Strains and mice for infection model
Before using geraniol to treat bovine mastitis, a preliminary experiment was conducted in mice to evaluate its efficacy, safety, and dosage. Geraniol was used to treat mice infected with Escherichia coli, one of the main pathogenic bacteria causing bovine mastitis. The bacterium was isolated from cow milk with clinical mastitis and stored in the Sichuan Experimental Teaching Demonstration Center for Medical Testing of Chengdu Medical College (Chengdu, China). Specific pathogen-free (SPF) Kunming (KM) mice (male; body weight, 20 ± 2 g) provided by Sichuan Chengdu Institute of Biological Products Co., Ltd. (Chengdu, China) were used in the experiment.
Mouse model of E. coli infection
After adaptive feeding for 2 weeks in the laboratory, 80 KM mice were randomly divided into eight groups (n = 10), including seven experimental groups and one blank group. The pathogen E. coli was cultured at a constant temperature of 37 °C in Luria-Bertani (LB) medium (50 mL) until the logarithmic phase. The bacterial culture was diluted using physiological saline to seven concentrations (5.85 × 107, 2.925 × 107, 3.663 × 106, 3.663 × 106, 1.831 × 106, 1.371 × 106, and 0.911 × 106) and mixed with 5% porcine stomach mucin (Sigma, St Louis, MO, USA). The E. coli solution was intramuscularly injected into seven experimental groups, and the mice in the blank group were intraperitoneally injected with the same volume of normal saline. On the premise that the blank control group mice did not die, the mortality of the mice in the experimental group was recorded within 72 hours (every 12 hours). The minimum concentration that caused the death of all mice in a group was considered as the minimum lethal dose (MLD) of E. coli. Meanwhile, the lowest bacterial load that killed half of the mice was considered the half lethal dose (LD50) of E. coli.
Further, 90 KM mice were randomly divided into eight groups (n = 10), including blank group, E. coli infection group, solvent group, positive control group, and five treatment groups. The blank group and E. coli infection group were injected intramuscularly with normal saline; the solvent group was injected intramuscularly with Twain-80; the positive control group was injected intramuscularly with cefotaxime (0.3 g/kg); and the geraniol treatment groups were injected intramuscularly with different doses of geraniol (0.261 g/kg, 0.166 g/kg, 0.107 g/kg, 0.07 g/kg, and 0.045 g/kg). Each mouse was injected once daily for 5 days. After the last intramuscular injection, mice in all groups, except the blank group, were intraperitoneally injected with the MLD of E. coli. The survival rate of the mice was observed within 7 days after E. coli challenge, and the dose of geraniol that maintained 50% survival rate of mice was considered ED50.
Assessment of antibacterial activity of geraniol in mouse infection model
Mice infected with LD50 of E. coli were treated with the ED50 of geraniol to evaluate its efficacy. Forty KM mice were randomly divided into four groups (n = 10), including the control group, half lethal E. coli infection group, the geraniol treatment group, and the antibiotic treatment group. Half lethal E. coli infection group, antibiotic treatment group, and geraniol treatment group were challenged with the LD50 of E. coli, and then intramuscularly injected with saline, cefotaxime, and geraniol (ED50), respectively. The control group was challenged with the same dose of inactivated E. coli, and intramuscularly injected with saline. Saline, cefotaxime (0.3 g/kg), and geraniol (0.26 g/kg) were administered once a day for five consecutive days.
Further, blood was collected from the eyeball venous plexus of mice at 0.5 d, 1d, 2d, 3d, and 7d after infection to determine the dynamic changes in routine blood index. TEK-Ⅱ MINI A3-0052 three-class blood cell counter (Tekang Technology, China) was used to determine the different blood parameters, including white blood cell (WBC) count, absolute lymphocyte (LYM) value, absolute intermediate cell (MID), absolute granulocyte (GRA) value, red blood cell (RBC) count, hemoglobin (HGB), hematocrit (HCT), mean red blood cell volume (MCV), mean hemoglobin content (MCH), mean hemoglobin concentration (MHCH), platelet (PLT) count, platelet packed volume (PCT), and average platelet volume (MPV). All the experiments were completed within 24 h after blood collection.
