Essential oil extraction
In this study, we used 200 g of fresh aerial parts of A. conyzoides collected in the early morning of July of 2017 at the Horto Berta Langes de Morretes, Federal University of Maranhão (UFMA), located in the municipality of São Luís, State of Maranhão (MA), northeast Brazil, Lat. 2°33′13.5″S 44°18′20.8″W. Samples were taxonomically identified and desiccated (voucher) specimens were deposited in the Herbarium of Maranhão—UFMA under the ID number Nº. 9.099. The essential oil of A. conyzoides was extracted from by hydrodistillation in a clevenger apparatus coupled to a Quimis ultrathermostatic bath with a temperature less than 12 °C. The aerial parts of the plant were crushed and placed in a conical flask added with ultrapure distilled water. After 2 h of distillation, the oil was removed from the water surface, centrifuged, and carefully separated from the water added with anhydrous sodium sulfate (JT Baker Chemical Co.), stored in an amber glass ampoule, hermetically sealed, and then stored in a cooled 4 °C for further analysis (Coutinho et al. 2007).
The access was registered under the ID number ADBBA07 in the National System of Management of Genetic Heritage and Associated Traditional Knowledge according to art. 41 of Decree No. 8.772/2016 of the Ministry of the Environment in Brazil.
Characterization of the essential oil by GC/MS
The chemical composition of the essential oil was analyzed by gas phase chromatography/mass spectrometry (GC/MS) with the injection of 1 μL (Auto Injector AOC-20i) in a GCMS-QP 2010 Ultra (Shimadzu) equipped with a Rtx-5MS silica capillary column (Restek, USA) 30 m long × 0.25 mm inner diameter coated with 5%—diphenyl/95%—dimethyl-polysiloxane (0.25 μm film thickness).
The temperature of the GC oven was programmed from 60 to 240 °C at 3 °C/min, injector (1:20 split). Transfer line and ionization chamber temperatures were 250 °C, 250 °C, and 200 °C, respectively. Helium was used as the entrainment gas at a rate of 1 μL/min.
The mass spectra were obtained by electronic impact at 70 eV with automatic scans in the mass range between 35 and 400 m/z at 0.30 scans/s.
The identification of the components was based on the time and linear retention index (series of C8–C28 n-alkanes) and on the interpretation and comparison of the mass spectra obtained from the libraries (Adams 2012; NIST 2011).
Microorganism
DH82 cells (Canine Histiocyte: ATCC No. CRL-10389) infected with 35th passage of the Cuiabá #1 strain of E. canis were cultured in Dulbecco’s Modified Eagle’s (DMEM) medium (Sigma Chemical Co., St. Louis), MO, USA) supplemented with 5% fetal calf serum (HyClone Laboratories, Logan, Utah, USA) and maintained in a 25 cm2 culture bottle at 37 °C with 5% CO2 as recommended by Aguiar et al. (2007). E. canis infection rate was determined by examining (screening) smears from a monolayer cell stained by the Diff-Quik Kit (Laborclin, Pinhais, PR, Brazil) under the light microscope.
When an rickettsial infection rate of 70% was detected using this method, the cells were resuspended with the same effect and the cell suspension was centrifuged at 4000g for 5 min. The experiments were run on 24-well culture plates at 37 °C with 5% CO2. The bacterial rate was standardized as 3000 cells per well and 70% of the cells infected with the rickettsia.
The access was registered under number A9463BB in the National System of Management of Genetic Heritage and Associated Traditional Knowledge according to art. 41 of Decree No. 8.772/2016 of the Brazilian Ministry of the Environment.
Biological assay
The assays were performed in triplicate at concentrations of 25, 50, 100, 200, 300, 400, and 500 µg/mL of the essential oil of A. conyzoides L. plus 1% Dimethylsulfoxide-DMSO (Merck Chemical Co.) in order to solubilize the sample and at concentrations of 0.25, 0.50, 0.75, 1.0, 1.5 µg/mL of the doxycycline plus 1% Dimethylsulfoxide-DMSO (Merck Chemical Co.). Analyses were performed at 18 h and 36 h after addition of the treatments to the medium. The analyses consisted of counting the percentage of infected cells on Diff-Quik (Laborclin, Pinhais, PR, Brazil) stained cell monolayer smear preparations examined under the light microscope (Aguiar et al. 2007).
