We found clear morphological differences in male reproductive organs between males, kept at 25°C for 1 or 3 weeks (C1, C3), and flies kept for 3 weeks in diapause conditions (D3) (Figure 1A). Thus, dormant males have very small seminal vesicles and male accessory glands compared to controls (C1 and C3). The immature reproductive organs of dormant males are similar to those of newly eclosed male flies (C0). Recovery from dormancy was achieved by transferring the flies back to 25°C and longer days. In flies that had recovered for 1–3 weeks (R1–R3) after 3 weeks of dormancy the seminal vesicles and accessory glands progressively increased in size, the latter eventually reached the size of 3-week-old controls (Figures 1B,C).

Since diapausing plant bug males have hypotrophied accessory glands, but still produce and accumulate significant amounts of sperm (Iarovaia and Razin, 1996), we monitored sperm number in control and diapausing, as well as recovering, D. melanogaster males. We found that males in dormancy have very few sperm cells (average of 5 sperm) in their seminal vesicles compared to fully filled vesicles in controls that contained between 1,000 (C1) and up to 2,500 (C3) sperm (Figure 1D). It seems that in D. melanogaster recovery of spermatogenesis after dormancy takes longer than recovery of oogenesis. Female flies had fully vitellogenic eggs in the ovaries after 1 week of recovery (Kubrak et al., 2014), whereas it took 3 weeks for males to completely restore sperm content after dormancy (Figure 1D). The condition of the few sperm cells produced by diapausing males was poor as determined by microscopical analysis of sperm motility (Figure 1E). A more detailed analysis of sperm production in dormant males revealed arrested spermatogenesis at the stage of sperm individualization (Bonaccorsi et al., 1990) (Figures 2A–C). Whereas, 1-week old control (C1) flies displayed clearly visible sperm individualization complexes and several thousand sperm cells in the seminal vesicles (Figure 2A), dormant males possessed very few sperm in these, but only early individualization complexes (Bonaccorsi et al., 1990) (Figure 2A). Correlated with this, the expression of scotti (also known as soti), a gene required for spermatid individualization (Fricke et al., 2014), is decreased in dormant flies (Figure 2C). Reduced meiosis during dormancy is supported by a reduction in twine (twe) mRNA, encoding the meiotic Cdc25 protein Twine (Apger-McGlaughon and Wolfner, 2013) (Figure 2D).

Male reproductive diapause results in a reversible inability of the male to inseminate receptive females (Pener, 1992), and the lack of mating appears to be part of the diapause phenotype in several insects (Swanson, 2003; Gioti et al., 2012). As expected, we found a very low number of mating attempts between dormant males and females of D. melanogaster (Figure S1A). We recorded no oviposition (and no progeny) after the negligible mating events between flies in dormancy (Figures S1B,C). Furthermore, we found that egg production, as well as egg fertility (scored as egg to adult viability), of flies that had recovered from dormancy is much lower than that of the control flies (Figures S1B,C). This agrees with data on the beetle Acanthoscelides pallidipennis where male diapause was shown to indirectly affect female reproductive performance (Chapman et al., 2003).

Interestingly, 1-week-old male flies kept under control conditions display strongly reduced attempts to mate with dormant females and dormant males rarely copulate with control females (Figure 3A). Also egg laying was strongly reduced when pairing a diapausing partner with a control one (Figure 3B), and the absence of progeny after crossing a dormant partner (D3) with a control one (C1) (Figure 3C) furthermore suggests that dormant partners are sterile. Males that had recovered from dormancy (R1–R3) displayed an increased tendency to mate with control females (Figure 3A), however they never reached the mating success of control males. Increased mating of the female normally results in an increase in fecundity, or fertility or both (Fan et al., 1999). In D. melanogaster females the fecundity of the recovered flies remained low (Figure 3B). A similar effect was seen in the post-diapausing grasshopper S. splendens (Zhu et al., 2013). Even after 3 weeks of recovery (R3) from diapause, when the male reproductive organs were fully mature, the number of copulation attempts was low between R3 and the control (C1) partners (Figure 3A), as well as between R3 flies of the two sexes (Figure S1A). This is also reflected in a low fecundity (egg production by females), since non-mated females lay much fewer eggs (Chapman et al., 2003). However, the egg-to-adult viability for the male progeny was fully restored after 2 weeks of recovery from dormancy (Figure 3C). This effect on males is not visible after crossing of partners that both had recovered from dormancy (compare Figure 3C and Figure S1C), probably as a result of low fecundity of the older females that had recovered (see Cirera and Aguade, 1997).