Fresh fecal samples of mice were collected for monitoring the dynamic changes in the gut microbiome. Fecal samples were collected before (0 h) and after (12 h) E. coli infection to evaluate the effect of infectious pathogens on the mouse gut microbiome. Samples collected before infection were used as a baseline (0 day) of the mouse gut microbiome. Fecal samples were also collected on 1st, 7th, 14th, and 28th days after treatment. All samples were quickly frozen in liquid nitrogen and transferred to -80 °C freezer until further analysis. The schematic diagram of the experiment is shown in Fig. S1A.
Oral administration of geraniol in healthy mice
Further, the effect of oral geraniol on the community structure of the gut microbiome in animals was analyzed. Healthy mice were fed with geraniol or antibiotics, and the dynamic changes in the gut microbiota of mice were examined. Thirty mice were divided into three groups (n = 10), including the control group, oral antibiotic group, and oral geraniol group. The mice of the antibiotic and geraniol groups were fed cefotaxime and geraniol, respectively, via gavage from 1th to 3th day, once a day. The therapeutic dose (ED50) was chosen as the oral feeding dose. The mice of the control group were fed normal saline by gavage. Fecal samples were collected from all mice on day 0 (prior to the experiment), 1, 3, 7, 14, and 28. The schematic diagram of the experiment is shown in Fig. S1B.
Experimental set up in dairy cows
The study was conducted on a dairy farm in Chengdu, Sichuan Province, China, from April 2020 to September 2020. All cows used in this study were the imported Holstein lactating adult cows (3–6 years) from Uruguay. This study did not involve any endangered or protected animal species and did not cause any harm to the experimental animals. Cows were provided a standard diet composed of grass-legume hay to meet the daily nutrient requirements for milk production. The cows did not receive antibiotics or other drugs during the past two months. Somatic cell count (SCC, the number of leukocytes per milliliter of fresh milk) and clinical symptoms were used to diagnose mastitis . A Bentley FTS/FCM400 Combi Instrument (Chaska, USA) was used to measure SCC in the milk. Cows with SCC < 2 ×104 cells/mL and no obvious signs of clinical mastitis, such as breast redness and heat, and no damaged nipple ends were diagnosed as healthy, while those with SCC > 2 ×106 cells/mL and symptoms such as breast redness, swelling, heat and pain, elevated body temperature, and abnormal milk color were considered as infected cows (mastitis). A total of 12 healthy and 25 infected (mastitis) cows were recruited for this study. The mastitis group was divided into the antibiotic treatment group (n = 13) and the geraniol treatment group (n = 12); the drug was injected into the cows' breasts. Detailed information on the individual cows and grouping is presented in Table S1.
Mastitis treatment and sample collection
In the geraniol treatment group, each cow with mastitis was injected 1.18 g of geraniol twice a day. In the antibiotic treatment group, each cow with mastitis was injected 10 g of cefalexin and kanamycin monosulfate intramammary infusion (Cephalexin: Kanamycin = 2:1; Tullyvin, Cootehill, Co. Cavan, Ireland), once a day. Treatments were carried out for 5 days. The cows with mastitis that showed no significant improvement on the 7 day after treatment were considered incurable by the drug. Fresh stool, milk, and blood samples collected at 0 days (before the experiment) were considered baseline. Materials were also sampled on 3rd, 5th, and 14th days. The middle part of the fresh stool was collected immediately after defecation using a sterile fecal collector. The nipples were wiped with iodine and 75% ethanol and cleaned using sterile distilled water to collect milk without pollution. While sampling, the first three milk strands were discarded, and then the milk (50 mL) was squeezed into a sterile centrifuge tube. The venous blood was obtained using a sterile syringe. Blood was centrifuged at 3000 rpm for 5 min to get the serum. These serum samples were stored at -4 °C while feces and milk samples were stored in a -80 °C freezer until further analysis. The schematic diagram of the experiment is shown in Fig. S1C.
