Aedes aegypti aegypti (Liverpool strain) were maintained in a colony held in an environmental chamber (25.0 ± 0.5°C, 65.0 ± 5.0% RH, and a photoperiod (L:D) of 12:12 h) at the Forensic Laboratory for Investigative Entomological Sciences (F.L.I.E.S. Facility) at Texas A&M University, College Station, Texas, USA. Mosquito larvae (~l,000) were reared in enamel pans (25 × 35 × 5 cm) containing 1.5 L of reverse osmosis (RO) water. Larvae were fed a diet of fish food, TetraMin (Tetra, Virginia, USA), on a standardized mosquito-rearing schedule (Gerberg et al., 1994). Pupae were collected daily and placed in a 50-ml cup of RO water at a density of 100 larvae/cup. Containers were divided into groups of three and placed into 30.5 × 30.5 × 30.5 cm aluminum screened wire mesh cages (BioQuip Products Inc., California, USA) for adult eclosion. Emergent adults were provided ad libitum with a 10% sucrose solution placed on absorbent cotton rolled in cotton-muslin gauze cloth and inserted in a 50-ml glass bottle placed inside each adult cage. Blood-feeding of 3–5-day-old (post-emergence) female mosquitoes was performed using a 1 ml aliquot of defibrinated rabbit blood (HemoStat Laboratories, California, USA) in an artificial membrane. At 48 h after blood feeding, a 2 × 5 cm filter paper placed in a 50-ml black cup containing 30 ml of RO water was provided as an oviposition site in each cage. Females were allowed to deposit eggs in the container for 3 days. Filter papers containing eggs were removed from the container and placed on a shelf in the incubator room, allowed to air dry, and then stored at room temperature until use.

Staphylococcus epidermidis (1457) strain (Kies et al., 2003) was grown on mannitol salt agar (MSA; Neogen Corp., Michigan, USA) at 37°C for 48 h, isolated onto a blood agar plate, and then incubated overnight at 37°C. For use in the mosquito behavior experiment, an inoculum of 10⁸ cfu (colony-forming units)/ml (2.1 ± 0.6 × 10⁸ cfu/ml) in phosphate-buffered saline (PBS) was used. Preliminary experiments showed no mosquito interactions induced with the PBS diluent used for S. epidermidis stock (Supplementary Figure 1). For QSI, 10 mg (0.01 g) of brominated furanone C-30 ((Z-)-4-Bromo-5-(bromomethylene)−2(5H)-furanone, C5H2Br2O2; Sigma-Aldrich Corp., Missouri, USA) (Chemical FW = 253.88 g/mole) was dissolved in 78.77 ml of methanol. Preliminary experiments showed no mosquito interactions induced by the methanol diluent used for the 500 μM furanone C-30 stock solution (Supplementary Figure 2).

The experiments were a modification of previously described methods (Zhang et al., 2015). At 2 h prior to each trial, 50 mated female mosquitoes (3–5 d-old post-emergence) that had never been offered a blood meal were collected using a battery-powered aspirator (Hausherrs Machine Works Co., New Jersey, USA). Mosquitoes were released into a clear Plexiglas^® cage (82 × 52 × 45 cm) with a wire mesh top and allowed to acclimate at a temperature of 25.0 ± 0.5°C and relative humidity of 65.0 ± 5.0%. Experiments were performed 30 min after sunrise (chamber at 12:12 L:D), which corresponded to the normal biting activity of Ae. aegypti (Yasuno and Tonn, 1970).

Blood feeders were individually constructed from a 25-ml sterile tissue culture flask (Corning Inc., New York, USA) tightly wrapped with parafilm and secured with cellophane tape. A 1-ml aliquot of defibrinated rabbit blood (HemoStat Laboratories, California, USA) was injected into the space between the culture flask and parafilm. A 5.0 × 5.0 cm piece of sterilized 100% cotton gauze (Dynarex Co., South Carolina, USA) was placed to absorb the inoculum over the parafilm and secured with two autoclaved rubber bands.

