Lipid extraction was carried out according to Hara and Radin (1978) with a slight modification for fish spermatozoa. Solvents and chemicals used were obtained from Merck (Darmstadt, Germany). The sample was weighed and homogenized in hexane-isopropanol (3:2, v/v). The homogenized sample was placed in a centrifuge tube, and 6.67% of Na₂SO₄ was added and centrifuged at 2,737 × g for 5 min. The upper phase of the homogenate containing lipids was transferred into pre-weighed tubes and evaporated using nitrogen. The weighed samples were dissolved in hexane and stored at −80°C until further analysis.

To investigate fatty acid (FA) composition, methylation of total lipids was performed using a combination of NaOH and boron trifluoride-methanol complex according to Appelqvist (1968). Fatty acid analysis was performed using gas chromatography (Trace Ultra FID; Thermo Scientific, Milan, Italy) on a BPX-70 50 m fused silica capillary column (i.d. 0.22 mm, 0.25 μm film thickness, SGE, United States) as described by Sampels et al. (2005). The FAs were identified by comparison of sample retention times with those of the standard mixture GLC-68-A (Nu-Chek Prep, Elysian, United States) and other previously validated standards (Nu-Chek Prep, Elysian, United States; Larodan, Sweden).

The procedures of Olsen and Henderson (1989) were used for separation of major lipid classes using a mobile phase consisting of hexane, diethyl ether, and acetic acid (85:15:2, v/v/v). Separation and quantification of phospholipid classes were carried out based on protocol described by Olsen and Henderson (1989) using a mobile phase consisting of 10 mL 0.25% KCl, 10 mL methanol, 25 ml chloroform, 25 mL methyl-acetate, and 25 ml isopropanol. Samples in both cases were applied with a CAMAG TLC Sampler ATS4 (Camag Switzerland) 2 cm from the base edge of the TLC plates (20 cm × 10 cm; Silica gel 60; 0.20 mm layer, Merck, Darmstadt, Germany) in 2 mm bands with an application speed of 250 nL s^–1. Nitrogen was used as spray gas. All samples were tested in duplicate, with distance between tracks of 9.8 mm. Separation of the lipid classes was executed with a CAMAG Automatic Developing Chamber 2 (ADC 2) (Camag, Switzerland). To identify major lipid and phospholipid classes, plates were dipped in a solution of 3% cupric acetate in 8% phosphoric acid and charred for 20 min at 160°C. For quantitative determination of the separated lipid classes, the plates were assessed by CAMAG TLC scanner 3 (Camag, Switzerland). Scanning was conducted at 20 mm/sec and data resolution of 100 μm step^–1, with a slit dimension of 6.00 mm × 0.45 mm at wavelength of 350 nm. The Savitzky–Golay 7 mode was used for data filtering. The lipid classes were identified by comparison with a standard mix of all lipid classes (TLC 18-4A NuCheck, Prep, Elysian, United States) or single phospholipid standards (Avanti Polar Lipids, Inc., Alabama, United States). WinCats integration software was used for peak integration, and manual baseline and peak correction were applied as required.

To prepare samples for protein analysis, separate milt samples from 5 burbot males were diluted in SF and AM (100 μL milt:500 μL medium), incubated for 2 min at 4 and 30°C, and centrifuged at 3,000 × g for 20 min at 4°C.

Protein samples were prepared as described by Xin et al. (2019) with minor modifications. Briefly, spermatozoa were suspended in 50 mM Tris/HCl pH 7.4 supplemented with 1% Triton X-100 and protease inhibitors (Thermo Fisher Scientific, Waltham, MA, United States). Samples were sonicated and vortexed continuously for 1 h, and subsequently centrifuged at 13,000 × g for 10 min at 4°C. The protein concentration was measured using a BCA Protein Assay Kit (Thermo Fisher Scientific, Waltham, MA, United States).

2D-PAGE electrophoresis was conducted according to Xin et al. (2019). The images of gels were captured and digitalized using G:BOX Chemi XX6 system (Syngene, Cambridge, United Kingdom) and analyzed using SameSpots software (TotalLab, Newcastle upon Tyne, United Kingdom). All samples were run at least in triplicate.

Protein spots of interest were excised from the corresponding SDS-PAGE gel and digested with trypsin according to the protocol by Shevchenko et al. (2006). Obtained peptides were separated using a microgradient device (Rehulka et al., 2018) with gradient of ACN as described (Loginov et al., 2019). MS and MS/MS spectra were acquired using an Autoflex Speed mass spectrometer controlled by flexControl 3.4 software (both Bruker Daltonics). WarpLC 1.3 (Bruker Daltonics, Billerica, MA, United States) was used for automatic measurement of microgradient MALDI fractions. Spectra were processed using flexAnalysis 3.4 and ProteinScape 3.1 software (Bruker Daltonics, Billerica, MA, United States).