Detection of inflammatory cytokines in cows
The levels of interleukin-6 (IL-6), interleukin-1β (IL-1β), tumor necrosis factor-α (TNF-α), cyclooxygenase-2 (COX-2), and inducible nitric oxide synthase (iNOS) in the serum of dairy cows with clinical mastitis were measured using enzyme-linked immunosorbent assay (ELISA) kits, according to the manufacturer’s instructions. All ELISA kits were purchased from Shanghai Future Industrial Co., Ltd. (Shanghai, China).
Bacterial DNA extraction, 16S rRNA sequencing, and bioinformatic analysis
Bacterial total genomic DNA was extracted using the Mo Bio PowerFecal DNA isolation kit (Mo Bio Laboratories, Carlsbad, CA, USA), following the manufacturer’s instructions. For the stool samples, a frozen aliquot (200 mg) of each sample was transferred directly to the PowerBead tubes for extracting total genomic DNA. For milk samples, 10 mL of each sample was centrifuged at 12,000 g for 5 min at 4 °C to remove the lipid layer from the milk. The supernatant was discarded, and the watercourse was collected and filtered through a filter membrane (0.2 µm). The filter membrane was transferred to the PowerBead tubes to extract the bacterial total genomic DNA. NanoDrop (Thermo Fisher Scientific, Waltham, USA) was used to measure the concentration and purity of the genomic DNA. The quality of DNA was assessed by agarose gel electrophoresis. Only samples that met the following criteria were used for PCR: (1) DNA concentration > 5 ng/uL; (2) total DNA > 150 ng; and (3) complete and contamination-free DNA fragment. The universal 515F/806R barcoded primer pair (515F: GTGCCAGCMGCCGCGGTAA, 806R: GGACTACHVGGGTWTCTAAT)  was used to amplify the V4 region of the bacterial 16S rRNA gene. The PCR reaction mixture (50 µL) contained 2 µL DNA (20 ng per sample), 25 µL 2×NEB Phusion High-Fidelity PCR Master Mix, 2 µL (0.4 µM) forward primer, 2 µL (0.4 µM) reverse primer, and 19 µL sterile double-distilled water. The thermocycling parameters were as follows: 95 °C for 3 min, followed by 30 cycles of 95 °C for 45 s, 56 °C for 30 s, and 72 °C for 90 s, and a final extension at 72 °C for 10 min. AmpureXP beads (Agencourt Beckman Coulter, Beverly, MA, USA) were used to purify the PCR products. Sequencing libraries were constructed using the MiSeq Reagent Kit v2 (Illumina, CA, USA). PCR, sequencing library construction, and paired-end sequencing (Illumina MiSeq platform) were executed at Novogene (Beijing, China).
The 16S rRNA gene raw sequences were pre-processed and analyzed using the QIIME2 pipeline (version 2019.9) . DADA2 software within QIIME2 was utilized to filter low quality (< Q20), chimeric, and erroneous reads . Sequence variants were aligned based on the MAFFT program . Further, a phylogenetic tree based on the clean reads was constructed using FASTTREE algorithm . Meanwhile, sequences identified as chloroplast and mitochondria were filtered out. Taxonomy was assigned to operational taxonomic units (OTUs) using BLAST to align representative sequences to the SILVA database (version 132) , and singleton OTUs were removed from downstream analysis. The OTU tables were rarefied to the minimum number of reads per sample before alpha or beta diversity analysis to eliminate the bias caused due to variant sequencing depth. Alpha-diversity indices, including the number of observed OTUs and Shannon’s diversity index, were calculated using QIIME2. PCoA (Principal Co-ordinate Analysis) was performed based on weighted UniFrac distances in QIIME2 to measure the similarity in the bacterial community between the samples .
