Animals. Ten-week-old inbred CBA, CD-1 and BALB/с mice, as well as 10-week-old and 21-day-old outbred ICR mice were obtained from the Vivarium of St. Petersburg State University (Russia) and from the Genetic Resources Center of the Federal Research Center Institute of Cytology and Genetics (Novosibirsk, Russia). All the experiments were conducted on male mice; female mice were used exclusively as chemosignal donors. Male mice were grouped (5–6 mice per cage) and female mice were placed in individual cages. Animals were housed under a regular 12 h dark/light cycle with food and water ad libitum. Experiments were carried out in accordance with the animal protocols approved by the St. Petersburg State University Ethics Committee (#131-03-1) and ARRIVE guidelines. Mice were euthanized by cervical dislocation. For the chromosome aberration assay, bone marrow samples were obtained by flushing the marrow from femur bones using a needle and syringe with a glacial acetic acid-ethanol mixture. For the alkaline comet assay, immunohistochemistry and RNA sequencing, the flushed cells were washed with phosphate buffered saline (PBS, Invitrogen, pH 7.4). All the experiments were conducted on CBA mice, except for the organ weight measurements (CD-1 mice), 2,3-DMP effects on DNA damage and fMRI (BALB/c mice), and VNO removal studies (ICR mice). The chromosome aberration assay was also performed on CD-1, BALB/c and ICR mice.
Animal treatments and used stimuli. The pheromones 2,5-DMP and 2,3-DMP were purchased from Sigma-Aldrich (USA). Stimuli were diluted 1:10000 (w:v) in dH2O21. For mouse pheromone exposure, 1.5 mL of the solution was applied onto filter paper balls in perforated plastic capsules (4 cm in diameter), which were placed on the male mice cell grid. The control was dH2O exposure. The direct contact of animals with the solutions was excluded. The exposure was carried out for 0.5, 1, 2, 4, 24, or 48 h or 30 days in various experiments. In the case of 48 h or 30 days of exposure, the capsule with 2,5-DMP solution was replaced with a new one every 24 h in order to compensate for active substance evaporation. Isolated female mice bedding was collected from 10 mature BALB/c females. Female mice were placed in individual cages 10 days prior to the experiment, and 5 days prior to the experiment the bedding was replaced. Immediately before the experiment (fMRI), the bedding was collected and mixed.
To assess the effects of prolonged 2,5-DMP treatment, male mice were exposed to 2,5-DMP, 2,3-DMP or H2O for 30 days. At the end of the treatment period, all tested animals were weighed and then sacrificed. The weight of the adrenal and preputial glands, testis and spleen of the male mice was determined.
Restraint stress was performed using a previously described procedure29. To induce DNA damages, restraint stress experiments were performed using a 50 ml plastic centrifuge tube with 3-mm holes for air flow and with tail protection. The intraperitoneal injection of acrylamide (100 mg/kg, dissolved in saline) was performed 8 h prior to the experiment.
To assess the involvement of circulating glucocorticoids in the observed pheromone physiological and genetic effects, metyrapone (30 mg per kg, Sigma-Aldrich, USA, dissolved in 150 µl of physiological solution) was intraperitoneally injected 30 min before the beginning of pheromone exposure. To assess the involvement of circulating catecholamines in the observed pheromone physiological and genetic effects, propranolol (10 mg per kg per day, > 99%, Alinda, Russia, dissolved in dH2O) was added to mouse drinking water for 3 days before pheromone exposure. The procedure was carried out according to a previously described method28.
For olfactory epithelium inactivation, mice received an intranasal administration of 10% ZnSO4 (Sigma) 3 days before the beginning of pheromone exposure57. For VNO inactivation, 21-day-old mice were subjected to VNO surgical removal according to a previously described method58. Ten days after the procedure, the physical state of mice was checked (by weight measurements and behavioral assessments). In the absence of significant differences related to control animals, mice were utilized in the pheromone exposure experiment. Sham operated animals were used as the control.
