Pistacia vera L. oleoresin and levofloxacin is a synergistic combination against resistant Helicobacter pylori strains

The increasing multidrug resistance in Helicobacter pylori, also correlated to its biofilm‐forming ability, underlines the need to search novel strategies to improve the eradication rate. Natural compounds are proposed as antibiotic-resistant-breakers capable to restore the efficacy of conventional drugs. Aim of this work was to evaluate the capability of Pistacia vera L. oleoresin (ORS) to synergize with levofloxacin (LVX) against resistant H. pylori strains. The antimicrobial activity of P. vera L. ORS and LVX and their combinations was determined by MIC/MBC (in neutral and acidic environments) and checkerboard tests. The anti-biofilm effect was determined by biomass quantification. In vivo Galleria mellonella model was used to confirm in vitro data. Pistacia vera L. ORS and LVX MICs ranged respectively from 780 to 3120 mg/l and from 0.12 to 2.00 mg/l, at pH 7.0 and 5.5. MBCs were similar to MICs. Pistacia vera L. ORS was able to synergize with LVX, restoring its effectiveness in LVX resistant strains. Pistacia vera L. ORS, LVX and their synergistic combinations displayed significant biofilm reduction. Pistacia vera L. ORS and LVX, showed protective effect against H. pylori infection on G. mellonella (62% and 63% of survival, respectively). Pistacia vera L. ORS can be considered a promising potentiator to restore the effectiveness of LVX tackling the H. pylori antibiotic resistance phenomenon.

Helicobacter pylori is a gastroduodenal pathogen that plays an important role in the pathogenesis of chronic gastritis, peptic ulcer, gastric adenocarcinoma and MALT (mucosa-associated lymphoid tissue) lymphoma. H. pylori infection is difficult to eradicate and requires the combination of different drugs such as clarithromycin, levofloxacin (LVX), amoxicillin, metronidazole, tetracycline and proton pump inhibitor [1][2][3] . The increase of antimicrobial resistance and the failure of therapeutic regimens, strongly underline the need to find novel strategies to enhance the eradication rate, also considering the capability of H. pylori to grow in biofilm mode [4][5][6] . Moreover, the H. pylori antimicrobial resistance profiles vary in different geographic areas, therefore the selection of therapeutic regimens needs to be adjusted according to local resistance patterns, if available [7][8][9][10] . For all these reasons, the search for alternative and effective new therapeutic schemes is important and urgent [11][12][13][14] .
New approaches to tackle the infections related to multidrug resistant (MDR) bacteria are recently proposed in the search for antibiotic-resistance-breakers capable to synergize with conventional drugs restoring their effectiveness [15][16][17] .
To this regards, much interest has been revived in the study of antimicrobial/antivirulence effects of formulations based on medicinal plants 15,18,19 . Several plants produce a variety of secondary metabolites such as phenolics, terpenoids and alkaloids, which possess a wide spectrum of biological activities, including the antibacterial ones 14,[19][20][21] . They interact with the lipidic bilayer of the cytoplasmic membrane, membrane proteins and enzymes involved in the synthesis of macromolecules, causing increased permeability, loss of proton-motive force and cellular material 14 .

Results
The neutral and acidic fractions of P. vera L. ORS were analyzed by Gas-chromatograph Mass Spectrometry (GS-MS). The neutral fraction contained a considerable amount of monoterpenes (27% both hydrocarbons and oxygenated) and a higher percentage of neutral triterpenes (59%). The acidic fraction had no monoterpenes and showed a higher percentage of acid triterpenes up to 69.7%. Mass spectra comparison allowed the identification of triterpenes, belonging to the 12-and 18-unsaturated oleanenes/ursenes, dammaranes and tirucallene derivative chemical families. The main compounds detected in P. vera L. ORS were hydroxydammarenone, tirucallol, isomasticadienoninc and masticadienoninc acids.
