Synergistic effects of Cinnamomum cassia L. essential oil in combination with polymyxin B against carbapenemase-producing Klebsiella pneumoniae and Serratia marcescens

Multidrug resistance prompts the search for new sources of antibiotics with new targets at bacteria cell. To investigate the antibacterial activity of Cinnamomum cassia L. essential oil (CCeo) alone and in combination with antibiotics against carbapenemase-producing Klebsiella pneumoniae and Serratia marcescens. The antimicrobial susceptibility of the strains was determined by Vitek® 2 and confirmed by MALDI-TOF/TOF. The antibacterial activity of CCeo and its synergism with antibiotics was determined using agar disk diffusion, broth microdilution, time-kill, and checkboard methods. The integrity of the bacterial cell membrane in S. marcescens was monitored by protein leakage assay. CCeo exhibited inhibitory effects with MIC = 281.25 μg.mL-1. The association between CCeo and polymyxin B showed a decrease in terms of viable cell counts on survival curves over time after a 4 hour-treatment with a FIC index value of 0.006. Protein leakage was observed with increasing concentrations for CCeo and CCeo + polymyxin B treatments. CCeo showed antibacterial activity against the studied strains. When associated with polymyxin B, a synergistic effect was able to inhibit bacterial growth rapidly and consistently, making it a potential candidate for the development of an alternative treatment and drug delivery system for carbapenemase-producing strains.


Introduction
Klebsiella pneumoniae and Serratia marcescens are able to acquire and share their resistance genes, leading to multidrug resistance, which can result in severe infections. One of the most important therapeutic challenges posed by Enterobacterales is their resistance to carbapenems [1]. Carbapenemase encoding-genes are located on self-conjugative plasmids, capable of disseminating among bacteria, resulting in the spread of resistance [2]. Clinical options for hard- Samples were taken from patients over 50 years old, hospitalized in different wards at a tertiary hospital in Brazil's Midwestern region and are representative of the carbapenemase-producing Gram-negative bacteria group. Bacterial strains evaluated in this study were representative from predominant clonal type (33 KPC-producing K. pneumoniae and 6 KPC-producing S. marcescens), previously characterized [14,15]. Selected isolates were maintained at -70˚C, then sub-cultured on MacConkey agar for 24 h before testing. These isolates were identified and characterized according to their sensibility to antibiotics by Vitek 1 2 (bioMérieux, Hazelwood, MO, USA). Their minimal inhibitory concentrations (MIC) were confirmed by broth microdilution. Preliminary screening for the presence of carbapenemase was performed using the modified Hodge test [16] and confirmed by matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF/ TOF) using a Microflex LT spectrometer (Bruker Daltonics, Billerica, MA, USA) [17].

Detection of resistance genes
Polymerase chain reaction (PCR) was performed to confirm the presence of the genes bla CTX-M-1-like , bla CTX-M-2-like , bla CTX-M-8-like , bla CTX-M-14-like , bla GES-like , bla GIM-like , bla IMP-10 , bla IMP-like , bla KPC-2 , bla NDM-like , bla OXA-23 , bla OXA-48-like , bla SHV-like , bla SIM-like , bla SME-like , bla SPM-like , bla TEM-like , and bla VIM-like in KP-KPC and SM-KPC strains. Total DNA of KP-KPC and SM-KPC isolates was extracted using the QIAamp DNA minikit 1 (Qiagen, Courtaboeuf, France) and used as the template in PCR experiments. Amplification was performed using specific primers and cycling parameters, according to the target sequence. Genomic DNA was quantified using a NanoDrop TM 1000 spectrophotometer (Thermo Fischer Scientific, Waltham, Massachusetts, MA, USA). Polymerase chain reactions (PCRs) were prepared as standard 25 μL volumes comprising 12.5 μL of 2 x Master mix [0.05 U/μL Taq DNA polymerase, reaction buffer, 4 mM MgCl 2 , 0.4 mM of each dATP, dCTP, dGTP and dTTp], 0.25 μL of each primer (1 μL), 1 μL template DNA (20−30 ng/μL) and 11 μL nuclease free water. PCR was performed using a Thermal cycler (model C1000 Touch) supplied by BIO-RAD, California, USA. Sequences of the primers used to the amplification of bla genes are shown in Table 1. The PCR products were analyzed in horizontal electrophoresis using a 1% agarose gel and a 50-bp DNA ladder (Ludwig, Brazil) as a molecular marker, then PCR products were visualized by UV light at 336 nm.