T-maze odor preference test. The aversiveness/attractiveness of the odorants (2,5-DMP, 2,3-DMP) for male mice was assessed using the T-maze odor preference test. It was conducted using a T-maze, consisting of an illuminated start box (plastic box opened at the top, 30x37x30 сm) and two darkened arms (opaque plastic boxes 10×10×5 cm), where mice could freely move from the start box. For pheromone exposure, filter paper balls moistened with 2,5-DMP, 2,3-DMP (diluted 1:10000 (w:v) in dH2O) or control water were placed in the T-maze arms. A 100 W electric lamp was suspended 1.5 m above the start box. Mice (10 per group) were placed in the middle of the start box individually and had to turn to the left or the right in order to continue through the maze and hide in the goal arm. In the first 2–3 minutes, mice walked around the start box alternately visiting both arms and generally stopped in one of them (for > 2 min). The latter was considered as a preference for the corresponding stimulus. If a mouse had not reached the end of either arm after 10 min, the experiment was discontinued. The experiment was repeated 6 times for each mouse. Mouse behavior was tested three times (days 1, 3, 5). Each test consisted of 60 preference/avoidance choices, where the corresponding stimulus was tested vs. water. The percentage of odor preference was calculated as (Y/60)x100%, where Y is number of specific odor preferences from 60 approaches.
Chromosome aberration analysis in bone marrow. Bone marrow samples obtained from mouse femurs were fixed with Clark solution (96% ethanol, glacial acetic acid, 3:1) and stored at 1–4оС. For chromosome aberration analysis, bone marrow cells were stained with aceto-orcein (4% solution) (Sigma-Aldrich, USA), squashed on slides and analyzed using the anaphase-telophase method59. The bridges, fragments, lagging chromosomes and multiple disturbances (two or more damage points in a dividing cell) were considered as chromosome aberrations. At least 200 divisions at the anaphase-telophase stage per animal were analyzed. For mitotic index estimation the sum of metaphases, anaphases and telophases per 1000 interphase cells was counted. Cell divisions were analyzed by a single blinded observer.
Alkaline comet assay. The DNA damage was detected by the alkaline comet assay based on the single-cell gel electrophoresis method (SCGE). Single-cell suspensions with a volume of 150 µl (3 × 105 ml− 1) were collected and added to 150 µl of 1% low-melting (tm < 42°C, Type VII, Sigma-Aldrich, USA) agarose gel and transferred to a CH-100 solid-state thermostat at 37°С. The obtained mixture was applied to heated poly-L-lysine (Sigma-Aldrich, USA) microscope slides (37°C), then covered with 1% universal agarose solution (tm < 65°C). Two slides were prepared for each animal. Then, the mixture was covered with a coverslip (24 × 24 mm) and stored at 4°C for 10 min. All further operations were conducted in the dark or under a green light. The slides were immersed in lysis buffer at 4°C for 1 h. After washing with phosphate buffer, the slides were placed into a horizontal electrophoresis chamber (Cleaver Scientific CSLCOM10, UK). Gel electrophoresis was performed for 20 min (1 V/cm) (C.B.S. Scientific, EPS-300 X Mini-Power Supply). The slides were fixed for 5 min in a 70% aqueous solution of ethyl alcohol, dried at room temperature for 1–2 h and stained with 0.1% SYBR Green I (Sigma-Aldrich, USA) solution for 20 min in the dark. The cell nuclei (not less than 300 nuclei per mouse) were imaged using an AxioScope.A1 (Zeiss, Germany) and digital CCD camera (QImaging, QI Click, Canada) with QCapturePro 7 software (Canada). Images were analyzed using TriTek Comet-Score™ Freeware v1.5 software (USA) and the percentage of DNA content in comet tails was calculated. Cell nuclei were analyzed by a single blinded observer.
Immunofluorescence. Histone H2AX S139 phosphorylation status in mouse bone marrow cells was examined using immunocytochemistry. Cells were fixed for 5 min with 3% paraformaldehyde at room temperature on microscope slides, permeabilized with 0.2% Triton X-100 (Helicon, Russia) for 20 min, and blocked with 1% bovine serum albumin (BSA) (Sigma-Aldrich, USA) in phosphate buffered saline (PBS) for 20 min at 37°C in a humid incubation box. Slides were then incubated with anti-phospho-H2AX (Ser 139) rabbit antibodies (#9719, Cell Signaling Technology, 1:200) in 1% BSA in PBS, for 45 min at 37°C in an incubation box. After that, cells were stained with antifade solution containing DAPI (Sigma-Aldrich, USA) and visualized on an Axio Scope.A1 microscope (Carl Zeiss, Germany). Quantitative analyses of the number of cells with foci were carried out in random areas using ImageJ Software (NIH, USA). Not less than 200 cells per mouse were analyzed. Cell nuclei were analyzed by a single blinded observer.
