Geraniol (purity > 98%), was obtained from Sigma Chemical Co. (St. Louis, MO), and its solubility was improved by emulsifying in anhydrous glycerol according to the previously reported method (Pavan et al., 2018). Specific pathogen-free (SPF) Kunming (KM) mice (male, body weight, 20 ± 2 g) were provided by Sichuan Chengdu Institute of Biological Products Co., Ltd. (Chengdu, China). The Escherichia coli strain ATCC25922 was obtained from the American Type Culture Collection (ATCC, Rockville, MD, USA). 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 the National Institute for Food and Drug Control (Beijing, China).

A preliminary experiment was conducted in mice to evaluate its efficacy, safety, and dosage before using geraniol to treat mice infected with Escherichia coli strain ATCC25922. After adaptive feeding for two weeks in the laboratory, 80 KM mice were randomly divided into eight groups (n = 10), including seven experimental groups and one blank group. The Escherichia coli strain ATCC25922 was cultured at a constant temperature of 37°C in Luria-Bertani (LB) medium (50 mL) until the culture reached a logarithmic phase.

Then, the bacterial culture was diluted to seven concentrations (5.85 × 10⁷, 2.925 × 10⁷, 3.663 × 10⁶, 2.75 × 10⁶, 1.831 × 10⁶, 1.371 × 10⁶, and 0.911 × 10⁶ CFU/mL) using PBS and mixed with 5% porcine stomach mucin (Sigma, St Louis, MO, USA).

The mouse model with E. coli infection was successfully established following the previously study (Long et al., 2022). The minimum concentration causing the death of all and half of the mice in a group was considered as the minimum lethal dose (MLD) and half lethal dose (LD₅₀) of E. coli, separately.

Afterwards, 90 KM mice were randomly divided into nine groups (n = 10), including the blank, E. coli infection, solvent, positive control, and five geraniol treatment groups. The blank and E. coli infection groups were intramuscularly injected with PBS; the solvent group was intramuscularly injected with Twain-80; the positive control group was intramuscularly injected with cefotaxime (0.3 g/kg); and the geraniol treatment groups were intramuscularly injected with different doses of geraniol (0.261 g/kg, 0.166 g/kg, 0.092 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, the mice in all groups, except for the blank group, were intraperitoneally injected with the MLD of E. coli. The survival rate of the mice was observed 7 days after E. coli challenge, and the dose of geraniol maintaining 50% survival rate was considered as ED₅₀.

The efficacy of geraniol was evaluated by treating the mice infected with LD₅₀ of E. coli with the ED₅₀ of geraniol. 40 KM mice were randomly divided into four groups (n = 10), including the control, half-lethal E. coli infection, geraniol treatment, and antibiotic treatment groups. The half-lethal E. coli infection, antibiotic treatment, and geraniol treatment groups were challenged with the LD₅₀ of E. coli and then intramuscularly injected with physiological saline, cefotaxime (0.3 g/kg), and geraniol (ED₅₀) once a day for five consecutive days, respectively. The control group was challenged with the same dose of inactivated E. coli, and intramuscularly injected with physiological saline.

Later, the blood was collected from the eyeball venous plexus of mice on 0.5d, 1d, 2d, 3d, and 7d after infection, and the blood parameters were determined using TEK-II MINI A3-0052 three-class blood cell counter (Tekang Technology, China). All the experiments were completed within 24 h after blood collection. The fecal samples were collected before (0 h) and after (12 h) E. coli infection to explore the effect of infectious pathogens on the mouse gut microbiome. The samples collected before infection were used as a baseline (0 day) of the mouse gut microbiome. The fecal samples were collected on the 1^st, 7^th, 14^th, and 28^th days after treatment. All samples were quickly frozen in liquid nitrogen and transferred to a -80°C freezer until further analysis. The schematic diagram of this experiment is illustrated in Figure S1A .

Theeffect of oral geraniol on gut microbiome community structure was analyzed. 30 KM mice were divided into three groups (n = 10), including the control, antibiotic, and geraniol groups. The mice of the antibiotic and geraniol groups were fed with the therapeutic dose of cefotaxime (0.3 g/kg) and geraniol (0.261 g/kg), respectively, by gavage from 1st to 3rd day, once a day. The mice of the control group were fed with physiological saline by gavage. The fecal samples were collected from all mice on days 0, 1, 3, 7, 14, and 28. The schematic diagram of this experiment is illustrated in Figure S1B .