Dual choice assays were performed with two blood feeders placed at equal distances horizontally and vertically apart with the cotton gauze inoculated with 1 ml of (1) H₂0 (reverse osmosis water) or PBS (phosphate-buffered saline), (2) with or without MeOH (methanol), (3) with or without 50 μM/ml QSI, and (4) with or without 10⁷ cfu/ml S. epidermidis and 50 μM/ml QSI side-by-side but separated by a parafilm dam to prevent mixing or physical contact (NSE-QSI). Each treatment solution was applied directly to the gauze 15 min before the experiment. For each trial, three or four replicates were performed in succession by rotating each of the two treatments to prevent positional bias.

For quadruple choice assays, four blood feeders were placed at equal distances horizontally and vertically (24 cm) apart in a square pattern, with the gauze side down on the wire mesh top while connected to a water bath (Thermo Fisher Scientific, Connecticut, USA), and maintained at 37°C (Zhang et al., 2015). Cotton gauze assigned to each replicate on a blood feeder was inoculated with 1 ml of either 10⁷ cfu/ml S. epidermidis (SE), 50 μM/ml QSI (QSI), 10⁷ cfu/ml S. epidermidis + 50 μM/ml QSI (SE+QSI), or only PBS (CONT) as a control. Each treatment solution (i.e., QSI and bacteria) was mixed in a 1.5-ml microfuge tube 15 min before application to the gauze. The effect of 10⁷ cfu/ml S. epidermidis + 50 μM/ml QSI that was neither mixed nor in physical contact (NSE-QSI), QSI alone, or solvent controls (PBS for S. epidermidis and methanol for QSI) on mosquito responses was also measured using a dual choice assay (methods provided in Supplementary material). For each trial, four replicates were performed in succession by rotating each of the four treatments to each of the four different corner locations initially assigned by a random number generator and rotated clockwise across trials to prevent positional bias (n = 12). The Plexiglas^® cages were cleaned with 3% Lysol and 95% ethanol and allowed to dry and air out between trials. During the experiments, mosquito landing and probing activity at each blood feeder was recorded on video, with no one present in the room during the trial, with cameras (2160 p/30 fps, LG, Korea) mounted on the outside of the cage. The number of ladings and probings of duration at each blood feeder (response) longer than 1 s were recorded over a 15-min assay period. A particular mosquito feeding blood on one of the treatments was continually counted at each second, which reduced a frequency variation biased in the total number of mosquitoes. This matrix, modified from the human landing catch (HLC) technique as a gold standard tool for mosquito host-seeking (landing) behavior (Shirai et al., 2002), engaged in sequential behaviors, including piercing and blood feeding on hosts, which provided a direct and sensitive estimate of mosquito behavioral responses and additionally intraspecific interaction over a given period.

The normality of the data was determined for the number of landings and probings by mosquitoes over a 15-min assay period as a function of treatment using the Shapiro–Wilk test (P ≤ 0.05 rejected normal distribution). All statistical procedures were conducted using JMP Statistics, Version 15.0 (SAS Institute Inc., North Carolina, USA). The alpha was set at 0.05 for all statistical tests. The odds ratios of choosing a particular treatment applied in a blood feeder were tested using R version 3.4.3 and used the DescTools package (https://cran.r-project.org/web/packages/DescTools/index.html). Generalized linear mixed models (GLMM) were performed using R version 3.4.3, with response variables being the number of landings and probings by Ae. aegypti with Poisson family distribution to the different treatments used as fixed effects, experimental groups (trials) as a random effect, and time as a covariate. We considered a repeated measures design with an autoregressive (1) covariance structure. The models were validated through the exploration of residual errors with graphical tools and overdispersion access (Zuur et al., 2010; Hartig, 2017). Nonparametric tests with either Wilcoxon's test or the Kruskal–Wallis test followed by the Steel-Dwass test were used to measure differences in mosquito behavioral responses, grouped by replicates, among treatments.

