly into LB medium containing the desired bacteria concentration (described in 2.2. chapter); 1 mL of the obtained mix solutions (LB containing bacteria + PSRE/PACN) was seeded in the wells of a 24 multiwell plate (SPL, BioSigma, Milan, Italy) and incubated 24 h at 37 ◦C. Bacteria cultivated into pure LB medium were considered as control. After incubation, bacteria viability was evaluated by means of the colorimetric metabolic assay Alamar blue (alamarBlue ®, Thermo-Fisher, Waltham, MA, USA) following the manufacturer's instructions. Briefly, the ready-to-use solution was added to each well in a 1:10 ratio and incubated for 4 h in the dark at 37 ◦C; then, fluorescence was recorded at 590 nm using a spectrophotometer (Spark, Tecan, Basel, Switzerland).

#### *2.4. Co-Cultures of Human Gingival Fibroblasts and Bacteria*

To verify PSRE and PACN ability to preserve cells metabolism in the presence of infection, a race for the surface cells-bacteria co-culture experiment was set up. *S. aureus* and *A. actinomycetemcomitans* were selected and used with cells to simulate the oral and mucosal environments, respectively. Human primary gingival fibroblasts (HGF, ATCC PCS-201-018) were used as a cellular model to be tested with bacteria. Cells were seeded at a defined number (1.5 × 10<sup>4</sup> cells/well) in the wells of a 24 multiwell plate and allowed to adhere overnight. Afterwards, the medium was removed and replaced by 1 mL of solution composed by an antibiotics-free medium (minimal essential medium Eagle alpha-modification, from Sigma) supplemented with 10% fetal bovine serum (FBS, Sigma) and 100 μg/mL of either PSRE or PACN and 1 × 10<sup>5</sup> bacteria. The plate was incubated 24 h at 37 ◦C to allow cells-bacteria direct contact. Then, cells and bacteria were detached from plate wells by a collagenase (1 mg/mL) trypsin-EDTA (0.25%) solution and collected. The number of viable cells was determined by trypan blue and Burker chamber count. Cells cultivated with fresh medium without bacteria were considered as a control.

#### *2.5. Rat Gingival Fibroblast Cell Culture and Treatments*

All experimental procedures were performed according to the Law of the Republic of Lithuanian Animal Welfare and Protection (License of the State Food and Veterinary Service for working with laboratory animals No. G2-80). The mice were maintained and handled at Lithuanian University of Health Sciences animal house in agreemen<sup>t</sup> with the ARRIVE guidelines. Primary gingival fibroblasts were isolated from gingiva of P5-7 rat pups. After isolation, the cells were grown in 75 cm<sup>2</sup> flasks in DMEM with high glucose and Glutamax (Thermo Fisher Scientific, Waltham, MA, USA), 10% FBS and

Pen/Strep. At 70%–90% confluence, the cells were detached by 0.025% Trypsin/EGTA and plated in 96 well plates at a density of 2 × 10<sup>5</sup> cells/well. The treatments were made 24 h after plating. All treatments were made simultaneously, without pre-incubations, and lasted 24 h. LPS was used at concentration of 2 mg/mL (1 mg/mL LPS did not induce significant increase in cell death), filtered PSRE and PACN solutions at 50 and 100 μg/mL (both preparations induced toxicity starting from 200 μg/mL).

#### *2.6. Bone Marrow-Derived Macrophages*

For bone marrow-derived macrophages (BMDM) isolation male C57BL6/J inbred mice (18–20 weeks old, Envigo, Netherlands) were used. The experimental procedures were carried out in accordance with the guidelines of the European Community (2010/63/EU), local laws and policies and were approved by the Latvian Animal Protection Ethical Committee, Food and Veterinary Service, Riga, Latvia. Mice were euthanized by decapitation, and bone marrow cells were extracted from femur bones and differentiated for 7 days in RPMI-1640 with Glutamax (Gibco,) supplemented with 10% FBS, 1% antibiotics and 10 ng/mL M-CSF (monocyte-colony stimulating factor, PeproTech, London, UK). Then cells were detached by 0.5% trypsin (Sigma Aldrich), and plated in a 12-well plate (11 × 10<sup>5</sup> cells/mL) in DMEM-high glucose medium supplemented with 10% FBS, 1% antibiotics. After 1h incubation in 37 ◦C incubator, cells were stimulated with PSRE and PACN at 100 μg/mL and LPS 10 ng/mL with murine IFNγ (interferon gamma, PeproTech) 100 U/mL for proinflammatory gene expression and macrophage polarization to M1 (pro-inflammatory) phenotype for 2 h and 24 h, respectively.