Hormone measurements. A commercially available enzyme-linked immunosorbent assay (ELISA) kits was used to determine the serum concentration of corticosterone (Mouse Corticosterone ELISA Kit, Cusabio, USA) and oxytocin (Mouse Oxytocin ELISA Kit, Cusabio, USA). All the procedures were performed following the manufacturer’s instructions. A spectrophotometer (Bio-Rad) was used to determine the optical density.
Cunningham test. The activity of humoral immunity after pheromone exposure was evaluated using a modification of the plaque-forming cell assay60. Ten days after the end of olfactory exposure, mice were immunized with sheep erythrocytes (1×108). The count of antibody-producing cells (APC) in the spleen was evaluated on day 4 after intraperitoneal immunization. For this purpose, mice were euthanized with diethyl ether, decapitated and the APC number in the spleen of each animal was quantified. Spleen cell suspensions were prepared by teasing the whole spleen through steel wire meshwork into cell culture medium. Cells were washed and made up to 15 ml with a balanced salt solution. Spleen cells were thoroughly mixed with sheep red blood cells and complement solution (to a final concentration 10% each) in a test tube at 37°C and then spread on microscope slides under sealed cover slips. Slides were sealed with heated paraffin wax and incubated at 37°C for 1 h (when a maximum number of plaques may be counted). Erythrocytes surrounding antibody-forming cells were coated with the antibody and lysed by complement. Such areas of lysis were observed using an Axio Scope.A1 microscope (Carl Zeiss, Germany). The count of plaque-forming cells per one thousand nucleated cells was calculated60.
Animal preparation and stimuli presentation for fMRI. Mice were anesthetized by an intraperitoneal injection of urethane (75 mg/kg) and placed on a heated mattress (surface temperature 30°С) on the magnetic resonance scanner’s table. A mask connected to a fluoroplastic tube with a system of a dosed supply of the odorants was brought to the animal. After setting the fMRI equipment and acquiring the morphological data, the synchronized odor delivery and fMRI registration were launched. Stimuli presentation was achieved by an olfactometer. The air carrying the odor stimulus was delivered at a speed of 20–25 ml/min and clean air (control) at speed of 30–35 ml/min. The odor stimulus was 50 µl of 0.01% 2,5-DMP applied on filter paper (45 mm2), or 10 g of solitary female bedding. fMRI registration was conducted over a period of 24.5 min, during which animals were alternately presented with stimuli and clean air (first 3 min clean air, then a combination of 1.5 min of stimulus and 3 min of clean air, repeated 5 times). Upon completion of the fMRI procedure, the animals were withdrawn from the magnetic resonance scanner and placed in separate cages until they returned to their normal physiological condition. In total, 136 tomograms were obtained per animal, including 40 made during stimuli presentation and 96 during the time interval without odor presentation.
Functional MRI. The BOLD response on odor stimuli in the mouse MOB was studied using an ultrahigh-field BioSpec 117/16 USR MRI scanner (Bruker, Germany) at 11.7 T in combination with a 1H transmitted-only surface coil. Five minutes before the study, mice were immobilized with 4% isoflurane anesthesia (Univentor 400 Anesthesia Unit, Univentor, Malta). During the MRI scan, urethane (20%; Sigma-Aldrich, Steinheim, Germany) was injected intraperitoneally (i.p.) at a dose of 1.5 g/kg (2 × 0.75 g/kg separated by 5 min) before positioning the mouse on the animal holder. BOLD fMRI experimental data were acquired using an echo-planar imaging (EPI) sequence with the following parameters: TE1/TE2 = 16/40 ms, TR = 2,000 ms; number of slices = 3; image parameters = 2 cm × 2 cm, matrix = 128 × 128 pixels, slice thickness = 0.5 mm; number of repetitions = 300, total scan time = 40 minutes. To improve the signal-to-noise ratio, additional adjustment of the field uniformity was carried out using the FastMap method, and the method of placing the slices of signal suppression for the areas of the brain not being examined was also used. Pre-processing was performed using SPM8 (Wellcome Trust Centre for Neuroimaging, London, UK) for MATLAB© (The MathWorks, Natick, USA). Data were fitted by a mono-exponential function as S0e− TE/T2*, where S0 is the signal at TE = 0 ms to obtain T2* (1/R*2). Individual animal BOLD response maps were generated with preprocessing and general linear model (GLM) analysis with a data-driven hemodynamic response function (HRF). The following preprocessing steps were performed to improve the detection of signal activation: slice timing correction, image realignment to each first volume for motion correction, linear detrending for signal drift removal, time course normalization by the average of the baseline volumes, and trial-averaging within the same imaging session. The two-gamma variate function was chosen for depicting both conventional positive BOLD and post-stimulus BOLD undershoot responses. In individual animal functional maps, the statistically significant activation threshold was set to uncorrected p < 0.01 and cluster size > 5 voxels. Individual t-value maps were overlaid on original EPI images.