Antibacterial assays of CCeo
The antibacterial activity of raw CCeo was determined using the standard method of agar disk diffusion. Polymyxin B and tigecycline were used as the positive standards. Filter paper disks of 6 mm diameter, containing 10 μl of the CCeo were evaluated against KP-KPC and SM-KPC. After 30 min at room temperature, the dishes were incubated at 37 ± 1˚C, for 24 h and the diameter of the inhibition zone formed was measured in millimeters [24]. The test was performed in duplicate and the means of the values obtained were used to classify the bacterium as sensitive (� 10 mm) or resistant (< 10 mm) [25]. The MIC of CCeo and trans-cinnamaldehyde (TC) were determined by broth microdilution, as described by Cavalcanti et al. [26] and was determined by resazurin colorimetric assay. CCeo was diluted with distilled water and 0.5% Tween-80.

Synergy tests
The agar disk diffusion assay was performed using the Kirby-Bauer methods for synergy screening. Here, CCeo was evaluated alone and in combination several antibiotics (amikacin, ciprofloxacin, imipenem (IPM), piperacillin/tazobactam, polymyxin B (POL), and tigecycline (Sigma-Aldrich). Antibiotic disks were placed on a Müller-Hinton plate inoculated with KP-KPC or SM-KPC and 10 μl of the CCeo were added in the antibiotic disk [24]. After incubation, the inhibition zone of the combination was comparatively examined with the independent test samples. Synergistic interactions were evaluated if the inhibition zone in the combination was larger than 2 mm or more in comparison to either antibiotics or CCeo alone [27]. The test was performed in duplicate.

Time-kill
The time-kill method was performed using the broth macrodilution technique [28]. CCeo alone and in combination with antibiotic (imipenem or polymyxin B) was tested against the , which was the same as the polymyxin B MIC for KP-KPC. The time-kill studies were performed with approximately 1.5 × 10 6 colony forming units (CFU).mL -1 in a final volume of 3.2 mL, verified with a spectrophotometer by Vitek 1 2 (bioMérieux). The tubes were shaken periodically and incubated at 37˚C. Samples were obtained after 0, 4,8,12,16,20, and 24 h of incubation. At each sample time, 1 μl was withdrawn from each tube using a sterile loop and seeded in MacConkey agar plates. The plates were incubated for 24 h at 37˚C and the colony counts were determined. The resulting colony count of the samples treated with the combination was compared to the samples treated with the most effective single agent. A decrease in colony count of 100-fold or more was considered synergism, while an increase in colony count of 100-fold or more was considered antagonism. Changes that were limited to 10-fold were defined as additive or indifferent [29]. A negative control of BHI was used for the bacterial strain, and saline was used as a sterility control. Polymyxin B and gentamicin (GEN) were used as the positive controls for KP-KPC and SM-KPC strains, respectively. The test was performed in triplicate.

Checkboard
To evaluate the potentiating effect of CCeo, combinations of CCeo and polymyxin B were assessed against the KP-KPC and SM-KPC strains. The concentration of each antimicrobial agent ranged from 1/32 to 2 × MIC. The polymyxin B concentration used as MIC for SM-KPC was 1 μg.mL -1 . Using a microtiter plate, two-fold serial dilutions of the oils were prepared in the horizontal rows and two-fold serial dilutions of the antibiotics were prepared in the vertical rows. The plates were prepared well-by-well to obtain a single plate in which both antimicrobial agents were cross-diluted. The inoculum used was 1.5 × 10 6 CFU.mL -1 . Resazurin was used to indicate viable bacteria. The fractional inhibitory concentration (FIC) of each combination was then calculated as the ratio of MIC of the antimicrobial agent in combination versus the MIC of the antimicrobial agent alone [29]. The SFIC is then calculated for each test sample independently.
The FIC index was calculated as follows: SFIC = FIC ( � i) + FIC ( � ii) . The results were interpreted as either synergistic (SFIC � 0.5), additive (0.5 < SFIC < 1.0), noninteractive (1.0 < SFIC � 4.0), or antagonistic effects (SFIC > 4.0) [29]. All experiments were carried out in duplicate as two independent experiments and the calculated FIC indexes were averaged. A negative control of BHI was used for the bacterial strains and saline was used as a sterility control. The sterility of CCeo, polymyxin B, and water was assessed. A second negative control, in the form of 0.5% Tween-80, was also used.