The antibacterial effect of P. vera L. ORS and LVX was evaluated against H. pylori strains to determine the susceptibility both at pH 7.0 and at 5.5 ( Table 1). The MIC values of P. vera L. ORS and LVX ranged from 780 to 3120 mg/l and from 0.12 to 2.00 mg/l, respectively, both in neutral and acid environments. In general, the MBC values of P. vera L. ORS, against H. pylori strains, were equal or one step above to the MIC values, except for H. pylori 9A/12 in which the MBC at pH 7.0 was two step above the MIC. The MBC values of LVX against H. pylori strains were always coincident to the MIC values, both at pH 7.0 and at 5.5, except for H. pylori 4A/13 (one step above). In Table 1 also are the best combinations of P. vera L. ORS and LVX with the values of FIC Index (FIC I) for each detected strain. Synergism was recorded in 31 of 32 strains (FIC I from 0.18 to 0.50) and an additive effect (FIC I = 0.75) was displayed in the reference H. pylori strain. Antagonism was not recorded. The MICs of LVX for all MDR and resistant strains (listed in Supplementary Table S1) were lower than the respective breakpoints when tested in combination with P. vera L. ORS with concentrations ranging from 0.03 to 0.25 mg/l ( Table 1). In particular, the best FIC I value was recorded for the resistant H. pylori 11F/11 with value of 0.18 with a MIC reduction of 4-and 2-fold for P. vera L. ORS and LVX, respectively. Figure 1 shows representative images of the checkerboard assays and the isobolograms combining different concentrations of P. vera L. ORS and LVX against two strains: the MDR H. pylori 2A/12 and the resistant H. pylori 11F/11. Pistacia vera L. ORS is able to restore the effectiveness of LVX in resistant H. pylori strains. The anti-biofilm effect of P. vera L. ORS and LVX alone ( Table 2) and their synergistic combinations (Table 3), were also evaluated for two MDR, two resistant and two susceptible H. pylori strains. Pistacia vera L. ORS and LVX at sub-MIC concentrations displayed reduction rates ranging from 8.43% ± 1.79 to 93.45% ± 4.34 and from 22.09% ± 2.12 to 91.98% ± 0.38, respectively (Table 2). Significant reductions of biofilm formation were obtained for all detected strains at all sub-MIC concentrations of P. vera L. ORS, except for H. pylori 3F/13 at 1/32 MIC. A similar trend was observed for LVX, in which a significant biofilm reduction was observed in all cases except for 1/32 MIC.
Pistacia vera L. ORS alone and combined with LVX displayed a significant H. pylori anti-biofilm activity. The Live/Dead assay was performed to evaluate the H. pylori viability showing also its capability to clusterize in presence of sub-MICs and sub-synergistic concentrations of P. vera L. ORS and LVX. A general reduction of H. pylori adhesion was observed in presence of all assayed sub-MIC concentrations of P. vera L. ORS and LVX as well as a marked loss of cell viability in presence of LVX. Figure 2 shows representative Live/Dead images of these effects in the resistant H. pylori 11F/11. In respect to the control ( Fig. 2A,B), P. vera L. ORS reduced the H. pylori aggregation with viable and prevalent coccoid bacteria (Fig. 2C), instead, a marked killing action together with anti-adhesive effect were detected in presence of LVX (Fig. 2D). This effect was visualized by few adhering cells on polystyrene surface organized in small clusters of red cells. A combined anti-adhesive and killing action was detected in presence of sub-synergistic concentrations of P. vera L. ORS and LVX (Fig. 2E,F).
The toxicity of P. vera L. ORS at different doses was evaluated with the in vivo G. mellonella model. The survival percentage of G. mellonella at 1000 mg/kg of P. vera L. ORS was 80% after 1 day and 60% after 5 days. At doses lower than 1000 mg/kg, analogous data was obtained with a survival rate of 80-100%, that was similar to those obtained with PBS injection; therefore P. vera L. ORS can be deemed not toxic ( Supplementary Fig. S1). During the experiments, G. mellonella control group was remained alive.
In the in vivo infection assay, the survival of G. mellonella larvae treated with H. pylori 11F/11 and LVX at MIC concentration, P. vera L. ORS at 1000 mg/kg, and the best combination of P. vera L. ORS and LVX (90 mg/l ORS + 0.12 mg/l LVX) was checked everyday until 5 days. The treatment with LVX rescued larvae injected with H. pylori with a survival rate between 90% and 100%. The treatment with P. vera L. ORS rescued larvae from H. pylori infection by 75% until 4 days and by 62% at 5 days. The best synergistic combination of P. vera L. ORS plus LVX showed a protective effect against H. pylori infection with larvae survival rate of 90% after 1 day and 63% after 5 days (Fig. 3A). The differences were compared with Long-rank test and the survival curves were statistically significant (p = 0.0008).
Pistacia vera L. ORS, both alone and combined with LVX, showed a protective effect against H. pylori infection over time.  www.nature.com/scientificreports www.nature.com/scientificreports/ The ability of H. pylori suspension of 1.8 × 10 6 CFUs to infect G. mellonella larvae was analyzed at different times, by CFU determination. As shown in Fig. 3B, H. pylori 11F/11 was able to infect G. mellonella larvae and to grow over time. In the presence of LVX, the high survival rate of G. mellonella larvae is related to the reduced H. pylori survival rate, demonstrated by the low CFU values detected after 1, 2, 3, 4 and 5 days (p ≤ 0.05). Pistacia vera L. ORS confirms its protective effect in G. mellonella larvae reducing significantly the bacterial load starting from 2 days (p ≤ 0.05). In presence of the best synergistic combination of P. vera L. ORS plus LVX, a significant H. pylori reduction was detected after 2 days (p ≤ 0.05) (Fig. 3B).