Bacterial volatiles were analyzed in triplicate experiments from the following samples in PBS as a diluent: (1) 0.5 ml of 10⁷ cfu/ml S. epidermidis; (2) each 0.5 ml of 50 μM/ml QSI plus 10⁷ cfu/ml S. epidermidis; (3) 0.5 ml of 50 μM/ml QSI; and (4) 0.5 ml of PBS. The protocol was modified by Groenhagen et al. (2013) and designed to improve filtration in the incoming air. Each sample was transferred to a 12-ml amber glass bottle, and VOCs were collected by the closed-loop-stripping-analysis (CLSA) technique at room temperature. Before every headspace sampling from treatment, the apparatus was thoroughly cleaned with dichloromethane (CH₂Cl₂) and autoclaved at 121°C for 15 min. Each amble glass bottle was placed in a 7.5 × 11-cm (O.D × H) glass filtering jar (Kimble Chase, New Jersey, USA) with a flat ground glass and sealed with a parafilm. The rubber stopper on the top of the glass filtering jar was equipped with one hole and inserted with a volatile trap packed with ~30.0 mg of Hayesep^® Q porous polymer (Volatile Assay Systems, New York, USA), connecting a vacuum pump (Rocker, Scientific Co., Ltd., New Taipei City, Taiwan) with Tygon^® tubing (Saint-Gobain S.A., Pennsylvania, USA). The tooled hose was connected with 3 cm of Tygon^® tubing piece inserted with a bacterial filter (Midwest Supplies, Minnesota, USA, 0.2 μm pore size) and a 14.6-cm carbon-filtered pipet (Marineland, Ohio, USA) to purify incoming air. Samples from each treatment were obtained by running the apparatus at 1 L min⁻¹ for 1 h. Samples were added with an additional 5.0 μl of 80 ng/μl n-octane (Sigma-Aldrich, MO, USA) as an internal standard and stored at −20°C until analysis.

GC-MS analyses were carried out on an Agilent 6890 Gas Chromatograph with an Agilent Technologies 5973N Mass Selective Detector (Agilent Technologies, California, USA) by the Environmental Research Group at Texas A&M University in College Station, Texas. The GC was programmed as follows: 5 min at 50°C, increasing at 5°C/min⁻¹ to 320°C, and operated in split/splitless mode: 60 s at 250°C. A carrier gas, helium, was used at 1.2 ml min⁻¹. Candidate identification of compounds was made by matching the comparison of mass spectra with the mass spectra fragmentation patterns in the National Institute of Standards and Technology (NIST) 05 mass spectra library for peaks observed in the chromatograms.

The GC-MS data were processed to estimate the percentage of the area of each compound in every sample across treatments, including control. To determine the difference among volatile profiles, permutational multivariate analysis of variance (PERMANOVA) was tested using the Adonis function in R version 3.4.3, vegan package (http://CRAN.R-project.org/package=vegan). VOC profiles were analyzed using non-metric multidimensional scaling (NMDS) based on the Bray–Curtis distance matrix to minimize the complex data of the area percentages in a two-dimensional space. An indicator species analysis was conducted to identify the compounds as influential species that may or may not be related to each group. The reliability of stress values was set at < 0.2. Compound abundance was also compared using a two-way ANOVA using JMP^® statistical software version 13 (SAS Institute Inc., North Carolina, USA) and the Tukey-Kramer Honestly Significant Difference (HSD). Significant levels were set at a P-value of ≤ 0.05.