Determination of cell membrane integrity
The integrity of the bacterial cell membrane was monitored by the release of proteins from the cell into the supernatant. Carbapenemase-producing SM-KPC was incubated in MacConkey agar at 37˚C for 24 h. Three groups with bacterial inoculum in the logarithmic growth phase (1.5 × 10 6 CFU.mL -1 ) were placed in a 96-well microplate, treated with the combination CCeo + polymyxin B, with 0.5, 1, and 2 × the MIC of each agent. The microplate was incubated at 37˚C for 0, 1, 2, and 4 h. Then, 1 μL aliquot of CCeo + polymyxin B treatment (concentration of MIC) were seeded in Müller-Hinton agar to confirm the viability of the strain at each timepoint. The content of each microplate well was then centrifuged at 2,500 rpm for 5 min at 4˚C. The concentration of the proteins released from the cytoplasm was then assessed in the supernatant using the Pierce TM BCA Protein Assay kit. The optical absorbance (OD) was read at 490 nm using the iMark TM Microplate Absorbance Reader (Bio-Rad, São Paulo, SP, Brazil). The experiment was performed in duplicate. Saline with bacterium was used as negative control. CCeo and polymyxin B were diluted with distilled water.

Statistical analysis
The experimental data derived from time-kill and determination of cell membrane integrity assays were analyzed using linear regression slopes comparison-GraphPad Prism 6.01 (GraphPad Inc., San Diego, CA, USA). For time kill, log CFU values were plotted against time for each antibiotic. The kill rate was determined at different time intervals (0, 4, 8, 12, 16, 20 and 24 h), undertaking a linear regression to find the slope for each antimicrobial to test the difference between the slopes and intercepts. For determination of cell membrane integrity assay, concentration of protein was plotted against antimicrobial agent's concentration. P value < 0.05 was considered significant.

Molecular analysis of 16S and arnB genes in SM-KPC strain
Reverse transcriptase quantitative polymerase chain reaction (RT-qPCR) analysis was performed to measure the changes in the expression levels of the small subunit 16S of ribosomal RNA during treatment with CCeo in association with polymyxin B. Total RNA was isolated from SM-KPC before and after treatment (at MIC values), transformed into cDNA using the iScript TM cDNA Synthesis kit (Bio-Rad). The cDNA was then subjected to RT-qPCR using the primers pairs 16S forward (16SF) and 16S reverse (16SR) ( Table 2) and SYBR 1 Green Jump-Start TM Taq ReadyMix TM (Sigma-Aldrich). The reactions were performed using the following reagents: SYBR Green mix, 16SF primer, and 16SR primer (10 μM), and template (20 ng), using double-distilled water to obtain the reaction final volume. RT-qPCR cycling was carried out using the CFX96 Touch TM Real-Time PCR Detection System (Bio-Rad) as follows: initial denaturation at 94˚C for 2 min, then 40 cycles of 94˚C (15 sec) and 60˚C (1 min). All PCR assay reactions were performed in triplicate. Negative controls were included with each PCR assay. Following amplification, the results were analysed on the PCR computer using the CFX Manager ver. 3.1.15 software. Once amplification was completed, the level of amplification was reported by the software as the mean cycle threshold (CT) value of the replicate samples. A sample was considered positive by real-time RT-qPCR and expressed the 16S gene if the CT value was above the threshold value before the 40 cycles of the PCR reaction and below the negative control CT.
Reverse transcriptase polymerase chain reaction (RT-PCR) was also performed on the arnB genes in the SM-KPC strain before and after treatment with CCeo + polymyxin B or polymyxin B alone [14,15]. The primers arnBF and arnBR were designed based on the gene sequences of S. marcescens subsp. marcescens Db11 and S. marcescens 2880STDY5682918 (GenBank Accession No. HG326223.1, NZ_FCJC01000002.1) ( Table 2). Total RNA was isolated and transformed into cDNA using iScriptTM cDNA Synthesis kit (Bio-Rad) and assessed by RT-PCR (Lin2014). The PCR products were examined by 1% agarose gel electrophoresis and ethidium bromide staining.

