Bed bugs
Four bed bug populations (one laboratory strain and three collected from infested homes) were used in this study (Table 1). All populations were reared in the laboratory as described by DeVries et al.28. Briefly, bed bugs were maintained in 168 cm3 plastic containers on paper substrate at 25 °C, 50% relative humidity, and a photoperiod of 12 h:12 h (Light:Dark). Bed bugs were fed defibrinated rabbit blood (Hemostat Laboratories, Dixon, CA, USA) weekly using an artificial feeding system. This system maintained blood at 35 °C by circulating water through custom-made water-jacketed glass feeders. An artificial membrane (plant budding tape, A.M. Leonared, Piqua, OH, USA) was stretched over the bottom of each glass feeder, containing the blood while simultaneously allowing bed bugs to feed through it. In all experiments, adult males starved for 7–10 days were used. All populations were used for documenting responses to human skin swabs. The WS population was used for bioassays with various human volunteers and hexane extracted swabs, and the JC population was used for testing various lipids.
Skin swab collection
The North Carolina State University Institutional Review Board approved this study (IRB #14173). Informed consent was obtained from all human participants, and all the methods were performed according to the relevant guidelines and regulations. Six human volunteers (3 males, 3 females) ranging from 25 to 50 years old representing several ethnicities (white/Caucasian, Hispanic, Asian) provided samples for this project. Skin swabs were collected following the exact methods outlined by DeVries et al.16. In our 2019 study, these swabs were reported to attract bed bugs independent of other cues in Y-tube olfactometer assays. Briefly, participants were asked to follow a standard operating procedure, which was reviewed with them prior to sample collection. Before collecting skin swabs, participants were asked to not to eat ‘spicy’ food for at least 24 h, take a morning shower, avoid the use of deodorant and cosmetics after showering, and avoid strenuous physical activity. Skin swabs were collected 4–8 h after showering. Hands were washed with water only before lifting filter paper. Swabs were collected using 4.5 cm diameter filter paper discs (#1; Whatman plc, Madistone, United Kingdom). Both sides of a single filter paper disc were rubbed over the left arm from hand to armpit for 12 s, left leg from lower thigh to ankle for 12 s, and left armpit for 6 s. This procedure was repeated on the right side using a new filter paper disc, so that two samples were collected during each swabbing session. The skin swab samples were then stored in glass vials at − 20 °C, and used within one month of collection. The swabs from all human volunteers were used to compare participants and establish that bed bugs responded similarly to all, and participant A’s skin swabs were used for all subsequent bioassays.
Two-choice arrestment bioassays
Two-choice bioassays were conducted in plastic Petri dishes of 6 cm diameter (Corning Life Sciences, Durham, NC, USA) (Fig. 1). The bottom surface of each Petri dish was roughened so that bed bugs could freely move about the arena. Two tents (3 × 1.5 cm) were created using filter paper (Whatman #1). One tent served as the control tent, and the other served as the treatment tent. Control tents were either untreated (nothing added) or treated with hexane only. Treatment tents were either made directly from human odor swabs, treated with human odor extract (in hexane), or treated with a specific compound (in hexane). Tents were allowed 60 min to acclimate to room conditions and allow for the solvent to evaporate prior to initiating bioassays. The positions of tents (treatment and control) were alternated to account for any side-bias.
Two-choice behavioral assay (top-view) consisting of two equal size paper shelter tents. A clean filter paper (control) was always paired with a treated filter paper that either represented a human skin swab, hexane extract of swabbed paper, SPE fraction of human skin swab extract, or authentic TAGs. A single male bed bug was introduced into the center of each arena and allowed to select a tent to arrest under.
Adult male bed bugs were housed in individual vials for 24 h prior to each experiment. A single adult male bed bug was released in the middle of the arena 5 h into the scotophase, by transferring it on its harborage. The harborage material was removed immediately after the bed bug moved off of it (the harborage). Bed bugs were allowed the remaining 7 h of the scotophase to freely move around the arena, with their final position reported 3 h into the photophase. Bed bugs that were in contact with the filter paper with any part of their body were recorded as making a choice (i.e. arrestment state); others not in contact with either filter paper tent were recorded as non-responders, reported in the figures, but not used in data analysis. It should be noted that momentary pauses in movement (feeding or other behaviors) are not referred to as arrestment in this study. In total, 15–39 replicates were performed for each experiment (reported for each bioassay).
Bioassays with human skin swabs
Bioassays with human skin swabs were performed to understand if bed bug arrestment behavior (1) differed among different bed bug populations, and (2) influenced by different host odors. Skin swabs were removed from the freezer, equilibrated to room temperature, divided into three equal parts and trimmed to a rectangular shape corresponding to the size of a shelter tent (Fig. 1). Skin swabs from participant A were used to evaluate the responses of four bed bug populations (Table 1). Skin swabs from all participants A–F were used to evaluate the robustness of our findings across multiple human hosts.
Skin swab extraction and fractionation
Skin swabs collected from volunteer A were pooled and extracted in hexane. Extraction procedures were carried out sequentially by placing a single skin swab into a 20 ml glass vial containing 5 ml of hexane, vortexing for 30 s, then moving the filter paper to a new 20 ml vial containing 5 ml of hexane and repeating the process. Three sequential extractions were performed for each skin swab, and a minimum of 10 skin swabs (collected over several days) were used for each extraction. After all skin swabs were extracted, all sequential hexane extracts were combined and concentrated to a final concentration of one skin swab equivalent per 300 µl, or one bioassay equivalent (BE) per 100 µl (since each swab was used for 3 bioassays; see “Bioassays with human skin swabs” for more information on the size used for each bioassay). Control swabs were also extracted. These swabs were treated identically to the skin swabs, except they did not contact human skin.