Staphylococcus epidermidis was grown as described above. Three replicates of 3 mL frozen (−80°C) 10⁷ cfu/ml bacterial suspensions in 30% glycerol in tryptic soy broth per treatment (S. epidermidis exposed to furanone: SE + fur1, SE + fur2, and SE + fur3) or controls in diluent only (S. epidermidis wildtype alone: SEwt1, SEwt2, and SEwt3) were centrifuged at 7,000 × g at 4°C for 5 min. The culture supernatant was removed, and total RNA was isolated from the bacterial pellet using Trizol^® (Thermo Fisher Scientific, Massachusetts, USA), following the manufacturer's instructions. Following this, RNA was treated with Turbo DNAse (Thermo Fisher Scientific, Massachusetts, USA) according to the manufacturer's instructions to remove trace DNA. RNA quality was analyzed by agarose gel electrophoresis, and RNA concentrations were determined using Qubit 2.0 (Thermo Fisher Scientific, Massachusetts, USA). All samples were stored at −80°C until further processing for library preparation. Total RNA libraries were created using the NEBNext^® Ultra™ RNA Library Prep Kit and NEBNext^® Multiplex Oligos (Dual Index Primers) (New England Biolabs, Massachusetts, USA) for Illumina^® (Illumina, Inc., California, USA) and associated protocols. High-throughput RNA sequencing was performed by St. Jude Children's Research Hospital on an Illumina HiSeq2000 with 2 × 100 bp PE (paired-end) read lengths.

Sequences were initially trimmed with the sequencing facility using TrimGlare v0.4.2 (Krueger, 2015), but a more stringent quality trimming was also performed using default parameters within the Qiagen CLC Workbench 12.0 (https://www.qiagenbioinformatics.com/) following the QC analysis of sequence reads. The results from high-quality reads were aligned to the S. epidermidis (SE) 1457 genome (downloaded from the NCBI database using accession numbers CP020462 and CP020463 corresponding to the S. epidermidis genome and plasmid, respectively).

RNASeq data were mapped with the following parameters: (a) maximum number of allowed mismatches was set at 2, with insertions and deletions set at 3; (b) length and similarity fractions were set to 0.8, with autodetection for both strands; and (c) a minimum number of hits per read was set to 10. Gene expression values were reported as reads per kilobase of transcript per million (RPKM) mapped reads. Treatment reads with an absolute fold change of 1.5 and a p-value of ≤ 0.5 were considered significant. Following this, transcripts were further annotated into pathways by linking protein ID with potential conserved domains and protein classifications archived within the Conserved Domain Database (Marchler-Bauer et al., 2017) and by using the KEGG and STRING databases (Jensen et al., 2008; Kanehisa et al., 2016). A heat map of expression values was created in the CLC workbench, measuring Euclidean distance with average cluster linkage. Gene expression was filtered with t-test statistics using Bonferonni corrected p-values with the minimum absolute fold change set at 1.5 and the P-value set at ≤ 0.05 for significant expression between S. epidermidis wild-type and QSI treatments.

Preliminary dual choice assays determined that the mosquito responses to blood feeders showed no significant difference (P > 0.7575) in responses to blood feeders with H₂O or PBS over the 15-min period (Supplementary Figure 1), indicating no attraction to the diluent PBS. There was no significant difference between mosquito responses to the blood feeders treated with or without MeOH (P > 0.3856), indicating no bias when using MeOH as a QSI solvent (Supplementary Figure 2). While initial dual choice results confirmed no bias in mosquito responses to solvent controls MeOH or PBS, a response was elicited in blood-feeders with QSI alone or NSE–QSI (not mixed and not in physical contact) (Supplementary Figures 3, 4).