Ethical approval
This study was conducted with the approval of the Research Ethics Committee from the Universidade Federal da Grande Dourados (Process No. 877.292/2014).

Results
The isolates showed resistance to carbapenems. Sensitivity was observed only to polymyxin B for KP-KPC and amikacin and gentamicin for SM-KPC ( Table 3 Table 4. The survival curves of the strains showed a decrease in viable cell counts over time when CCeo was in association with polymyxin B (Fig 1). CCeo + polymyxin B promoted decreases in cellular count in a range of 5 log 10 CFU.mL -1 . Considering the time of cell death, CCeo inhibited KP-KPC and SM-KPC strains within 24 h. Moreover, CCeo + polymyxin B reduced the bacterial load to undetectable levels within only four hours of treatment, which was faster than polymyxin B (20 h) (p < 0.05) and gentamicin (8 h) (p < 0.05) for KP-KPC and SM-KPC, respectively. For KP-KPC, CCeo + polymyxin B time-kill curve presented statistical difference against the negative control (p < 0.05) but not from POL curve (p > 0.05) showing that the synergy CCeo + polymyxin B is as effective as polymyxin B, an used antibiotic against this carbapenemase-producing strain. For SM-KPC, the CCeo + polymyxin B time-kill curve was significantly different from the polymyxin B curve (p < 0.05). The combination of CCeo and imipenem was unable to reduce the bacterial load (p > 0.05). For all treatments, imipenem had no activity within 24 h. The control antibiotics (polymyxin B and gentamicin) successfully inhibited both strains before 24 h. Saline was used as negative control, for which no growth was observed. The effect of the addition of Tween-80 (0.5%) as an oil solubilizer to the antibacterial assays did not interfere with bacterial growth.
In the checkboard assay, 49 different combinations of CCeo and polymyxin B were tested for each strain, ranging from twice the MIC to several dilutions below the MIC. The FIC index values were calculated by considering the lowest combinations of CCeo and polymyxin B in which there was no visible growth. The FIC index values were 0.006 for both KP-KPC and SM-KPC. A 320-fold reduction of the polymyxin B MIC was able to inhibit the growth of the strains (Fig 2). Checkboard assay was not performed using imipenem because the time-kill curves did not prove synergism.
When SM-KPC was treated with increasing concentrations of CCeo and CCeo + polymyxin B, the cells externalized the proteins. The protein leakage from the bacteria in the negative control at 1 h resulted in 12.0 μg.mL -1 . At a sub-inhibitory concentration (0.5 × MIC), 33.8 μg.mL -1 of protein was present in the supernatant of CCeo treatment and 36.2 μg.mL -1 for CCeo + polymyxin B, both at 1 h. With CCeo treatment (2 × MIC), 80.6 μg.mL -1 of protein was present in the supernatant. With CCeo + polymyxin B (1 h), 96.8 μg.mL -1 of protein was present in the supernatant, showing an increase in the protein leakage levels. The concentrations of externalized proteins when cells were treated with CCeo and CCeo + polymyxin B differed significantly from cells treated only with polymyxin B (p < 0.05). Polymyxin B alone did not show any leakage of protein (Fig 3). In addition, no difference was observed in the protein leakage Table 4

Method
Antimicrobial agent Bacteria between the time-points studied (1, 2 and 4h). Protein leakage after 1 h in all treatments was the time-point chosen to be represented graphically because the bacterial cells are still viable to grow in the culture media at this time.
When the cells were treated with CCeo + polymyxin B, Müller-Hinton agar plates showed abundant growth at 0 h, an expressive growth reduction at 1 h, only 2 CFU at 2 h, and a total inhibition at 4 h. No difference in bacterium growth at 0, 1, 2, and 4 h was observed when the cells were treated with CCeo alone (Fig 4), consistent with the time-kill assay results.
The 16S rRNA transcripts were reduced in SM-KPC cells after treatment CCeo + polymyxin B, as verified by RT-qPCR, confirming a decrease in the bacterial cell number (S2 Fig). The