This 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 imported Holstein lactating adult cows (3–6 years) from Uruguay. The cows were provided with 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. The somatic cell count (SCC) and clinical symptoms were used to diagnose mastitis (Pilla et al., 2013). The SSC in milk was measured using a Bentley FTS/FCM400 Combi Instrument (Chaska, USA). A total of 12 healthy and 27 infected (with subacute clinical mastitis) cows were recruited for this study. The mastitis group was divided into the antibiotic treatment (n = 14) and the geraniol treatment groups (n = 13). The drug was injected into the cows’ breasts. Detailed information on individual cows and grouping is presented in Table S1.

A drug dosage of 0.0944 g/kg was considered as the ED50 of geraniol against E. coli infection in mice. The dosage per unit body weight of cow was about 1/24 of the mice. Therefore, the dosage of geraniol administered to cows was about 0.00393 g/kg. If the average weight of cows is around 600 kg, the dosage of each cow is 2.36 g. In the pre-experiment, high (2.36 g), medium (1.18 g), and low (0.59 g) dosage of geraniol were administered, respectively. The effects of high and medium dosages were relatively good. Therefore, the medium dosage (1.18 g) was selected to treat the cows in this study.

In the geraniol treatment group, each cow was injected with 1.18 g of geraniol twice a day. In the antibiotic treatment group, each cow was injected with 10 g of cefalexin and kanamycin monosulfate intramammary infusion (Cephalexin: Kanamycin = 2:1; Tullyvin, Cootehill, Co. Cavan, Ireland) once a day. The treatments were carried out for 5 consecutive days. The cows with mastitis with no significant improvement after 7 days of treatment were considered incurable by the drug. The fresh stool, milk, and blood samples collected on 0 days were considered baseline and sampled on the 3^rd, 5^th, and 14^th 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 the unpolluted milk. During 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 and centrifuged at 3000 rpm for 5 min to obtain the serum. The serum samples were stored at -4°C, while the feces and milk samples were stored in a -80°C freezer until further analysis. The schematic diagram of this experiment is illustrated in Figure S1C .

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 the enzyme-linked immunosorbent assay (ELISA) kits (Shanghai Future Industrial Co., Ltd., Shanghai, China) according to the manufacturer’s instructions (Long et al., 2022).

The total bacterial genomic DNA was extracted using the Mo Bio PowerFecal DNA isolation kit (Mo Bio Laboratories, Carlsbad, CA, USA), following the manufacturer’s instructions. Later, a frozen aliquot (200 mg) of each stool sample was transferred directly to the PowerBead tubes to extract the total genomic DNA. Then, 10 mL of each milk sample was centrifuged at 12,000 g for 5 min at 4°C to remove the lipid layer. The supernatant was discarded, and the watercourse was collected and filtered through a 0.2 µm filter membrane. The filter membrane was transferred to the PowerBead tubes to extract the total bacterial genomic DNA. The concentration and purity of the genomic DNA were measured using NanoDrop (Thermo Fisher Scientific, Waltham, USA). The universal 515F/806R barcoded primer pair (Caporaso et al., 2011) was used to amplify the V4 region of the bacterial 16S rRNA gene. The 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 performed at Novogene (Beijing, China).

The 16S rRNA gene raw sequences were pre-processed and analyzed using the QIIME2 pipeline (version 2019.9) (Bolyen et al., 2018). Low quality (< Q20), chimeric, and erroneous reads were filtered by DADA2 software (Callahan et al., 2016). The amplicon sequences variants (ASVs) were generated from clean reads using the DADA2 pipeline. Later, aphylogenetic tree based on the clean reads was constructed using the FASTTREE algorithm (Price et al., 2012). The taxonomy was determined by BLAST using the representative sequences against the SILVA reference database (version 132) (Quast et al., 2013). The ASVs 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. The alpha-diversity indices, including the number of observed ASVs and Shannon’s diversity index, were calculated using QIIME2. PCoA (Principal Co-ordinate Analysis) was performed based on weighted UniFrac distances in QIIME2 to explore the similarity in the bacterial community between the samples (Lozupone and Knight, 2005).

