Male specific-pathogen-free (SPF) C57BL/6 J mice (6–8 weeks old) were purchased from the Experimental Animal Center, Air Force Medical University (Xi'an, China). They acclimated for −7 days with tap water and a pelleted basal diet before the start of the experiments. All animal experiments were performed according to the Animal Protection Guidelines of Air Force Medical University (SYXK2019-001).

To test the impact of RF-EMR on male fertility, C57BL/6 J mice (n = 8 in each group) were randomly assigned to CTRL (sham exposure group) or Radiation group (RF-EMR exposure) (Figure 1). After 10 weeks of exposure, male mice were housed with one female mouse overnight to test fecundity. This mating experiment was conducted once in a week for four consecutive times. The vaginal plugs were checked, and the mice were separated the next morning once the copulation had taken place. Litter sizes and male pups were recorded on delivery. After mating, the Radiation group mice were again subject to radiation. Sperm and testis were collected for further analysis when mating experiments were done. Thus, it was 14 weeks of radiation when we analyzed testis organization and sperm quality (Figure 1).

Furthermore, offspring from CTRL or Radiation mice were raised to explore the effects of paternal RF-EMR exposure on F1 development and glucose metabolism. Pups were numbered per the order of weighing on the postnatal day 1 and culled to 6–8/litter on the postnatal day 3 (maximally balancing the sex ratio). After weaning, the No.1 and No.2 male and female F1 in each litter were weighed weekly to monitor the growth trajectory until 12-week-old (Figure 1). In addition, postnatal 6-week-old No.1 and No.2 male F1 in each litter were picked to analyze body composition as described below. Then, at 7 weeks of birth, the No.1 and No.2 male and female F1 in each litter were chosen to perform glucose metabolism analysis. After all these metabolic tests were done, F1 mice were euthanized, and their livers were harvested at 10 weeks of age for gene expression analysis (Figure 1).

As illustrated in Supplementary Figure 1, the in-house 2.0 GHz RF-EMR exposure system was comprised of a CW signal generator (f = 2.0 GHz, power = −22.0 dBm; 1441A, China Electronics Technology Instruments Co., Ltd., China), an amplifier (output powers = 8.4 W; SWSPA-C200S/C-2, Chengdu SWIEE Power Electronics Technology Co., Ltd, China), and a horn antenna (8.94 dB gain, CW; LB-1080, A-INFO, China). The distance between the antenna and the target animals was 1.3 m, ensuring that the mice were in the far-field of the antenna.

The power density in the point of mice cage was 2.5 W/m², as measured by an electromagnetic fields meter (PMM8053A, PMM Costruzioni Electtroniche Centro Misure Radio Electriche S.r.l., Milan, Italy). The C57BL/6 J mice in the Radiation group were placed freely in a plexi glass cage (one mouse per cage), allowing for the transmission of electromagnetic waves. Mice in the CTRL group were raised in similar cages without the exposure of RF-EMR. As indicated, mice were exposed to 2.0 GHz frequency, 3 h/day (9:00 a.m. to 12:00 a.m.) RF-EMR. The animals had free access to food and water when they were exposed. Since the incidence and polarization of free movement mice kept changing during exposure, the estimated whole-body specific absorption rate (SAR) value ranged from 0.125 to 0.5 W/kg, which were calculated with the help of a prolate spheroidal model of a medium mouse (20–22).

For sperm quality analysis, cauda epididymis was cut into pieces by placing in the Human Tubal Fluid (HTF) Medium (FUJIFILM) for sperm swim up at 37°C for 30 min. Hemocytometer was used for sperm count. The sperm suspension was also detected by flow cytometry (FACS Aria II, BD Biosciences) to detect sperm apoptosis and viability via Annexin-V/7-AAD staining (Biolegend). For reactive oxygen species (ROS) detection, the sperm was washed with PBS and incubated with 10 μMDCFH-DA (Beyotime, China) in PBS for 30 min at 37°C. After incubation, samples were washed again and detected by flow cytometry.