The experiments were performed according to the method published by Chou (2006). The treatments were tested at constant ratios of equipotent concentrations, ranging from 0.0625-, 0.125-, 0.25-, 0.5-, 1-fold of their respective IC50 value that was determined for each experiment. The synergism, antagonism, or additive effect of each of the combinations was assessed by calculating the combination index value (CI). According to Chou (2006) and Chou (1976):
$$ \frac{{d_{1} }}{{d_{1} x}} + \frac{{d_{2} }}{{d_{2} x}} $$
(1)
in which D1 and D2 are the doses of doxycycline (1) and EO (2), respectively, that are responsible for an effect x in combination whereas D1x and D2x are the doses of 1 and 2, respectively, that are responsible for the same effect individually. If CI < 1, the treatments have a synergistic effect; if CI > 1, they are antagonistic; if CI = 1, an additive effect is observed. A normalised isobologram was created by plotting the normalised concentrations \( \frac{{d_{1} }}{{d_{1} x}} \) of 1 and \( \frac{{d_{2} }}{{d_{2} x}} \) of 2 on the y- and x-axis, respectively, in which the denominators represent the respective doses of c 1 and 2 alone reducing antibacterial load by x%, and the numerators represent the respective doses of 1 and 2 reducing bacterial load by x% in combination. The normalised concentrations were calculated considering that
$$ \frac{{D_{1} }}{{D_{1} x}} = \frac{{D_{1} }}{{Dm_{1} \cdot \left( {\frac{fa}{1 - fa}} \right)^{{\frac{1}{{m_{1} }}^{{}} }} }} $$
(2)
in which \( D_{{m_{1} }} \) is the IC50 of 1 in vitro, fa is the fraction affected (or (% effect) ÷ 100), and m1 is the slope of linear regressions from median effect plots using the function
$$ log\left( {\frac{fa}{1 - fa}} \right) = f\left( {log\left( {D_{1} x} \right)} \right) $$
(3)
The CI-effect graph representing the CI as a function of the associated antibacterial effect was also plotted as well as the log (DRI)-effect plot representing the log of the dose reduction index (DRI) as a function of the associated antibacterial effect. The DRI is the ratio of the concentration of a treatment resulting in an effect x alone (D1x) to the concentration of the same treatment resulting in an effect x in combination (D1):
$$ DRI = \frac{{D_{1} x}}{{D_{1} }} $$
(4)
Ehrlichia canis suspensions were standardized at 800 cells/well with a 70% infection rate using 96-well culture plates. Solutions of the tested products were used in concentrations determined from their respective IC50 The protocol used to determine the antimicrobial effect of the essential oil and doxycycline were an adaptation from Rolain et al. (1998) and Rolain et al. (2002) regarding serial dilutions 0.0625 0.25, 0.5 and 1 times the respective IC50 value that was determined for each treatment. Initially, we added 200 μL of medium into each well of a sterile microplate. Subsequently, 50 μL of each product tested in serial dilutions were arranged in an orderly fashion so that we were able to evaluate activity according to the decrease of the essential oil and the synthetic drug. From top to bottom, there is decrease of the essential oil IC50, and horizontally from the right to the left there is a decrease of the synthetic drug (Nightingale et al. 2007).
Our results show that in each well there is a unique combination of concentrations between the two substances (i.e., essential oil of A. conyzoides and doxycycline).
Statistical analysis
The analyses were performed using the software GraphPad Prism 5.0 (GraphPad Software, La Jolla, California, USA). The Student’s test was for toxicity analysis and the analysis of variance (ANOVA) was used to obtain data on the rate of inhibition of the microorganisms and the treatment time of the groups. Statistically, significant differences were found with values of p < 0.05. The IC50 value was also acquired by linear regression using the software GraphPad Prism 5.0 (GraphPad Software, La Jolla California USA). The effects of the interaction between the treatments and doxycycline were also evaluated by the analysis of the combination of multiple drugs using the software CompuSyn® (Chou and Talalay 1984).