To determine what compound classes were responsible for the observed behavior, hexane extracts were fractionated using solid phase extraction (SPE). Extracted samples were concentrated to 1 BE/10 µl hexane, then loaded onto a 1 g silica SPE column (6 ml total volume; J.T. Baker, Phillipsburg, NJ, USA). The column was eluted with the following solvents (4 ml of each, each repeated twice sequentially): hexane, 2% ether (in hexane), 5% ether (in hexane), 10% ether (in hexane), 20% ether (in hexane), 50% ether (in hexane), 100% ether, ethyl acetate, and methanol (all solvents acquired from Sigma Aldrich, St. Louis, MO, USA). Each solvent fraction was then concentrated to a final concentration of 1 BE/100 µl and stored at − 20 °C.
Bioassays with extracted and fractionated human skin swabs
For all extraction and fractionation bioassays, filter paper tents were cut to a size of 3 cm × 1.5 cm (Fig. 1) and treated with 100 µl (1 BE) of extracted or fractionated human skin swabs (50 µl on each side). A dose–response bioassay was run first to determine if the compounds responsible for bed bug arrestment responses could be extracted and at what concentration (BE) they were behaviorally active. Dilutions were made in hexane, with control tents receiving extracts of control filter paper. At least 20 replicates were conducted for each concentration. After validating an appropriate BE that could be used in future experiments, SPE fractions were diluted in hexane to 0.1 BE and applied to filter paper tents as previously described (50 µl per side). A minimum of 15 replicates were conducted for each fraction to identify behaviorally active fractions.
Compound identification
To better understand what classes of compounds were present in behaviorally active fractions, we conducted thin layer chromatography (TLC) with known standards. A flexible, silica (250 µm) TLC plate (Whatman) was placed into a glass chamber containing a solvent layer of 1.5 cm. The plate was cleaned twice with acetone, then standards (triglyceride [TAG], wax ester, squalene) and samples (fractions) were each loaded into separate lanes. The plate was developed twice in 10% ether (in hexane), then visualized non-destructively with iodine.
In addition, behaviorally active fractions were further evaluated for their composition with GC–MS and LC–MS. GC–MS was employed to analyze free fatty acids, squalene, and cholesterol29, whereas LC–MS was employed to characterize the intact skin lipids as previously described30. Samples were analyzed with a GC 7890A coupled to the MS 5975 VL analyzer (Agilent Technologies, CA, USA) following derivatization. Briefly, 50 µL of the extract dissolved in isopropanol were dried under nitrogen and derivatized with 100 µL BSTFA containing 1% trimethylchlorosilane (TCMS) in pyridine to generate the trimethylsilyl (TMS) derivatives at 60 °C for 60 min. GC separation was performed with a 30 m × 0.250 mm (i.d.) × 0.25 µm film thickness DB-5MS fused silica column (Agilent). Helium was used as the carrier gas. Samples were acquired in scan mode by means of electron impact (EI) MS.
Liquid-chromatography coupled to the MS analyzer by means of an electrospray interface (ESI) was used to determine abundance and ESI tandem MS of non-volatile lipids as previously described29,30. LC separation was performed with a reverse phase Zorbax SB-C8 column (2.1 × 100 mm, 1.8 μm particle size, Agilent). Data were acquired in the positive ion mode at unit mass resolving power by scanning ions between m/z 100 and 1000 with G6410A series triple quadrupole (QqQ) (Agilent). LC runs and MS spectra were processed with the Mass Hunter software (B.09.00 version).
Bioassays with triglycerides
After determining that TAGs were prominent compounds in bioactive skin swab fractions, commercially available TAGs were evaluated for behavioral activity. Filter paper tents were treated with 100 µl of hexane (50 µl to each side) containing TAG standards. First, tripalmitin (16:0/16:0/16:0) (Sigma-Aldrich) was evaluated in a dose–response fashion (60 µg to 0.6 µg) to determine what level of TAG was appropriate for bioassays. The upper level of testing was set at 60 µg as a conservative estimate of the amount of TAGs bed bugs may be exposed to, based on calculations of our arena size and previous reports of TAGs on human skin and sebum. Specifically, previous reports documented that 1.5 mg of sebum could be passively collected using Sebutape from an area of 4.7 cm230,31. Because TAGs typically constitute 60% of human sebum32, it is reasonable to assume that passive collection of sebum can result in > 190 µg/cm2 of TAGs in a short amount of time (30 min). Our sampling methods involved swabbing rather than passive collection, but our use of 60 µg over a 9 cm2 (two sides of 4.5 cm2) shelter tent (6.67 µg/cm2) is a low-estimate of the amount of TAGs collected (although this was not directly measured in the current study). Other TAGs that we tested at a concentration of 60 µg per 9 cm2 included the saturated TAGs trimyristin (14:0/14:0/14:0) and tristearin (18:0/18:0/18:0) and the unsaturated TAGs triolein (18:1/18:1/18:1), trilinolein (18:2/18:2/18:2), and trilinolenin (18:3/18:3/18:3) (all from Sigma-Aldrich). A minimum of 30 replicates were conducted with each TAG.
Statistical analysis
A Chi-square goodness of fit test was used to compare the responses of bed bugs to control versus treated tents in all two-choice bioassays, with the null hypothesis that if bed bugs do not respond differentially to treated tents they should display a 1:1 preference ratio for both sides of the assay. All tests were conducted in SPSS Version 26 (IBM Corp., Armonk, NY).
Source: Ecology - nature.com