In a comparison between blood feeders with or without QSI, a significant difference (P ≤ 0.0001) was found in mosquito responses over time, as the total number of responses to blood feeders treated with QSI was 51.7% higher than those to a blood feeder alone (Supplementary Figure 3) indicating some attraction of mosquitoes to QSI over a blood feeder alone. Bacterial VOCs mix through indirect gas diffusion by co-cultivation, without physical contact, of the bacteria that produce them. This can result in new product formation or differential functionality, such as upregulation or downregulation of antimicrobial activity (Singh, 2011; Tyc et al., 2015; Lammers et al., 2022). Therefore, mosquito interactions with blood feeders treated with or without SE and QSI on the same feeder but not mixed (NSE-QSI) were evaluated. The mosquito responses to these NSE-QSI feeders were significantly greater over the 15-min period (P ≤ 0.0443) than those to blood feeders alone. The total number of attraction responses [1,449 from the three experiments of triplicate trials (n = 9)] to blood-feeders treated with or without NSE-QSI were 800 (55.2%) and 649 (44.8%), respectively (Supplementary Figure 4). Blood-feeders with NSE-QSI elicited a greater response (23.2% in the total number of responses) at all time points compared to the blood-feeder solvent controls, indicating some attraction of mosquitoes to NSE-QSI. This could be owing to the attraction of the mosquitoes to the SE, as observed in the previous pairwise experiment, the QSI, or to both. The mechanism of action of the halogenated furanone C-30 is due to its similar structure to that of quorum sensing autoinducers, allowing it to bind to the QS response regulator while failing to activate it, resulting in inhibition of the QS system (Markus et al., 2021). It is unclear whether this would occur through gas diffusion, but the results of this experiment do not support that conclusion under these specific experimental parameters. The pairwise assays offered limited choices for the mosquitoes that may have skewed the attraction when no other options were presented but allowed us to exclude concerns of possible repellency as these results showed no evidence for QSI-induced mosquito repellency. However, the results required further delineation of mosquito attraction to SE vs. QSI-treated SE; therefore, quadruple choice assays were conducted to parse out the stronger attraction behavior when a wider variety of choices were offered simultaneously.

By determining the odds ratio of mosquito attraction to blood feeders treated with fully functional S. epidermidis and S. epidermidis with QS inhibited by brominated furanone C-30, we aimed to capture behavioral alterations that are directly linked to bacterial QS functionality. The analysis indicated the mosquito response to the blood-feeders treated with S. epidermidis (3.58) was significantly greater than the treatment of S. epidermidis + QSI (quorum sensing inhibitor) (1.24) or QSI alone (0.89) relative to the PBS control (Figure 1). While treatments (S. epidermidis, S. epidermidis + QSI, QSI, and PBS) (P ≤ 0.0001) and time (P ≤ 0.0071) significantly impacted mosquito response, no significant trial and time interactions (P ≤ 0.3398) were measured. In general, mosquitoes spent 2.36-, 2.83-, and 2.55-fold more time on the blood feeders treated with S. epidermidis than with S. epidermidis + QSI, QSI, or PBS, respectively, at every time point over the 15-min experimental period (Figure 2). Overall, S. epidermidis alone accounted for 46.1% of the total responses, and the inhibition of QS with the addition of QSI to S. epidermidis accounted for 19.6% of the recorded responses or 55.1% fewer responses than S. epidermidis alone (Figure 3). The QSI compound alone was equally as attractive as S. epidermidis + QSI, as well as the PBS control, indicating that the QSI was not repellent. No significant difference was measured between S. epidermidis + QSI, QSI, and PBS over time, although mosquito responses to S. epidermidis + QSI accounted for 20.1%, which was 8.5% greater than QSI or PBS individually. The results of this assay directly linked QS functionality to a change in mosquito behavior, thereby providing support for the hypothesis that prokaryote QS mechanisms are monitored and interpreted by eukaryotes.