For BrdU staining, mice were intraperitoneally injected with BrdU (Sigma, B5002, 50 mg/kg) 4 h before testis harvesting. Paraffin sections (3 μm) were deparaffinized in xylol and rehydrated. Antigen retrieval was performed by heating in citric acid buffer (pH = 6.0) at 95°C for 15 min. Then slides were incubated with 4 M HCl at room temperature for 30 min, and washed thrice with 1.5 M Tris-HCl (pH = 8.8). Slides were incubated with primary antibodies (BrdU, Santa Cruz, sc-32323; 1:100) overnight at 4°C, and subsequently incubated with Alexa Fluor 488 anti-mouse secondary antibody (Thermo Scientific, A11001, 1:500). Nuclei were stained with DAPI and tissues were analyzed by a confocal microscope (Nikon, Japan).

Sections were prepared and rehydrated for TUNEL staining according to the instructions (Beyotime, China). Briefly, sections were firstly incubated with proteinase K (20 μg/ml) for 15 min at 37°C followed by washing, then incubated with 50 μL TUNEL mix for 60 min at 37°C, and then were incubated with DAPI for 10 min at room temperature.

The permeability of the blood-testis barrier was assessed using Inulin-FITC (Sigma, F3272). Briefly, a total of 20 μL inulin-FITC (10 mg/mL in PBS) was injected into the interstitium of the testes. The animals were euthanized 40 min later, and the testes were immediately collected and embedded in optimal cutting temperature (OCT) tissue freezing medium (Leica). Nuclei were stained with DAPI, and fluorescence images of the cryosections were captured with a fluorescence microscope (Olympus).

After 10 weeks of radiation, each male mouse was housed with one 8-week-old female C57BL/6 J mouse (in proestrus or estrus) overnight inside a well-ventilated room kept at 24°C with a 12-h light-dark cycle. The female mice were checked for vaginal plugs and separated, if copulation had taken place. Litter sizes and pup numbers (for all mice) were recorded on delivery.

All the newborns' weights were measured at the age of 1 day. After weaning at 3 weeks of age, we weighed some of the offspring every week from 3 to 12 weeks of age. At 6 weeks of age, the body composition of F1 male mice was measured to obtain lean mass and fat mass by nuclear magnetic resonance (NMR) by the EchoMRI-500 (USA).

For all experiments, mice blood glucose was determined with a blood glucose meter (YUWELL, China) from the tail vein blood. In brief, mice were given free access to the food before measuring fed glucose at 4 pm. Fasting blood glucose levels were measured at 8 am after 12 h fasting. Both fed and fasting blood glucose were measured at 7 weeks of age for F1 offspring.

The intraperitoneal glucose tolerance test (IPGTT) and pyruvate tolerance test (PTT) were performed at 9 am, after 12 hrs fasting. In contrast, the insulin tolerance test (ITT) was performed at 2 pm, following a 6 hrs fasting. Glucose (1 g/kg, Sigma), pyruvate (1.5 g/kg, Sigma), or insulin (0.75 IU/kg, Novo Nordisk, Denmark) was administrated intraperitoneally to the subject. The glucose level from the tail vein blood was measured at 0, 15-, 30-, 60-, and 120-mins post injection. IPGTT, PTT, and ITT were performed in F1 mice at 7-, 8- and 9- weeks of age, respectively.

Total RNA was extracted with RNAiso Plus Reagent (TaKaRa, Japan) and then reverse-transcribed by Prime Script RT reagent kit (TaKaRa, Japan). Quantitative PCR was performed using TB Green (TaKaRa) on a Real-time PCR Detection System (Bio-Rad, United States). Cycle threshold (CT) value was used to calculate the fold change by the 2^−ΔΔCT method, and the relative mRNA expression was normalized to Actin. The primer sequences of Gck, Pklr, G6pc, Pck1, Pygl, Foxo1, Irs1, Irs2, and Actin are listed in Supplementary Table 1.

All experiments were performed at least three times, data are showed as means ± standard error of the mean (SEM), and statistical analyses were carried out using GraphPad Prism 7. The area under the curves (AUCs) of glucose levels in IPGTT, ITT, and PTT were calculated with GraphPad Prism 7. Normality of data was tested first, and then a two-tailed Student's t-test was used to test differences between samples. p < 0.05 was considered as significant.