Extended lifespan during diapause is enabled by a strong regulation of metabolism and energy storage (Hahn and Denlinger, 2011). In addition, increased nutrient reserves are critical for post-diapause fitness (Hahn and Denlinger, 2007, 2011). Previously we found that dormant D. melanogaster females have lower body mass and display increased circulating and stored carbohydrates and triacylglycerides (TAG) compared to non-dormant ones (Kubrak et al., 2014). We find here that D. melanogaster males kept in dormancy for the same duration display a lower body weight (Figure S2) and a slight hyperglycemia, with hemolymph glucose levels higher than in C1 controls (Figure 4A). However, the circulating trehalose concentration was not affected by dormancy (Figure 4B). Whole-body levels of glucose are higher after 3 weeks of dormancy than in controls, but decreased to the control level during recovery (R1, R3) (Figure 4C). Whole body trehalose is slightly higher in males after 3 weeks of dormancy and, in contrast to glucose, did not decrease after recovery (Figure 4D). The main stored carbohydrate, glycogen, is lower at the beginning of dormancy (D1), compared to non-dormant males, then increased somewhat, but still remained at a slightly lower level than in control flies (Figure 4E). Surprisingly, the dormant male D3 flies exhausted stored TAG (Figure 4F), while females kept high levels of TAG over 12 weeks of diapause conditions (Kubrak et al., 2014). Rapid use of stored nutrients might result in a shorter dormancy in males than in females. Indeed, a shorter duration of male adult diapause was seen in several insect species (reviewed in Pener, 1992).

In D. melanogaster carbohydrates and lipids are regulated by adipokinetic hormone (AKH) and insulin-like peptides (DILPs) (Rulifson et al., 2002; Bharucha et al., 2008; Owusu-Ansah and Perrimon, 2014; Padmanabha and Baker, 2014; Gáliková et al., 2015). Our investigation of female flies suggested that AKH signaling is upregulated and insulin signaling downregulated during dormancy (Kucerová et al., 2016). In the present study we did not detect a significant increase in Akh mRNA expression, but after recovery the levels dropped significantly (Figure 5A). This correlates with the less prominent shift in carbohydrate metabolism, and suggests a difference between males and females.

We did not monitor DILP transcript levels since we found that a better readout for insulin signaling (IIS) in diapausing flies is to measure mRNA of target genes (Kucerová et al., 2016). A downregulation of IIS in dormant D. melanogaster males is suggested from our data by an activation of transcription of two FOXO targets, phosphoenolpyruvate carboxykinase (pepck) and brummer (bmm) TAG lipase in flies dormant for 1 week (Wang et al., 2011) (Figures 5B,C). This was also observed in dormant D. melanogaster females (Kubrak et al., 2014; Kucerová et al., 2016). We also found an enhanced transcription of bmm, the fly homolog of adipose triglyceride lipase (ATGL) (Grönke et al., 2005), which indicates an activation of lipolysis during diapause. Finally, the increased pepck transcription we detected in dormant males suggests an increase of both gluconeogenesis and glyceroneogenesis (Okamura et al., 2007) and was previously recorded in diapausing larvae of the mosquito Wyeomyia smithii (Emerson et al., 2010) and pupae of the flesh fly Sarcophaga crassipalpis (Ragland et al., 2010).

One of the characteristics of the diapause phenotype is enhanced stress resistance (Schmidt et al., 1993, 2005b; MacRae, 2010; Kucerová et al., 2016). This includes increased resistance to heat, cold, starvation desiccation and oxidative stress (Tatar and Yin, 2001; Schmidt et al., 2005a,b). Stress responses are reflected in activation of several conserved signaling pathways. These include the Jun-N-terminal Kinase (JNK) signaling (Biteau et al., 2011) and Janus kinase/signal transducer and activator of transcription (JAK/STAT) (Agaisse and Perrimon, 2004). The JAK/STAT signaling together with the Toll immune pathway furthermore regulate innate immune responses (Lemaitre et al., 1996; Agaisse and Perrimon, 2004; Valanne et al., 2011). We found increased expression of read-out genes from the above pathways in dormant D. melanogaster males (Figures 5D–F). Thus, Neural Lazarillo (NLaz) from JNK signaling (Hull-Thompson et al., 2009), Turandot A (TotA), a read-out of JAK-STAT signaling (Agaisse et al., 2003), and a target of Toll signaling, Drosomycin (Drs) (Fehlbaum et al., 1994) are up-regulated in dormant males compared to C1 controls, but are decreased back to the control values after recovery (Figures 5D–F). These results are consistent with those obtained for dormant D. melanogaster females (Kucerová et al., 2016) and are in accordance with a stress-resistant diapause phenotype.