Mascot Server 2.4 (Matrix Science, Boston, MA, United States) was used to conduct searches against in-house prepared databases containing protein sequences from the Gadus morhua National Center for Biotechnology Information (NCBI) database (accessed 2018-06-05; 533,418 sequences) supplemented with sequences of common contaminants (Max Planck Institute of Biochemistry, Martinsried, Germany). Parameters used for the database search were enzyme specificity, trypsin, two missed cleavages were allowed; fixed modifications, carbamidomethylation of cysteine; variable modifications, N-terminal protein acetylation and methionine oxidation. Precursor ion tolerance was set at 50 ppm, and the mass tolerance for MS/MS fragment ions was set at 0.5 Da.

Motility duration, velocity, and fertilizing ability of fish spermatozoa may be strongly influenced by ambient temperature (Cosson et al., 1985; Billard et al., 1995; Bombardelli et al., 2013; Dadras et al., 2016) and osmolality (Billard et al., 1995; Wilson-Leedy et al., 2009). The osmolality of media used for activation of spermatozoa has a significant effect on ATP level (Boryshpolets et al., 2009). However, still no defined metabolic pathway responsible for these changes has been identified.

Burbot spermatozoon motility behavior at different osmolalities in relation to temperature was investigated in vitro. Trends of changes in spermatozoon VCL and LIN in the six studied media (Tris, Tris + Ca²⁺, Suc100, Suc 100 + Ca²⁺, Suc 270 and Suc 270 + Ca²⁺) are shown in Figures 1, 2, respectively. The trends of spermatozoon VCL and LIN in SF are presented in Supplementary Figures 1,2, respectively. The trend of VCL in spermatozoa diluted in SF (Supplementary Figure 1) was similar to that in non-ionic isotonic medium (Figure 1). The VCL showed nearly identical patterns in Tris and Tris + Ca²⁺ media (the media with lowest osmolality among those used) with the lowest VCL observed at 30°C and the highest VCL at 4°C during the course of motility (Figure 1). In contrast, in a study of sterlet Acipenser ruthenus spermatozoa, Dadras et al. (2016) observed the highest VCL at a high temperature (24°C) in hypotonic activation medium.

The spermatozoa diluted in non-ionic isotonic media (Suc 270 and Suc 270 + Ca²⁺) showed similar initial VCL at 4 and 30°C, however, at 30°C, a sharp decrease of VCL in Suc 270 and a gradual decrease of VCL in Suc 270 + Ca²⁺ were observed (Figure 1). This may be due to higher sensitivity of burbot spermatozoa to Ca²⁺ at high temperature (Dadras et al., 2019).

The spermatozoa diluted in hypotonic media (Tris and Tris + Ca²⁺) exhibited shorter motility (∼20–30 s) and lower VCL at 30°C compared to media with higher osmolality, possibly due to osmotic shock and consequent hypotonic cell swelling (Perchec-Poupard et al., 1997; Cosson, 2008; Bondarenko et al., 2013). The short duration of spermatozoon motility and lower VCL in Tris and Tris + Ca²⁺ at 30°C compared to the other temperatures in the same media (Figure 1) might be attributed to greater vulnerability of spermatozoa to osmotic pressure at higher temperatures (Dadras et al., 2019).

The trend of LIN with time post-activation was similar in all studied conditions (Figure 2 and Supplementary Figure 2) and not affected by medium osmolality or temperature, similar to reports for osmolality by Dziewulska and Pilarska (2018).

The VCL of spermatozoa was significantly lower at higher temperatures in hypotonic media (Tris ± Ca²⁺), particularly at the initial phase of motility, with the maximum at 15 s post-activation at 30°C of 27 and 73 μm s^–1 at 4°C. In activation media with higher osmolality, the initial VCL (15 s post-activation) showed smaller differences among the studied temperatures with activation by Sucrose 100 ± Ca²⁺ (maximum of 55 μm s^–1 at 30°C and 73 μm s^–1 at 4°C) and by Suc 270 ± Ca²⁺ (maximum of 68 μm s^–1 at 30°C and 52 μm s^–1 at 4°C). At 20–25 s post-activation in all studied media, VCL remained at the highest level at 4°C compared to the higher temperatures.

After comparison of general trend of VCL and LIN, the motility at 15 s post-activation at different temperatures was compared in each medium as well as at each temperature among target media (Figures 3A,B).