After 14 weeks of exposure, mice testis in the Radiation group (RF-EMR exposure) presented no morphology difference compared with the CTRL group (without exposure) by H&E staining (Figure 2A). Apoptosis of testis germ cells was significantly induced in the Radiation group as determined by TUNEL assay, while the proliferation rate detected by BrdU staining was not significantly influenced (Figures 2B–E). The blood-testis barrier plays a pivotal role in male fertility (23); we observed its integrity by Inulin-FITC injection (Figure 2F). The fluorescent image indicated that Inulin-FITC distribution was restricted to the testis interstitial space in both groups, suggesting that the murine blood-testis barrier might not be affected after 14 weeks of RF-EMR exposure. Collectively, these data indicated that testis germ cells apoptosis is induced after RF-EMR exposure in our experimental conditions.

We then analyzed the influence of RF-EMR exposure on murine sperm quality. Early apoptosis and death rate of sperm in the radiation mice group showed no significant difference compared with that of the control group, as presented by Annex in-V/7-AAD staining (Figures 3A,B). Epididymis sperm concentrations were also not changed (Figure 3C). Furthermore, there was no difference in sperm ROS between the two groups (Figures 3D,E). These data revealed that RF-EMR exposure in our experimental conditions does not influence sperm quality.

We further investigated the effect of RF-EMR exposure on mice fertility by detecting the female mice pregnancy rate and sex ratio of pups. The results showed that RF-EMR exposure did not influence male mice fertility and had no effects on litter size and sex ratio (Figures 4A–D). These data underscored that our experiment's RF-EMR exposure does not affect male mice fertility.

To assess the effects of RF-EMR on glucose metabolic homeostasis in male mice, we measured the body weight, the blood glucose level, and the glucose tolerance (IPGTT) in mice following a period of 14 weeks RF-EMR. There was no difference in body weight between the CTRL and the Radiation mice (Supplementary Figure 2A). Neither fasting nor fed blood glucose was altered in the F0 exposed male mice (Supplementary Figure 2B). Also, the glucose tolerance was not influenced (Supplementary Figure 2C). These results demonstrated that RF-EMR does not impact F0 mice glucose metabolism.

As Figure 5A indicated, there was no difference in birth weights of F1 between the CTRL father (C-F1) group and the Radiation father (R-F1) group. Interestingly, the weights of male offspring but not female offspring remained significantly lower in the R-F1 group than those in the C-F1 group after weaning. However, no significant difference was detected after 12 weeks of age between the R-F1 and C-F1 groups (Figure 5A). These results suggested that the body weight trajectories of F1 males are influenced during the period of adolescence while that of females are not.

To explore the potential effects of paternal RF-EMR exposure on the female offspring's metabolism, the blood glucose level and the glucose homeostasis (evaluated by IPGTT, ITT, and the PTT) in female F1 mice were investigated. There was no difference in parameters as aforementioned between female C-F1 and R-F1 groups (Figures 5B–E). These results underscored that paternal RF-EMR exposure does not impact F1 female mice glucose metabolism.

As an altered growth curve was observed in the male F1 mice, we next determined the body composition of F1 mice, and the data indicated that R-F1 male mice gained significantly less fat mass than C-F1 (Figure 6A, 45 days of age). Furthermore, the fasting blood glucose level and the AUC (the area under the curve) of IPGTT were significantly lower than that of C-F1 (7 weeks of age), while the AUC of ITT (8 weeks of age)and PTT (9 weeks of age) did not change (Figures 6B–E).

Hepatic glucose production plays an essential role in regulating glucose homeostasis through glycolysis, gluconeogenesis, and glycogenolysis (24, 25). To examine the underlining mechanism of male R-F1 glucose metabolism alteration, next, we focused on glucose metabolism associated enzymes of the liver. As shown in Figure 7A, of these key enzymes, the mRNA level of glycolytic enzyme Gck was significantly down regulated, while the other enzymes were not influenced. In addition, insulin signaling, which is essential to the regulation of live glucose metabolism, was not altered (Figure 7B).