Determination of drug residues in milk of dairy cows after antibiotic or geraniol injection
After injection of antibiotics and geraniol into the mammary gland, the drug residues in milk were determined every day from the 1st to the 7th day after stopping administration. The geraniol standard (purity > 98%, No. SHBL2152), HPLC-grade acetonitrile (ACN), and methanol (MeOH) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Cephalexin (purity > 98%, No. 130408-201411) and kanamycin (purity > 98%, No. 130556-201502) were obtained from National Institute for Food and Drug Control (Beijing, China). The residue of geraniol in milk was detected and analyzed by gas chromatography-mass spectrometer (GC-MS; Agilent 7890A/5977B gas chromatography-mass spectrometer system) (Santa Clara, California, USA) , whereas cephalexin and kanamycin residues were analyzed by high-performance liquid chromatography (HPLC; Agilent LC-1260, Agilent Technologies, Santa Clara, CA, USA). GC-MS analysis for the residues in the milk samples was carried out according to a previously reported method , and the residues of cephalexin and kanamycin in milk samples was detected in accordance with the national standards GB/T 21314-2007  and GB/T 22969-2008 , separately. Further, the specificity of the method was determined by analyzing the chromatographic peaks of the standard and the milk samples to assess the accuracy of the method.
Assessment of geraniol and antibiotics induced bacterial drug resistance
Geraniol and amoxicillin were used to induce drug resistance of bacteria in vitro to verify whether geraniol causes pathogenic bacteria to develop drug resistance. The standard strain (E. coli, ATCC25922) stored at - 80 ℃ was inoculated onto an agar plate and cultured at 37 ℃ for 24 h to obtain a single colony. Prepare E. coli into 0.5 Michaelis concentration unit (1.5 × 108 CFU/mL) with sterile normal saline and diluted 100 times with Mueller-Hinton Broth (MHB) medium. Then, geraniol emulsion (86 mg/mL) mother liquor was prepared with Tween-80 as an emulsifier, while amoxicillin was dissolved in sterile water, and the amoxicillin mother liquor was prepared at a concentration of 4096 µg/mL. The geraniol emulsion and amoxicillin mother liquor were diluted into 12 different concentrations by double dilution to determine their minimum inhibitory concentration (MIC). Further, 1 mL of MHB culture solution, 0.5 mL of diluted geraniol solution, and 0.5 mL of the standard strain (1.5 × 108 CFU/mL) were added into the test tube one after the other. The final concentration of geraniol and amoxicillin in the test tube was maintained ranging from 0.02 to 43.00 mg/mL and 1 to 2048 µg/mL, respectively. This mixture was incubated at 37 ℃ for 24 h to observe the growth of the standard strain. The standard strain without drug was used as the positive control, while drug alone with no bacteria was used as the negative drug control, and tubes containing only culture medium were blank controls. The minimum drug concentration that inhibited the growth of the standard strain was taken as the MIC value of the bacteria. After the MIC value was determined, geraniol/amoxicillin solution at the sub-inhibitory concentration (1/2MIC) and the standard strain (0.5 Michaelis concentration) were added to the MHB medium. These strains were cultured at 37 ℃ and 200 rpm for 24 h as the first-generation resistant strains induced by drugs. Following the above method, the first-generation resistant strain was repeatedly cultured with geraniol/amoxicillin to obtain the second-generation strain, and the MIC values of geraniol and amoxicillin were determined every ten generations.
Significance differences between the mean values of three or more groups were analyzed using the Kruskal-Wallis test followed by the post-hoc Dunn's multiple comparison test in GraphPad Prism 7 (GraphPad Software, Inc., USA). Mann-Whitney U test was used to compare the differences in mean values between two groups. The result graphs were created using“boxplot,” “barplot,” “ggplot2”, and “plot” functions in base R (version 3.5) .