The quantitative PCR primers for Enterobacteriaceae, Streptococcus, Mycoplasma, Lactobacillus, and Bifidobacterium were designed according to the 16S rRNA representative sequences annotated as these genera. The DNA of cow milk bacteria in the antibiotic treatment group and the geraniol treatment group was used as a template for quantitative PCR.The primers and thermocycling parameters are presented in Table S2 and Table S3. Taking the bacteria in the dairy cows’ milk obtained on day 0 as the baseline, the relative quantification of Enterobacteriaceae, Streptococcus, Mycoplasma, Lactobacillus, and Bifidobacterium was calculated on the 3^rd, 5^th, and 14^th days. The 16S rRNA gene was used for bacterial normalization.

After injecting antibiotics and geraniol into the mammary gland, the drug residues in milk were determined every day from the 1st to the 7th day till the discontinuation treatment. The geraniol residue 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) following a previously reported method (Min et al., 2016), whereas cephalexin and kanamycin residues were analyzed by high-performance liquid chromatography (HPLC; Agilent LC-1260, Agilent Technologies, Santa Clara, CA, USA), according to the national standards GB/T 21314-2007 (Dasenaki et al., 2015) and GB/T 22969-2008 (GB/T 22969-2008, 2008), separately. Afterwards, a quarter of the antibiotic treatment dose (about 2.0 g) was injected into the cow’s (n=5) breast, and its residue in milk was detected following the above method to verify the consequence of different injection dose on longer residual time of antibiotics in milk.

The standard strain (E. coli, ATCC25922, and S. aureus, ATCC25923) was diluted into 0.5 Michaelis concentration units (1.5 × 10⁸ CFU/mL) with sterile normal saline, and then diluted 100 times with Mueller-Hinton Broth (MHB) medium. Later, a mother liquor of amoxicillin and ampicillin was prepared at a concentration of 4096 µg/mL using Tween-80 as an emulsifier. The minimum inhibitory concentration (MIC) of geraniol emulsion, amoxicillin, and ampicillin were determine by diluting their mother liquor into 12 different concentrations by double dilution. Afterwards, 1 mL of the MHB culture solution, 0.5 mL of the diluted geraniol solution, and 0.5 mL of the standard strain (1.5 × 10⁸ CFU/mL) were added to the test tube. The final concentrations of geraniol, amoxicillin, and ampicillin in the test tube were maintained from 0.02 to 43.00 mg/mL, 1 to 2048 µg/mL, and 1 to 2048 µg/mL, respectively. Later, the mixture was incubated at 37 °C for 24 h to observe the growth of the standard strain. The minimum drug concentration inhibiting the growth of the standard strain was regarded as the MIC value of the bacteria. The strains cultured in the MHB medium supplemented with sub-inhibitory concentration (1/2MIC) induction drugs at 37 °C and 200 rpm for 24 h were regarded as the first-generation resistant strains. The first-generation resistant strain was repeatedly cultured with geraniol, amoxicillin, and ampicillin to obtain the second-generation strain, and the new MIC values of geraniol, amoxicillin, and ampicillin were determined every ten generations following the above method.

Significance differences between the mean values of three or more groups were analyzed using one-way ANOVA, followed by the post-hoc Dunn’s multiple comparison tests in GraphPad Prism 7 (GraphPad Software, Inc., USA). The differences in mean values between the two groups were compared using Mann-Whitney U test. The graphs were plotted using the “boxplot,” “barplot,” “ggplot2”, and “plot” functions in base R (version 3.5) (Stiglic et al., 2019).

The MLD of the mice and LD₅₀ of E. coli were 2.925×10⁷ and 0.911×10⁶ CFU/kg, respectively (Table S4). After 5 days of prophylactic administration of geraniol, significant antibacterial activity was observed in the mice challenged with MLD of E. coli. Geraniol at 0.261 g/kg completely (100%) prevented E. coli infection. The prevention rate decreased with the decrease in dose, and the drug dose of 0.0944 g/kg was considered as the ED₅₀ of geraniol against E. coli infected mice (Table S5).