Discussion
As far as we know, there are no studies on the antibacterial action of C. cassia alone or in combination with antibiotics against carbapenemase-producing Enterobacterales bacteria. Here, C. cassia was able to inhibit two strains of Gram-negative bacteria resistant to almost all classes of clinical antibiotics, either alone or in combination with polymyxin B. Antibacterial assays demonstrated the potential of action of CCeo against KP-KPC and SM-KPC (MIC of 281.25 μg.mL -1 ). Trans-cinnamaldehyde is the main component with antibacterial action of CCeo, and showed the same MIC as CCeo. Following a new tendency to study the whole plant instead of single compounds [29], assays were conducted with CCeo as a whole. CCeo + polymyxin B was found to be faster than polymyxin B or gentamicin in inhibiting KP-KPC and SM-KPC. This susceptibility was independent of the acquired resistant gene, since KP-KPC expressed bla KPC-2 , and SM-KPC expressed bla KPC-2 and bla IMP-10 , indicating a mechanism of action not related to the genus, species, or resistance genes. The FIC index value of CCeo and polymyxin B in combination confirmed the synergistic effects between these two agents for KP-KPC and SM-KPC. This indicated that the best dosages of the antimicrobial agents where obtained when these compounds were used in combination, providing a basis for the use of different combinations in the development of alternative treatments.  Imipenem and polymyxin B were chosen for synergism confirmatory tests because they are used clinically for the therapeutic treatment of Gram-negative infections. The doses of CCeo and antibiotics in this study were selected according to the MIC values presented by the strains studied. For carbapenem-resistant bacteria, polymyxin B alone or in combination with other antibiotics is currently the treatment of choice for these infections [30] and this study´s aim was to reach the lowest doses of the antimicrobial agents tested in synergism for an efficient alternative treatment. As the species Serratia marcescens has intrinsic resistance to polymyxins, the experimental polymyxin B concentration used as MIC for SM-KPC was 1 μg.mL -1 , which was the same as the polymyxin B MIC for KP-KPC. The widespread distribution of carbapenemase-producing Enterobacterales has resulted in the increased use of this drug, with the inevitable risk of developing resistance. However, the use of polymyxin is associated with a high incidence of toxicity, such as nephrotoxicity and neurotoxicity [31]. Despite the polymyxins toxicity, when there is no other pharmacological option for treatment, this dosage is administered in patients who present infections with bacteria resistant of carbapenem [32]. The possibility of reducing the necessary dose for treatment with polymyxin B, when associated with CCeo, is a important result since the polymyxin B dose is directly related to its toxicity when used therapeutically. Our study demonstrates that when polymyxin B is associated with CCeo there is a drastic reduction of 320-fold in the dose required for in vitro treatment of the studied multidrug-resistant strains, which brings a benefit to the patient who will not have to receive higher doses of a toxic medication. Moreover, lower doses are expected to reduce the selective pressure upon carbapenemase-producing bacteria, thereby minimizing the appearance of resistance [30]. Concerning CCeo toxicity, the major components of C. cassia are considered to be non-toxic and safe agents with no acute or chronic toxicity, no mutagenicity or genotoxicity, and no carcinogenicity detected in mammalian studies [33].
The synergism results in our study also demonstrate that a dose of 8.79 μg.mL -1 of C. cassia essential oil is able to inhibit the growth of the multi-resistant strains tested. Cinnamaldehyde has been reported as the main component in CCeo and studies showed that Cinnamaldehyde absorption appears to be rapid and complete in mice [34] and in humans. The elimination of cinnamaldehyde by two adult volunteers, who had received a single oral dose of 0.7 mg/kg, was rapid with 100% recovered in the urine within 8 h [35]. Humans also clear systemically available cinnamic acid quickly. Eleven adult human volunteers received single intravenous doses of cinnamic acid, equivalent to 5 mg/kg bodyweight. Plasma was cleared of cinnamic acid (cinnamaldehyde is converted to cinnamic acid in organism) within 20 min [36]. Acute toxicity, oral LD 50 values for cinnamaldehyde in rats and guinea pigs have been reported as 2220 mg/kg and 1160 mg/kg, respectively [33]. Mutagenicity and genotoxicity in mammalian, there was no evidence of an increase in unscheduled DNA synthesis when rats were administered up to 500 mg/kg bodyweight by intraperitoneal injection in a micronucleus assay [37]. The CCeo oil are intended to be used at low levels of exposure relative to doses that elicit adverse effects in laboratory animals via systemic exposure. A C. cassia oral safe dose is 0.7 mg/ kg/day [33], and the bio-accessibility of C. cassia by oral route was approximately 79% [38]. The minor C. cassia essential oil dose, demonstrated in our study (8.79 μg.mL -1 ), able to inhibit the growth of the multi-resistant strains tested when associated with polymyxin B is lower than the literature describes as safe concentrations of C. cassia for humans.