Pistacia vera L. ORS exhibits its anti-H. pylori effect also in in vivo model.

Discussion
In this study, the antimicrobial and anti-biofilm activities of P. vera L. ORS alone and combined with LVX against resistant clinical H. pylori strains were evaluated. Antibiotic resistance is presently a remarkable issue and, recently, a great number of studies suggest that the addition of medicinal plants to standard treatments, commonly used in therapy could emphasize the eradication rate of H. pylori 9,13,22,28 .
Our data shows an antimicrobial effect of P. vera L. ORS against resistant H. pylori strains both in neutral and acidic environments suggesting its potential inclusion in treatments of H. pylori related gastric infections. In particular, the real goal of this study is the evaluation of the combined action of P. vera L. ORS and LVX that synergize to each other restoring the antimicrobial drug efficacy. Pistacia vera L. ORS is able to restore the effectiveness of LVX by reducing its MIC values under the breakpoint against MDR and resistant H. pylori strains.
Safavi et al. 13 demonstrated that the antibacterial action of P. vera L. was correlated to the acid fraction containing terpenes. Another possible mechanism of antimicrobial activity of P. vera L. ORS could be related to its volatile components. In fact, ORS is characterized by a percentage of volatile monoterpenes of 15-30% which make it more fluid than the resins properly so-called. This ORS characteristic could favor the diffusion of antibiotics across bacterial membranes and/or hinder the efflux pumps that are a common resistance mechanism www.nature.com/scientificreports www.nature.com/scientificreports/ in Gram-negative bacteria 29 . Mechanisms of the pharmacological synergism are related to the multi-action in different target sites in the bacterial cell, pharmacokinetic or physicochemical effects (e.g. improvement of solubility or bioavailability) and/or action on bacterial resistance mechanism 30,31 . As well known, P. vera L. induces protrusions, morphological abnormalities and cellular fragmentation in H. pylori cells and probably facilitates the penetration and killing action of LVX 11,22 .
The anti-biofilm properties of P. vera L. ORS alone and in association with LVX were also evaluated on H. pylori biofilm formation. Pistacia vera L. ORS exhibits a strong anti-biofilm activity. The resin, a product of the Pistacia genus, contains a high molecular weight polymer cis-1,4-poly-β-myrcene, that shows antimicrobial and anti-adhesive actions 32 . Sharif et al. 33 showed that this natural compound exhibited both anti-plaque activity and H. pylori growth reduction. In our study, each sub-synergistic concentration displayed an important anti-biofilm effect. In particular, a stronger anti-adhesive effect was more evident when in the synergistic combinations P. vera L. ORS prevailed vs LVX, and a remarkable killing effect when LVX prevailed vs P. vera L. ORS.
Galleria mellonella is a validated in vivo model chosen for studies of pathogenesis of bacterial or fungal infections and it realizes the similar condition of disease development in the human body. Moreover, G. mellonella is used to study the toxicity and the efficacy of antibiotics and antifungal agents. Giannuoli et al. 34 showed that G. mellonella larvae are susceptible to H. pylori infection and it is a good in vivo model to evaluate the virulence factors and pathogenic mechanisms of H. pylori 34,35 . In some studies, the in vitro activity of natural products did not correlate with in vivo data due to the lack of specificity action of the molecule 25 . Here, we infected G. mellonella with the resistant H. pylori 11F/11 to evaluate the antimicrobial efficacy of P. vera L. ORS observing a protective effect against H. pylori infection, confirming the in vitro data. Levofloxacin, at MIC concentration, reduces significantly the H. pylori replication in G. mellonella larvae with a high percentage of survival larvae. Interestingly, P. vera L. ORS alone or combined with LVX shows a significant protective effect and this remarkable data suggests its potential use in H. pylori infection.  www.nature.com/scientificreports www.nature.com/scientificreports/ Pistacia vera L. ORS is a phyto-complex able to restore the effectiveness of LVX by reducing its MICs. The co-administration with the antibiotics commonly used in therapy could improve the potency of the treatment by repositioning certain marketed antimicrobial compounds. In this way, the use of natural compounds can overcome many of the commercial barriers regarding the developing of new antibiotics restoring the activity of ineffective drugs 16 .