Mosquitoes use olfactory cues in part to interpret their surroundings (Davis and Sokolove, 1975; Gillies, 1980; Bentley and Day, 1989; Lu et al., 2007), and bacteria produce a multitude of volatiles (Schulz and Dickschat, 2007). The previous assay determined that the application of a QSI to S. epidermidis significantly reduced mosquito attraction to a blood meal. By determining the production of VOCs from fully functional S. epidermidis and S. epidermidis inhibited by QSI, alterations in volatile production directly linked to bacterial QS functionality were captured. A total of 26 compounds were identified by comparing experimental mass spectra with the NIST14 Mass Spectral Library from the headspace volatiles among samples in PBS as a diluent: (1) 10⁷ cfu/ml S. epidermidis; (2) 50 μM/ml QSI plus 10⁷ cfu/ml S. epidermidis; (3) 50 μM/ml QSI; and (4) PBS (Table 1). Excluding octane, which was added as an internal standard, 19 compounds were detected from S. epidermidis, of which 13 were from S. epidermidis inhibited by QSI. Staphylococcus epidermidis had a mean of 16.00 ± 1.53 compounds compared to 12.00 ± 1.00 with QSI application. The mean number of compounds detected from each of the QSI and control was 13.33 ± 0.33 and 13.67 ± 0.88, respectively. Based on the relative frequency and abundance across treatments and excluding octane, nine VOCs were shared by all treatments: furfural; benzene, 1,3-dimethyl; benzaldehyde; phenol; nonanal; benzothiazole; 2,5 cyclohexadiene; butylated hydroxytoluene; and diethyl phthalate. A total of 19 total compounds were identified from S. epidermidis, 10 of which were common to the PBS control. Major changes (>10% difference) in the relative abundance of VOC from SE once exposed to QSI were measured (Table 1), demonstrating a decrease in the proportion of furfural (−45%), benzaldehyde (−39%), phenol (−54%), and butylated hydroxytoluene (−80%) and an increase in the proportion of 1-heptene,4-methyl (24%), and 2,5 cyclohexadiene (40%). Of the compounds not common to PBS, 8 were completely inhibited by the addition of QSI to S. epidermidis: 1-heptene,4-methyl; 4,7-methano-1H-indene; pregnane-3d; morphine; octacosane; heneicosane; heptadecane 9-octyl; tetracosane; and lanosta.

The eight volatiles from S. epidermidis inhibited by treatment with QSI have a variety of functions. For example, 4,7-methano-1H-indene (4,7-methanoindene) has been used in plasticizers, as an ingredient in synthetic waxes and resins and perfume materials or as an intermediate for drugs and insecticides. Novel 4,7-methanoindene derivatives substituted with esters or acids have been patented as useful perfume ingredients (US3557188A United States) and described as having a pleasant odor (Dunkel, 1971). Pregnane-3 (pregnane-3 20-dione) is a steroid hydrocarbon in a form that can be modified to produce several GABA-modulating hormones, including progesterone and other steroid forms capable of urinary excretion. Morphine, an opioid, is known for its analgesic properties. Endogenous opioids can, through the involvement of quorum sensing circuitry, enhance virulence pathways in bacterial pathogens (Zaborina et al., 2007). Heneicosane is a bioactive compound that has been isolated from plants (e.g., Periploca laevigata. Labill., Plumbago zeylanica, L.) and exhibits antimicrobial capabilities against fungus and gram-positive and negative pathogenic bacteria (Vanitha et al., 2020). It is also an insect oviposition pheromone that can attract or repel gravid female mosquitoes dependent on its concentration (Seenivasagan et al., 2009). Octacosane, heptadecane 9-octyl (9-octylheptadecane), and tetracosane are all alkane hydrocarbons found in plants. Octacosane is a constituent of wood oils that has antimicrobial and cytotoxic activities (Martins et al., 2015). Heptadecane 9-octyl has the potential to be used as an antifungal agent (Abubacker and Devi, 2014). Tetracosane is also a component of the sex pheromone bouquet of the female mining bee [Andrena nigroaenea (Kirby) (Francke and Schulz, 2010)]. Lanostane [20.xi.-lanosta-7, 9(11)-diene-3.beta.,18, 20-triol], a triterpenoid, is a polycyclic hydrocarbon. Triterpenoids are common in fungi, marine organisms, and higher plants and participate in the environmental defense mechanisms of the organism. Lanostane specifically has anti-inflammatory and anti-peroxidative properties (Ríos et al., 2000). Therefore, many of these compounds have already been described by their ecological olfactory influences, which, when disrupted by QSI treatment, would presumably affect those appraising their environment through these volatile occurrences.