#### *2.7. Human Peripheral Blood Mononuclear Cells*

Human peripheral blood mononuclear cells (PBMCs) were purchased from ATCC (ATCC ® PCS-800-011 ™, Manassas, VA, USA). The cells were cultured at 3.3 × 10<sup>6</sup> cells/mL (12 well plate) in RPMI medium supplemented with 10% FBS, 1% antibiotics. After 1 h incubation in 37 ◦C incubator, cells were stimulated with 1 μg/mL LPS (lipopolysaccharide, Sigma-Aldrich) in the presence of 100 μg/mL PSRE or PACN, for 6 h.

#### *2.8. Analysis of Cell Viability by Lactate Dehydrogenase Release, Alamarblue and MTT Assay*

PBMCs' viability was assessed by measuring lactate dehydrogenase (LDH) release in cell culture media. LDH activity was measured using a method based on the reduction of a tetrazolium salt (yellow) to formazan (red) [22]. The absorbance of kinetic parameters was determined spectrophotometrically at 503 nm on Hidex Sense microplate reader. The reaction mixture contained 30 mM lactate, 150 μM NAD+ (nicotinamide adenine dinucleotide), 0.4 mM 2-(4-iodophenyl)3-(4-nitrophenyl)-5-phenyltetrazolium chloride (INT) and 3.25 mM *N*-methylphenazonium methyl sulphate (PMS) in 100 mM Tris bu ffer solution (pH 8.0). In addition, PBMCs viability after 24 h incubation with di fferent concentrations of PSRE and PACN was determined with alamarBlue ®, (Bio-Rad Laboratories, Hercules, CA, USA) following the manufacturer's instructions. Briefly, the ready-to-use solution was added to each well in a 1:10 ratio and incubated for 2 h in the dark at 37 ◦C; then, fluorescence (Ex 544 nm/Em 590 nm) and optical density (570 and 600 nm) using Hidex Sense microplate reader.

BMDM viability after 24 incubation with different concentrations of PSRE and PACN was determined using MTT assay. After incubation, BMDM were incubated with MTT (TCI Europe) solution (1 mg/mL) for 1 h, the formazan crystals formed during incubation were dissolved in isopropanol, and optical density at 570 nm corresponding to the amount of viable cells was measured in a Hidex Sense microplate reader.

#### *2.9. Necrosis Evaluation by Double Nuclear Staining*

The level of necrotic cell death was assessed by double nuclear fluorescent staining with Hoechst33342 (10 mg/mL) and propidium iodide (PI, 5 mg/mL), 5 min at 37 ◦C. PI-positive nuclei indicating lost nuclear membrane integrity were considered necrotic. Cells were visualized under fluorescent microscope OLYMPUS IX71S1F-3 (Olympus Corporation, Tokyo, Japan), counted in fluorescent micrographs and expressed as the percentage of total cell number per image. The data are

presented as averages ± standard deviation. LD50 was calculated by SigmaPlot v.13 (Systat Software Inc, London, UK) using the equation proposed by a dynamic curve fitting tool.

#### *2.10. Apoptosis Evaluation by Annexin V staining*

The viable cells were analyzed on a BD FACS Melody ™ (BD Biosciences, San Jose, CA, USA) flow cytometer using Annexin V-allophycocyanin (BioLegend, San Diego, CA, USA) staining. For analysis of BMDMs apoptosis, treated cells were stained with fluorescent annexin V antibody and, afterwards, the proportion of apoptotic (Annexin V positive) cells was evaluated.