Overall, our findings underline that the use of this natural compound, combined to LVX, could represents an effective and innovative strategy to tackle the antibiotic resistance and biofilm forming capability in H. pylori.
To the best of our knowledge, this is the first work that shows the effect of P. vera L. ORS against H. pylori strains with remarkable results relative to its antimicrobial and anti-biofilm effects. Future studies will be required

Materials and Methods
Pistacia vera L. ORS recovery and chemical fingerprinting. Pistacia vera L. raw ORS was obtained by making incisions from the base of the trunk of plants from a private pistachio garden in Bronte (Catania, Sicily, Italy) 36 . α-pinene, β-pinene, and alkanes standard solutions C 8 -C 20 were purchased from Sigma-Aldrich (USA), pure standards were purchased from Labochem (Italy), Chemfaces (China) and Extrasynthese (France). The raw ORS was filtered with 100 ml of diethyl ether on filter paper (Whatman, cat. n° 1004-930, grade 4). The volume was reduced gently under vacuum with a rotary evaporator until the filtrate was a yellow-amber crystalline solid. The ORS was fractioned to acidic and neutral fraction according to Barton and Seoane 37 . To improve its chromatographic behavior, as reported by Assimopoulou and Papageorgiou 38,39 , the acidic fraction was methylated under reflux with CH 3 I in presence of K 2 CO 3 to obtain the methylated fraction used for analytical purposes only. The identity of the ORS components was established from its GC retention indices, relative to C 8 -C 20 alkanes, by comparing their fragmentation patterns with those reported in the literature and by computer matching with the NIST MS 107, NIST 21 libraries, using the software GCMS solution version 1.02 (Lab solution, Shimatzu), by co-injection with authentic samples 40 .

Bacterial cultures.
For experiments, thirty resistant and MDR H. pylori clinical strains were used. All these strains were resistant to LVX (MICs ≥ 0.5 mg/l), MDR strains displayed the resistance at least three antimicrobial classes. The susceptible clinical strain H. pylori 9F/13 and the reference strain H. pylori ATCC 43629 were also included. The used strains together with their susceptible panel to antibiotics were listed in Supplementary  Table S1. The bacteria were cultured on chocolate agar containing Columbia agar base (CA, Oxoid, Milan, Italy) with 10% (v/v) lacked horse blood plus IsoVitalex 1% (v/v) (BBL, Microbiology System, Milan, Italy) and stored at −80 °C. The bacterial suspension were prepared in Brucella Broth (BB) plus fetal calf serum 2% (FS) (Biolife Italiana, Milan, Italy) and adjusted to an optical density at 600 nm (OD 600 ) of 0.2 (1.8 × 10 7 CFU/ml approximately), by using a Biophotometer (Eppendorf, Milan, Italy) 21 .
Antibacterial susceptibility assay. The determination of P. vera L. ORS and LVX MICs against standardized broth culture of H. pylori strains (prepared as describe above) was determined with broth microdilution method in 96-wells-microtitre plates 41   www.nature.com/scientificreports www.nature.com/scientificreports/ FS. Levofloxacin (Sigma Aldrich S.R.L, Milan, Italy) was prepared in BB plus FS in twofold dilutions from 2.00 to 0.01 mg/l. One hundred µl of P. vera L. ORS or 100 µl LVX and 100 μl standardized bacterial suspension were dispensed in each well of 96-wells-microtitre plate and incubated in microaerobic condition for 3 days at 37 °C. MIC values were measured by determining the lowest concentration of P. vera L. ORS and LVX able to inhibit the visible growth of the microorganisms. MBCs were determined by sub-culturing 10 μl of suspensions from the non-turbid wells on CA and incubated as describe above. The MBC represents the lowest concentration of P. vera L. ORS or LVX that inhibited the bacterial growth on plates. To evaluate a possible effect of pH variation on P. vera L. ORS, the MICs and MBCs were also determined in acid conditions by aseptically adding adequate amounts of 1N HCl to the medium to achieve a final pH value 5.5.