We used D. melanogaster, of the Canton S strain obtained from Bloomington Drosophila Stock Center (BDSC), Bloomington, IN, USA. Experimental flies were grown on a medium containing 100 g/L sucrose, 50 g/L yeast, 12 g/L agar, 3 mL/L propionic acid and 3 g/L nipagin under uncrowded conditions at 25°C and normal photoperiod (12L:12D). Reproductive dormancy in males of D. melanogaster was induced by the same conditions as applied before for females (Kubrak et al., 2014). Briefly, newly eclosed 3–6 h old unmated male flies (designated C0) were collected under mild CO₂ anesthesia, put into 50 mL vials (10–15 flies per vial) with 7 mL of food medium and transferred to diapause conditions, with low temperature and short photoperiod (11°C, 10L:14D) (Saunders et al., 1989; Schmidt et al., 1993). Flies were sampled after 1 (D1) and 3 weeks (D3) of diapause. Concomitantly, newly eclosed unmated flies (C0) were kept under control conditions (25°C, 12L:12D) for 1 and 3 weeks (C1 and C3). To test the reproductive ability of flies after diapause, flies dormant for 3 weeks (D3) were transferred to control conditions (25°C and 12L:12D) and kept for 1–3 weeks to recover (R1–R3).

The morphology of male reproductive organs was investigated in flies kept under diapause conditions for 3 weeks (D3), compared to flies kept under control conditions for up to 3 weeks (C0, C1 and C3) as well as to flies that had recovered after 3 weeks of dormancy (R1–R3). Testes and male accessory glands from 18 to 20 males for each time point were dissected in phosphate-buffered saline PBS (pH 7.2), fixed in 4% paraformaldehyde in PBS for 20 min and rinsed in PBS with 0.1% Triton twice for 10 min. Nuclei/DNA and actin filaments in samples were stained with 4′,6-Diamidino-2-phenylindole dichloride (DAPI) (Sigma-Aldrich, 1:1000 dilution) and rhodamine-phalloidin (Invitrogen, 1:1000 dilution), respectively. Samples were washed with PBS with 0.1% Triton-X three times, mounted in Fluoromount-G (SouthernBiotech) and analyzed on a Zeiss LSM 780 confocal microscope or Zeiss Axioplan 2 fluorescence microscope. The size of seminal vesicles, testes and accessory glands were analyzed from images using Image J software from NIH, Bethesda, Maryland, USA (http://rsb.info.nih.gov/ij/). Data represent areas of measured portions of reproductive organs and are shown in arbitrary units (AU), and represent mean values from measurements of 18–20 randomly selected flies for each replicate.

Testes and accessory glands from 18 to 20 males for each time point were dissected in PBS (pH 7.2). For each experimental group three intact testes were transferred into 50 μL of saline (130 mM NaCl, 5 mM KCl, 1.5 mM CaCl₂.2H₂O, 2 mM Na₂HPO₄, 0.37 mM KH₂PO₄, pH 6.0), where seminal vesicles were carefully opened to release the sperm cells. After 30 min aliquots of diluted sperm solution were sampled in a Bürker CE cell counting chamber (chamber depth 0.1 mm Marca: Marienfeld, Germany) to count the number of sperm cells.

Sperm motility, the ability of sperm to move and swim forward, was scored after releasing sperm into saline. Data are shown as proportion of males with present motile sperm in testes, relative to the total number of analyzed males (40).