#### *2.11. Caspase Activity Assessment*

Caspase-3 activity in cell lysates after treatments was measured by means of a Caspase 3 Assay Kit (Sigma-Aldrich) according to the manufacturer's protocol, by assessing Ac-DEVD-7-amido-4- methylcoumarin cleavage and subsequent increase in 7-amido-4-methylcoumarin (AMC) fluorescence in a fluorometric plate reader Ascent Fluoroskan (Thermo Fisher Scientific, Waltham, MA, USA; λex = 360 nm, λem = 460 nm). The data are presented as averages ± standard deviation of AMC concentration increase rate normalized for mg of cellular protein, averages ± standard deviation.

Caspase-8 in cell lysates after treatments was measured by means of a colorimetric Caspase 8 Assay Kit (Sigma-Aldrich) according to the manufacturer's protocol, by assessing Ac-IETD-*p*-Nitroaniline cleavage and increase in *p*-Nitroaniline (pNA) concentration in a spectrophotometric plate reader Multiskan Go 1510 (Thermo Fisher Scientific, Waltham, MA, USA) reading absorbance at λex = 405 nm. The data are presented as averages ± standard deviation of pNA concentration increase rate normalized for mg of cellular protein, averages ± standard deviation.

#### *2.12. Detection of Secreted Inflammatory Mediators*

Medium collected after treatments was assayed for cytokines tumor necrosis factorα (TNFα), interleukin-6 (IL-6), interleukin-8 (IL-8) and prostaglandin E2 (PGE2) production using TNFα mouse (Millipore), IL-6 human (Sabbiotech), IL-8 rat (Abbexa) and PGE2 rat (Abbexa) kits following the manufacturer's protocols.

#### *2.13. Bone Marrow-Derived Macrophage Polarisation to M1 Phenotype and Analysis by Flow Cytometry*

BMDMs were incubated with PSRE and PACN (100 μg/mL each) and LPS+IFNγ (10 ng/mL/100 U/mL) for 24 h. The cells were washed twice with HBSS and harvested by trypsin (0.5%), then DMEM-high glucose medium with 10% FBS was added and cell suspension was centrifuged at 300× *g* for 5 min. Then cells were incubated with specific conjugated antibody mixtures (in concentration 1:100 in cell wash buffer) for 30 min on ice in the dark. The mixture contained following monoclonal antibodies purchased from BioLegend (San Diego, CA, USA): FITC-conjugated anti-mouse F4/80, phycoerythrin (PE)-conjugated anti-mouse CD86 and biotin-conjugated anti-mouse CD80. Then cells were washed and stained with streptavidin-APC-Cy7. After 30 min, cells were washed and stained for evaluation for apoptosis (see method above). After staining samples were analyzed by flow cytometry (BD FACS Melody ™, BD Biosciences, San Jose, CA, USA).

#### *2.14. mRNA Isolation and Quantitative RT-PCR Analysis*

Total RNA from cells was isolated using Ambion PureLink RNA Mini Kit (catalogue No. 12183025) according to the manufacturer's protocol. The first-strand cDNA synthesis was carried out using a High Capacity cDNA Reverse Transcription Kit (Applied BiosystemsTM, Foster City, CA, USA) following the manufacturer's instructions. The quantitative RT-PCR analysis of gene expression was performed by mixing SYBR Green Master Mix (Applied BiosystemsTM), synthesized cDNA, forward and reverse primers specific for interleukin-1β (IL-1β), interleukin 10 (IL-10), inducible nitric oxide synthase (iNOS), TNFα, cyclooxygenase 2 (COX-2) and running the reactions on a Mic Real-Time PCR instrument. The relative expression levels for each gene were calculated with the ΔΔCt method and normalized to the expression of glucose-6-phosphate isomerase gene.

## *2.15. Statistical Analysis*

The quantitative results are presented as mean ± standard deviation (SD) of 3–7 replicates. The data were processed using Microsoft Office Excel 2010 (Microsoft, Redmond, WA, USA) and SPSS 20 (IBM, Armonk, NY, USA) software. For antibacterial activity testing and bacterial-fibroblast co-culture experiments, the statistical data analysis was performed by applying the ANOVA with a Tukey HSD post hoc test. For experiments on rat gingival fibroblasts, ANOVA with Dunn's test was used and data analyzed by SigmaPlot v.13 (Systat Software Inc., London, UK). In all cases, differences were considered statistically significant when *p* < 0.05.