Checkerboard assay. Antibacterial synergisms between P. vera L. ORS and LVX was determined by checkerboard test. Dilutions of the two substances, from MICs to serial dilution below, were inoculated in microtiter plates and incubated as described above 15  Biofilm biomass quantification and cell viability analysis. Anti-biofilm activity of P. vera L. ORS, LVX (at sub-MICs) and all their synergistic concentrations 44 (FIC I ≤ 0.5, at sub-synergistic concentrations) were evaluated on the biofilm formation of two representative MDR (2A/12, 10A/13), two resistant (11F/11, 3F/13), one susceptible (9F/13) and the reference ATCC 43629 strains. Broth cultures of H. pylori, harvested in BB supplemented with 2% (w/v) SF (Biolife, Italy) and 0.3% (w/v) glucose, were gently shaken and incubated overnight at 37 °C in a microaerobic atmosphere. After incubation, each broth culture was adjusted to OD 600 = 0.1 and 100 µl of standardized broth culture was inoculated on flat-bottomed 96-wells-polystyrene-microtiter plates with P. vera L. ORS (100 µl), LVX (100 µl) or with sub-synergistic concentrations (50 μl of each substance). After incubation at 37 °C in microaerobic atmosphere for 48 hours, the produced biomasses of the treated and untreated biofilms were determined by safranin staining method 21 .
For the evaluation of cell viability, biofilms were grown in presence of sub-MIC concentrations of P. vera L. ORS, LVX and with all sub-synergistic combinations. Briefly, 1 ml of P. vera L. ORS or 1 ml of LVX or 1 ml (500 µl + 500 µl) of the all sub-synergistic combinations and 1 ml of standardized broth cultures of H. pylori 11F/11 were inoculated in Petri dish (3.5 cm) and incubated as describe above. After incubation, the planktonic cells were removed from each Petri dish and the sessile bacterial populations were washed with PBS and stained with Backlight Live/Dead Viability staining (Molecular Probes, Invitrogen detection technologies, USA) as indicated by manufacturer 45 . The images were observed at fluorescent Leica 4000 DM microscopy (Leica Microsystems, Milan, Italy), and more fields of view were examined randomly.
Pistacia vera L. ORS toxicity test. The toxicity was evaluated by using wax moth G. mellonella larvae.
Stock solution of P. vera L. ORS was diluted in PBS to obtain the following twofold final concentrations: 1000, 500, 250, 125 and 62.5 mg/kg. Seven groups of ten randomly-selected G. mellonella larvae, weighing 0.2-0.3 g, were treated as follows: five groups were injected in the last left proleg with 10 μl of each concentration of P. vera L. ORS and one group was injected with PBS by using Hamilton syringe, one group was uninjected. A total of seventy larvae were incubated at 37 °C in Petri dishes in dark for 5 days. The G. mellonella survival was evaluated every 24 hours; larvae were considered dead when were unresponsive to touch 46 . During assays, larvae did not receive nutrition.
In vivo G. mellonella infection assay. The in vivo activity of P. vera L. ORS against H. pylori was evaluated by using G. mellonella larvae, that represents a recognized model for H. pylori infection 34 and does not require ethical approval. H. pylori 11F/11, standardized at OD 600 = 0.25 (1.8 × 10 6 CFU/ml), was the clinical strain used for experiments. Five groups of ten randomly selected G. mellonella larvae were injected with 10 μl of H. pylori suspension, in the last left proleg of each larva for a total of fifty larvae. After two hours, ten larvae were treated with 10 μl of LVX at MIC value, ten larvae were treated with P. vera L. ORS at 1000 mg/kg, ten larvae were treated with the best synergistic combination of P. vera L. ORS plus LVX (90 mg/l ORS + 0.12 mg/l LVX) on the last right proleg. A control group of ten larvae was treated with 10 μl of PBS and ten larvae with a sham injection. For sham injection, the larvae were nicked with the syringe to evaluate the effect of the larval puncture. Larvae were incubated at 37 °C in Petri dishes in dark for 5 days. The G. mellonella survival was evaluated everyday; larvae were considered dead when unresponsive to touch.
To determine the survival rate of bacteria in larvae after 1, 2, 3, 4 and 5 days post-infection, fifty larvae were infected with 10 μl of H. pylori suspension (1.8 × 10 6 CFU/ml) and after two hours, ten larvae were treated with 10 μl of LVX at MIC values, ten larvae were treated with P. vera L. ORS at 1000 mg/kg , ten larvae were treated with the best synergistic combination of P. vera L. ORS and LVX, ten larvae was treated with PBS and ten larvae received a sham injection as describe above. After incubation, three larvae, for each group, were chilled on ice for 10 min, aseptically removed and the haemocoel was drained into a sterile 1.5 ml Eppendorf tube. Haemocoel was serially diluted in PBS and the bacterial cells were quantified by enumeration of CFUs on Campylobacter selective agar (CP Dent) with 7% defibrinated horse blood and 0.4% of Dent supplement (Oxoid) and incubated under microaerobic condition at 37 °C. The CFU/larva were counted after 3 days. During assays, larvae did not receive nutrition.