Male mating was scored by counting copulation trials. Prior to copulation unmated males and females were kept separately in the experimental conditions on the enriched food medium (10–15 flies per a 50 mL vial with 7 mL food medium). For each mating a fly couple, 1 male and 1 female, were placed in a vial with 7 mL of food medium and the number of copulation attempts was monitored during 2 h. Dormant flies (D3) were used in mating experiments about 10 min after their removal from 11°C and 10L:14D (they were given 10 min recovery before exposure to the opposite sex). Unmated dormant (D3) or recovered (R1–R3) males were coupled with females kept at the same conditions or control females, kept for 1 week at 25°C and 12L:12D (C1). We also paired dormant females with control (C1) males. After 2 h of coupling males were removed and females left for oviposition for 24 h. After 24 h of egg laying females were discarded and number of egg was counted for each fly couple. Eggs were kept at 25°C and 12L:12D until off-spring eclosed. Copulation data are shown as a relative number (percentage) of copulation trials among total number of fly couples subjected to copulation (Partridge and Farquhar, 1981). Egg-to-adult viability is the ratio of the total number of emerged adults to the total number of laid eggs (fraction of eggs developing into adulthood) (Vijendravarma and Kawecki, 2013). Data are shown as an average of 6 independent replicates with 5–7 fly couples for each replicate (30–42 males and females).

We used 10–15 unmated male flies in each of 6 independent replicates from all investigated groups: control (C), dormancy (D), and recovery (R) to measure concentration of circulating (hemolymph) glucose together with stored (whole body) glucose, trehalose, and glycogen, as well as stored triacylglycerides (TAG). Hemolymph was collected following a protocol from Broughton et al. (2008) and Demontis and Perrimon (2010) including centrifugation (3000 g, 4°C, for 6 min). After hemolymph extraction the pelleted fly bodies were homogenized in PBS (pH 7.2) in ratio of 1:10 (w/v) with subsequent centrifugation (16,000 g, 4°C, 15 min) and collection of supernatants. Carbohydrate assays of hemolymph and whole body supernatants were done as described in detail in our previous study (Kubrak et al., 2014), using a glucose assay kit with glucose oxidase and peroxidase (Liquick Cor-Glucose diagnostic kit, Cormay, Poland). Concentrations of hemolymph glucose are given in mM, whereas amount of body glucose and trehalose, as well as glycogen are expressed as micrograms per milligram of wet mass of flies (μg/mgwm).

Amount of triacylglycerides (TAG) was measured by a two-step colorimetric assay using Triglyceride Reagent (Sigma-Aldrich, T2449) and Free Glycerol Reagent (Sigma-Aldrich, F6428) in accordance with (Palanker et al., 2009; Bai et al., 2012). Data are expressed as micrograms per milligram of wet mass of flies (μg/mgwm) and represent an average of 6 independent replicates with 5–7 flies in each replicate.

Total RNA was isolated from 10 to 15 whole flies each from five independent biological replicates using Trizol-chloroform. Extracted RNA was further treated with DNAase (EN0521, Thermo Fisher Scientific). Quality and concentration of the RNA were determined with a NanoDrop 8000 spectrophotometer (Thermo Fisher Scientific). Reversible transcription (cDNA synthesis) was done following (Sajid et al., 2011) in 20 μL reaction mixture, containing 2 μg of total RNA, 1 μl of 10 mM dNTPs (R0192, Thermo Fisher Scientific), 0.4 μl of 100 μM random hexamer primer (SO142, Thermo Fisher Scientific), and 2 μL of M-MuLV reversible transcriptase (EP0352, Thermo Fisher Scientific) were used. The cDNA was then applied for quantitative real-time PCR (qPCR) using a StepOnePlus (Applied Biosystems) instrument and SensiFAST SYBR Hi-ROX Kit (Bioline) as recommended by the manufacturer. For each sample duplicate reactions of the total volume of 20 μl were conducted with a primer concentration of 400 nM and 4 μL of diluted 1:10 cDNA template. The mRNA levels were normalized to rp49 levels in the same samples. Relative expression values were determined by the 2^−ΔΔCt method (Livak and Schmittgen, 2001). The sequences of the primers are shown in Table S1.

The experimental data are presented as means ± S.E.M. Statistical analysis was performed using R Statistical Software (Foundation for Statistical Computing, Vienna, Austria) version 3.0.3. Prior to statistical treatment all data were tested for homogeneity of variances using the Fligner–Killeen test and for normal distribution by the Shapiro–Wilk normality test. Unless otherwise stated statistical analysis was performed by one-way analysis of variance (ANOVA) followed by Tukey multiple comparisons test, when data had a normal distribution, and a non-parametric Kruskal–Wallis test followed by pairwise comparisons using Wilcoxon rank sum test when data lacked normal distribution. A 95% confidence limit (P 0.05) was used throughout the study. Graphs were produced in OriginPro 7.5 software.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.