The relative abundance and quantity range percentages for each treatment are summarized in Table 2. VOCs have been organized by retention times. The differential VOC profiles across treatments were statistically determined by ANOSIM (R = 0.4134, P ≤ 0.037). The stress value representing the accuracy in spatial similarity/dissimilarity was 0.1046. The results of this assay delineated the possible bacterial volatile compounds elicited by QS activity that could be interpreted by mosquitoes, thus triggering specific behaviors.

The messenger RNA molecules expressed by bacterial genes ultimately lead to the production of proteins that influence cellular processes such as volatile production. By analyzing the bacterial transcriptome of fully functional S. epidermidis and S. epidermidis treated with QSI, we aimed to elucidate which cellular processes are directly linked to bacterial QS functionality. Triplicate samples of Staphylococcus epidermidis exposed to furanone (QSI), SE + fur1, SE + fur2, and SE + fur3, and diluent controls not exposed to furanone (S. epidermidis wildtype alone, SEwt1, SEwt2, and SEwt3, were processed to elucidate their mRNA expression. Sequencing and trimming yielded an average fragment and read length of 132 and 1,249,203, respectively. Altogether, 65 genes were differentially regulated between S. epidermidis+ QSI and S. epidermidis. Complete gene lists are shown in Supplementary Table 1. A total of 29 genes were significantly upregulated when S. epidermidis was treated with a QSI. A heat map of the top 59 significantly expressed genes from S. epidermidis with and without QSI is shown in Figure 4. Of the significantly upregulated genes, four were involved in environmental information processing and included gene encoding for bacterial secretion, lipoprotein export, a hexose-6-phosphate phosphate antiporter, and membrane transport. Seven were involved in genetic information processing, including translation, chaperones, folding catalysts, replication, repair, and (d)NTP-pool sanitation involving gyrB, dnaB, ung, and a gene encoding a putative YabN. Twelve were found to be involved in metabolism, including riboflavin metabolism, carbohydrate metabolism, nitrogen or urea metabolism, glycan biosynthesis and metabolism, and lipid metabolism. Moreover, four of these 12 were genes encoding for amino acid transport and metabolism. In addition, three significantly upregulated genes were classified as participating in signaling and cellular processes, including genes encoding a multidrug efflux transporter, a nucleoside transporter, and a gene involved in cell wall metabolism. Finally, one gene from the S. epidermidis plasmid p1457, whose function is currently unknown, was upregulated.

A total of 36 genes were significantly downregulated when S. epidermidis was treated with a QSI. Five genes were associated with environmental information processing pathways, all of which are involved in response to various stresses, such as general stress response or alkaline shock. In addition, 10 were involved in genetic information processing, with one involved in translation, one in the biosynthesis of the modified nucleoside queuosine tRNA, and three in a stress response. Involvement in stress response included per, a gene peroxide-responsive transcriptional repressor; spxA, a transcription factor that may function to reduce growth and development processes during periods of stress; and clpB, involved in cell recovery from heat, oxidative, and other stress. Moreover, dps, a stationary phase nucleoid protein that sequesters iron and protects DNA from damage, was downregulated.

Then, five significantly downregulated genes were associated with metabolism, including three for energy metabolism, one for riboflavin metabolism, and one for encoding a protein arginine kinase associated with general metabolism. In addition, six significantly downregulated genes were associated with signaling and cellular processes, including one encoding a serine protease, one encoding a zinc metallopeptidase, one encoding an M50 family peptidase, and one encoding a RidA family reactive intermediate/imine deaminase. Genes also downregulated and associated with signaling and cellular processes included F0F1 ATP synthase subunit alpha, believed to be involved in cell motility, intracellular trafficking, secretion, vesicular transport, and spoVG, involved in the regulation of cell wall metabolism that plays a role in sporulation and other functions in other organisms. Moreover, we identified two genes encoding two hypothetical proteins and three downregulated genes that were not associated with any of the above pathways, including a gene encoding a predicted lipoprotein, one encoding a putative gas vesicle protein, and one encoding a transmembrane protein. The results of this assay delineated the cellular processes within the bacteria that control the volatile compound production altered by QSI activity and that ultimately led to changes in mosquito behavior.