Similar to cefotaxime, geraniol cured 100% of infection in the mice challenged with LD₅₀ of E. coli (Table S6). The routine blood indices, including WBC, LYM, MID, GRA, PCT, PLT, RBC, MPV, and MCV, significantly decreased after infection (p < 0.05; Figure S2 ). These indices increased after treatment with cefotaxime and geraniol ( Figure 1 ), and reached a similar level to that of the healthy mice. of them, LYM, MID, RBC, MPV, and MCV were restored to significantly higher levels in the geraniol treatment group than in the antibiotic treatment group.

There was no significant difference in the alpha diversity indices (Shannon index and the number of observed ASVs; Figures S3A, B ) and PCoA ( Figure S3C ) of mice gut microbiota infected with E. coli than the pre-infected mice. After treatment with cefotaxime, the Shannon index ( Figure 2A ) and the number of observed ASVs ( Figure 2B ) significantly decreased. The gut microbial diversity significantly decreased on the 1^st, 7^th, and 14^th days after intraperitoneal injection of antibiotics (p < 0.05) and did not return to normal until the 28^th day. PCoA by the weighted UniFrac matrix suggested that antibiotic treatment destroyed the gut microbiota ( Figure 2E ). The phylogenetic distance between the gut microbiota after (1^st/7^th/14^th/28^th day) and before treatment was significantly larger than that within the samples before treatment (the 0^th day) ( Figure 2F ; p < 0.05), where the gut microbiota of the mice before treatment was considered as the baseline. After geraniol treatment, the Shannon index and the number of observed ASVs of gut microbiota showed no significant decrease, while those on the 28^th day were significantly higher than those on the 1^st and 7^th days (p < 0.05; Figures 2C, D ). Meanwhile, the PCoA of gut microbiota showed no significant changes after geraniol treatment ( Figure 2G ). The phylogenetic distance between the gut microbiota after (1^st/7^th/14^th/28^th day) and before treatment was not significantly different from the distance between the gut microbiota before treatment (p > 0.05; Figure 2H ).

The Shannon index ( Figure 3B ) and the number of observed ASVs ( Figure 3E ), significantly (p < 0.05) decreased after the mice were fed with cefotaxime. The gut microbial diversity of the mice continued to decrease from the 1^st to the 7^th day after providing cefotaxime by oral gavage. The diversity was recovered on the 14^th day after the experiment (the 10^th day after stopping cefotaxime orally); however, it did not return to the normal level until the 28^th day. In contrast, the Shannon index ( Figure 3C ) and the number of observed ASVs ( Figure 3F ) of gut microbiota in mice did not decrease after being fed with geraniol. The alpha diversity indices were similar to that in the control mice ( Figures 3A, D ). The number of observed ASVs in the mice fed with geraniol on the 28^th day after the experiment was significantly higher than before (0 days; Figure 3F ). PCoA indicated that the gut microbiota composition in the geraniol group and the control group did not change significantly during the whole experiment (0^th to 28^th; p > 0.05; Figure 3G ). However, the PCoA of the gut microbiota was significantly destroyed after being fed with cefotaxime, and it did not return to the state before administration by the 28^th day (24^th day after stopping treatment; p < 0.001; Figure 3G ).

The dairy cows withmastitis showed a significant increase in SCC in the milk ( Figure S4A ). After 5 days of treatment with geraniol and antibiotics (cefotaxime and kanamycin), the SCC in the dairy cows’ milk with mastitis significantly reduced (less than 2× 10⁶ cells/mL) ( Figures S4B, C ), and the clinical symptoms subsided (Table S7). During the treatment by breast infusionof geraniol, 10 of the 13 sick cows recovered with anaverage cure time of 7 days, while 10 of the 14 sick cows recovered with antibiotics wiyh anaverage cure time of 6 days (Table S7). These results indicated that the average cure rate of geraniol was similar to that of antibiotics.