RNA extraction and sequencing. The bone marrow plug was flushed from the tibia and femur using 500 µl of cold PBS into a 1.5 µl microcentrifuge tube and re-suspended. The cell solutions were centrifuged at 1,300 × g for 5 min at 4°C, and the supernatant was removed. Seven replicates of mouse bone marrow samples for each condition were used for RNA extraction. Tubes with tissues were snap frozen in liquid nitrogen and stored at -80°C for later use. Total RNA was isolated with the QIAzol Lysis Reagent following the manufacturer's instructions (Qiagen, Hilden, Germany). The RNA yields were quantified by a Qubit 2.0 fluorometer (Life Technologies, Carlsbad, CA, USA), and RNA integrity was verified by the 2100 Bioanalyser (Agilent Technologies, Palo Alto, CA, USA) on RNA Nano chips. Total RNA was processed for paired-end deep sequencing on an Illumina HiSeq 4000 platform following the manufacturer’s instructions for the Illumina TruSeq Sample Preparation Kit v2 (Illumina, San Diego, CA, USA). All 21 indexed libraries were 150 nucleotides paired-end, with uniform sequencing depth across the samples and a median depth of 23 million reads (5% and 95% quantiles of 18 and 50 million reads).
Estimation of gene expression levels. Adapter sequences were trimmed from reads by Cutadapt (parameters -m 75 --trim-n)61. Original sequencing reads or adapter-trimmed reads were mapped to the reference UCSC (mm10) mouse genome using the spliced aligner HISAT2 (parameters -p 10)62. Duplicated reads arising from the library preparation procedure were removed by SAMTools63. Uniquely mapped reads were counted by HTSeq (parameters -r pos -f bam -s s = no)64 using corresponding data from the mouse (mm10) GENCODE VM15 GTF file as the annotation source. Differential expression analysis was performed using the R Bioconductor package DESeq265. A statistically significant set of genes was selected using a Benjamini and Hochberg false discovery rate (FDR) cutoff set to 0.05. To estimate gene expression changes between conditions, a linear model was fitted for each gene using the binary logarithm of RPKM for each sample as the dependent variables and the experimental condition for the same sample as the independent variables (Benjamini and Hochberg FDR-corrected P < 0.05). The conditions were randomly classified as one of three states: 0, 1 or 2. Linear regression was applied to the genes with statistically significant expression changes between the restraint stress group and the control group (FDR < 0.05), and genes without significant changes in the difference between the pheromone 2,5-DMP stress group and the control group (p < 0.05, FDR > 0.05). We defined linear-model genes (or condition-dependent genes) as the ones with an absolute difference of more than 0.1 in the log2-transformed mean RPKM between the pheromone 2,5-DMP stress group and the control group and monotonically increasing or decreasing expression. The Bioconductor packages “topGO” and “clusterProfiler”66 were used for Gene Ontology (GO) enrichment analysis. GO term categories were considered significantly overrepresented using the adjusted p < 0.05 cutoff of Fisher’s exact test. The Kyoto Encyclopedia of Genes and Genomes (KEGG) database was used to determine signaling pathways. Unsupervised hierarchical clustering (function "hclust" in R) was used to identify co-expression modules based on 821 genes differentially expressed in the bone marrow of mice under restraint stress. Gene expression levels were calculated as RPKM (reads per kilobase per million mapped reads). To assess concordance between transcriptional responses induced by the stress pheromone 2,5-DMP and restraint stress in mouse bone marrow tissue, the Pearson correlation coefficient of the gene expression fold change was calculated.
Statistics. Data are presented as mean ± S.D or mean ± S.E., unless otherwise stated. All the collected material was coded for the unbiased primary data assessment. The statistical analysis began with the decoding of the primary data. The normality of the intragroup distribution was assessed using the Kolmogorov-Smirnov test. All of the analyses were conducted using Graph Pad Prism 7™ software (GraphPad Software, San Diego, USA). In the case of the normal intragroup data distribution, a two-tailed Student’s t-test was applied for single comparisons between two groups, and ANOVA with the Tukey HSD post-hoc test was used for multiple comparisons. In the case of a non-normal intragroup distribution, the Mann-Whitney test was applied for single comparisons between two groups.