In this study, CCeo was found to be efficient as an excipient for polymyxin B, reducing the dose of antibiotic required for treatment. Thus, micro-emulsions and nano-emulsions synthesized using CCeo may be used as drug delivery systems. Due to their potent antibacterial activity and physicochemical properties, these drug delivery systems could be used to improve the treatment options for human diseases and provide better delivery vehicles for drugs with low bioavailability, thereby decreasing drug toxicity and prolonging the usable life of current antimicrobial drugs [39]. Nanostructured systems have been designed intending essential oils encapsulation as approach to enhance their bio-availability and bio-efficacy as a result of high cellular uptake and controlled release delivery. According to the literature, polymer-based nano-carriers are extensively used for this purpose and no significant cytotoxicity on the normal growth and the development of cultured diploid human cells were observed [40].
The concentrations of leaked proteins after sub-inhibitory (0.5 × MIC) treatments were 2.8-fold (CCeo) and 3.0-fold (CCeo + polymyxin B) higher than the negative control at 1 h. These sub-inhibitory treatment results indicated that the bacterial cells start to leak proteins when they are exposed to CCeo. However, the cells are still viable to grow in culture media. The protein leakage indicates an affected membrane [9] is still able to allow the entrance of polymyxin B to conclude the lysis of the bacterial cell membrane. At 1 × MIC and 2 × MIC of the CCeo and CCeo + polymyxin B treatments, protein leakage was higher. However, late intracellular content extravasation may occur due to other apoptosis pathways being activated [7]. It is possible to assume that CCeo induced concentration-dependent damage in cell membranes, destabilized the outer membrane of the SM-KPC, allowing polymyxin B to enter the periplasm of the cell, disrupting outer membrane integrity, with leakage of cellular contents, and cell death [7]. Moreover, it has already been shown that cinnamon reduces the negativity of the electric charge of E. coli cell membranes, allowing antibiotics to access the penicillinbinding proteins to induce of cell death [27].
A low FIC value index indicates the low MICs of CCeo and polymyxin B in the inhibition of SM-KPC growth. The mechanisms of action of this synergistic effect could involve CCeo blocking the SM-KPC strategies for resistance to polymyxin B or increasing the outer membrane permeability, allowing polymyxin B to enter the bacterial periplasmic space. The resistance strategies of S. marcescens to polymyxins include modifications on their LPS, which normally have negative charges and are the initial targets of polymyxins [4]. Several bacterial cell injuries have already been reported for bacteria treated with cinnamon essential oils [27] and cinnamaldehyde [41]. In this study, the expression of the arnB gene was observed in SM-KPC before and after treatment with polymyxin B. It is known that polymyxin B can upregulate the arnB expression in S. marcescens, indicating that the bacterial cell is able to sense the presence of polymyxin B and undergo a positive feedback reaction by enhancing the arnB expression in order to promote better protection against an imminent drug threat [4]. These results confirm the hypothesis that CCeo disturbs the bacterial outer membrane allowing polymyxin B to enter the cell since, when it was associated with polymyxin B, CCeo was able to inhibit arnB overexpression.
The findings presented here are of great significance given the current importance of polymyxin B in clinical practice and an increase in the generalized bacterial resistance to this drug. However, this research was limited to in vitro studies. Since our findings were promising with KP-KPC and SM-KPC, strains classified with "Critical" priority pathogens on the list for the research and development of new antibiotics by The World Health Organization (WHO) [42], the development of research with another clinically relevant bacteria, should be extended to in vivo investigations, including models of infection.

Conclusion
CCeo was able to inhibit KP-KPC and SM-KPC and was able to exert this inhibition in combination with polymyxin B at a reduced antibiotic dose. The results presented here, therefore, provide a basis for further in vivo and clinical studies for the development of novel treatments using a combination of CCeo and polymyxin B for use against multidrug resistant strains.