During inflammation, the pro-inflammatory mediators, including IL-1β, IL-6, TNF-α, COX-2, and iNOS, are produced at the inflammation site. The levels of IL-1β, IL-6, TNF-α, COX-2, and iNOS in the serum of cows with clinical mastitis were significantly higher than those in the healthy cows (p < 0.001; Figure S5 ). The levels of these inflammatory factors (L-1β, IL -6, TNF-α, COX-2, and iNOS) in blood decreased after injecting the antibiotics and geraniol into the breast of the dairy cows with clinical mastitis ( Figure 4 ). However, the levels of these inflammatory factors returned to their normal levels by the 14^th day after geraniol or antibiotics administration ( Figure 4 ).

The alpha diversities of the bacteria in the milk ( Figures 5A, B ) and gut ( Figures 6A, B ) of the dairy cows with mastitis were significantly lower (p < 0.05) than those in the healthy dairy cows. PCoA indicated that the milk ( Figure 5C ) and gut ( Figure 6C ) microbiota structure of cows with mastitis was significantly different from the healthy dairy cows. Mastitis resulted in a significant increase in the intra-individual variations in milk ( Figure 5D ) and gut ( Figure 6D ) microbiota, and the phylogenetic distance between the healthy group and mastitis group was significantly larger than that in the healthy or mastitis group. The dairy cows with mastitis showed a significant imbalance of gut microbiota in milk. Mastitis significantly increased the relative abundances of Enterobacteriaceae (cows with mastitis, 0.546 ± 0.075; healthy cows, 0.027 ± 0.009), Streptococcus (cows with mastitis, 0.184 ± 0.041; healthy cows, 0.005 ± 0.004), and Mycoplasma (cows with mastitis, 0.072 ± 0.079; healthy cows, 0.005 ± 0.037), which was significantly lower than that of the normal bacteria in milk ( Figures 5E ), such as Ruminococcaceae (cows with mastitis, 0.006 ± 0.002; healthy cows, 0.085 ± 0.043), Clostridiales (cows with mastitis, 0.007 ± 0.004; healthy cows, 0.048 ± 0.014), Bacillales (cows with mastitis, 0.001 ± 0.000; healthy cows, 0.031 ± 0.022), and Clostridium (cows with mastitis,0.002 ± 0.001; healthy cows, 0.045 ± 0.012). However, no significant change in the relative abundance of predominant phylum (top 10) and genus (top 20) was observed between the gut microbiota of cows with mastitis and healthy cows ( Figures 6E, F ).

On the 3^rd, 5^th, and 14^th days after antibiotic treatment, the Shannon index of milk microbiota in the cows with mastitis was significantly higher (p < 0.05) than that of before treatment ( Figure 7A ), and the number of observed ASVs was significantly higher (p < 0.05) than that of before treatment on the 5^th day after antibiotic treatment ( Figure 7C ). Meanwhile, the Shannon index of milk microbiota in the cows with mastitis significantly increased (p < 0.05) on the 5^th and 14^th days after geraniol treatment ( Figure 7B ). Additionally, the number of observed ASVs was significantly higher (p < 0.05) on the 5^th and 14^th days after geraniol treatment than that of before treatment ( Figure 7D ). The antibiotic treatment group and geraniol treatment group showed significant recovery in the community structure, which was consistent with the normal milk microbiota ( Figures 7E, F ). Notably, antibiotics showed a faster recovery of milk microbiota community structure in cows with mastitis than geraniol. After three days of antibiotic treatment, the microbial community in the milk of dairy cows with mastitis was closer to the healthy dairy cows, while geraniol treatment took 14 days to restore the milk microbiota( Figures 7E, F ).

The relative abundance of the pathogenic bacteria (Enterobacteriaceae, Streptococcus, and Mycoplasma) in milk significantly decreased following the treatment with antibiotics and geraniol (p < 0.05) ( Figures 8A, B ). In the antibiotic treatment group, the relative abundance of Enterobacteriaceae, Streptococcus, and Mycoplasma decreased to 0.026 ± 0.006, 0.003 ± 0.001, and 0.06 ± 0.005, respectively, on the 14^th day after treatment ( Figure 8A ), while it decreased to 0.04 ± 0.01, 0.004 ± 0.002, and 0.02 ± 0.01, respectively, in the geraniol treatment group ( Figure 8B ). The relative abundance of probiotics, including Lactobacillus and Bifidobacterium, in the milk of cows with mastitis significantly increased (p < 0.05) with geraniol treatment (3^rd and 5^th days; Figures 8D, F ); however, the relative abundance of these probiotics did not increase significantly in the antibiotic treatment group ( Figures 8C, E ).

Similarly, the quantitative PCR analysis results were consistent with the 16S rRNA sequencing results showing similar changing patterns for the number of pathogens ( Figure S6 ) and probiotics ( Figure S7 ) in the cows’ milk in the geraniol treatment group and antibiotic treatment group.

After antibiotic injection, the Shannon index and the number of observed ASVs of gut microbiota in the cows with mastitis descreated, showing no recovery until the 14^th day (10 days after discontinuing antibiotics) ( Figures 9A, B ). Conversely, the Shannon index ( Figure 9C ) and the number of observed ASVs ( Figure 9D ) increased steadily after treatment with geraniol. On the 14^th day (10^th day after stopping geraniol) after geraniol treatment, the Shannon index and the number of observed ASVs of gut microbiota in the cows with clinical mastitis were significantly higher (p <0.05) than before treatment.

Furthermore, the PCoA based on weighted Unifrac distance showed that antibiotic treatment significantly destroyed the gut microbiota structure of cows with mastitis ( Figure 9E ). On the 10^th day after discontinuing antibiotic treatment (14^th day), the gut microbiota structure showed the largest variation than that of before treatment. The phylogenetic distance between the gut microbiota after (5^th/14^th day) and before treatment was significantly larger (p < 0.05) than the samples before treatment ( Figures 9F ). However, geraniol did not destroy the gut microbiota community structure in the cows with mastitis during treatment ( Figure 9G ). Compared with the samples before treatment (the 0^th day), the phylogenetic distance in the gut microbiota of cows with mastitis show no significant increase between the 3^th/5^th/14^th days after receiving geraniol treatment and before treatment (p > 0.05) ( Figure 9H )

A total of 140 milk samples were collected from 20 cured cows (each cow was sampled once a day for seven consecutive days) of the geraniol treatment group (n = 10) and antibiotic treatment group (n = 10) to detect the drug residues in milk. Cephalexin and kanamycin remained in the milk of cured cows on the 7^th day after cephalexin and kanamycin injection ( Table 1 ). Meanwhile, the quantitative analysis of the antibiotic residues indicated that metabolism could only reduce the residue dose of antibiotics by 0.1 to 0.4 times daily, without any reduction in cephalexin and kanamycin residues in the milk of five cows within two or three days. No geraniol residue was detected in the cows’ milk on the 4^th day after stopping geraniol treatment ( Table 2 ). Furthermore, the quantitative analysis of the drug residues showed that the geraniol residues decreased 6 times day by day after stopping geraniol.

The reduction in cefalexin and kanamycin injection dose did not change its residual time in the dairy cows. These two antibiotics were detected in the milk of cured cows on the 10th day after injecting a low dose (2 g) of cephalexin and kanamycin into the cow’s breast (Table S8).

The initial MIC values of geraniol and amoxicillin for the Escherichia coli strain ATCC25922 were 33.6 µg/mL and 4 µg/mL, respectively. The initial MIC values of geraniol and ampicillin for the Staphylococcus aureus strain ATCC25923 were 33.6 µg/mL and 0.5 µg/mL, respectively. The MIC value of geraniol for Escherichia coli and Staphylococcus aureus did not change after being induced by sub-inhibitory concentration for 150 generations ( Figures 10A, B ). The MIC value of amoxicillin for Escherichia coli ATCC25922 increased eight times (32 µg/mL) by the 10^th generation, 256 times (1024 µg/mL) by the 20^th generation, and 512 times (2048 µg/mL) by the 40^th to 150th generations ( Figure 10A ), and the MIC value of ampicillin for Staphylococcus aureus increased two times (1.0 µg/mL) by the 10th generation, 4 times (2.0 µg/mL) by the 80th generations, 8 times (4.0 µg/mL) by the 90th generations, 16 times (8.0 µg/mL) by the 110th generations, 32 times (16.0 µg/mL) by the 130th generations, and 62 times (32.0 µg/mL) by the 150th generation ( Figure 10B ). This result suggested that prolonged use of geraniol might not induce bacterial resistance.