The spermatozoon VCL in Tris medium showed significant differences between low (4°C) and high (20 and 30°C) temperatures (Figure 3A), with the lowest recorded at 30°C (23.33 ± 3.56 μ ms^–1). A similar relationship was observed in Tris + Ca²⁺ (Figure 3A). Comparison of spermatozoon motility in Suc 100 at different temperatures showed VCL at 30°C (55.35 ± 3.82 μm s^–1) to be significantly lower than at other temperatures (Figure 3A). Significant differences of VCL in Suc 100 + Ca²⁺ were revealed between 30 (51.65 ± 7.7 μm s^–1) and 10°C (73.68 ± 2.6 μm s^–1) as well as between 30 and 20°C (72.4 ± 14.49 μm s^–1) (Figure 3A). The VCL in Suc 270 did not differ with temperature (Figure 3A).

At 4°C, the VCL in Suc 270 + Ca²⁺ (41.58 ± 19.5 μm s^–1) was significantly lower than that recorded in Sucr 100, Tris, and Tris + Ca²⁺ (Figure 3A). At 10°C, spermatozoon VCL in Tris (59.94 ± 5.87 μm s^–1) was significantly lower than in Suc100 (88.04 ± 4 μ ms^–1) (Figure 3A). In addition, the mean spermatozoon VCL in Tris at 20°C (39.91 ± 3.53 μm s^–1) was significantly lower than was recorded in Suc 100, Suc 100 + Ca²⁺, Suc 270 and Suc 270 + Ca²⁺ (Figure 3A). The spermatozoa activated in Tris (23.33 ± 3.56 μm s^–1) and Tris + Ca²⁺ (27.16 ± 1.96 μm s^–1) showed significantly lower VCL compared to all other groups at 30°C (Figure 3A). Comparison of the LIN in Tris, Tris + Ca², Suc 100, Suc 100 + Ca², Sucrose 270, and Suc 270 + Ca² did not differ with temperature (Figure 3B).

In line with reported threshold of osmolality of sucrose (>450–480 mOsm kg^–1) for burbot spermatozoa motility (Dziewulska and Pilarska, 2018), the spermatozoa were motile in non-ionic solutions (Suc 100 and Suc 270 ± Ca²⁺) in the present study. The results showed that higher temperatures increase burbot spermatozoa vulnerability to osmotic pressure, affecting negatively motility of burbot spermatozoa in extremely hypotonic media (Tris and Tris + Ca²⁺). This hypotonic signal can induce swelling and efflux of K⁺, resulting in membrane hyperpolarization (Morisawa et al., 1983a; Takai and Morisawa, 1995; Krasznai et al., 2000).

Information about the role of swelling in motility activation of freshwater fish spermatozoa is scarce. A significant correlation of hypo-osmotic swelling with progressive motility of spermatozoon has been reported in mammals (Padrik et al., 2012). Trends of spermatozoon volume changes illustrated by relative optical density in the six studied media (Tris, Tris + Ca²⁺, Suc 100, Suc 100 + Ca²⁺, Suc 270, and Suc 270 + Ca²⁺) are shown in Figure 4. The trend of spermatozoa volume change in AM is presented in Supplementary Figure 3.

Spermatozoon swelling in hypotonic conditions might play different roles depending on the spermatozoon motility signaling mode of fish species (Bondarenko et al., 2013). A considerable increase of ROD, representing spermatozoon swelling, in spermatozoa diluted in Tris was observed at 20 and 30°C compared to 4°C (Figure 4). In spermatozoa diluted in Tris + Ca²⁺, higher ROD was observed at 10 and 30°C (Figure 4), showing temperature dependency of spermatozoon swelling in burbot. The trend of change in ROD in Tris at 30°C was steeper compared to Tris + Ca²⁺ at the same temperature, suggesting that spermatozoon swelling at high temperature is inhibited by external Ca²⁺ in burbot.

In Suc100 and Suc 100 + Ca²⁺, extremely low levels of swelling were observed with slight increase of ROD at all temperatures over the course of 60 s, with no difference among the studied temperatures (Figure 4). No changes of ROD were observed in Suc 270, Suc 270 + Ca²⁺ (Figure 4) and AM (Supplementary Figure 3) during the 60 s period. This lack of swelling in artificial isotonic media can be generalized to spermatozoa diluted in SF. The swelling of burbot spermatozoa at 4°C, a temperature close to the natural spawning temperature, was not observed at a high level, while higher spermatozoon swelling was detected at 30°C. Swelling of burbot spermatozoa was higher in extremely hypotonic media (Tris and Tris + Ca²⁺) and increased at high temperatures (20 and 30°C). However, in species with an osmotic-dependent activation mode of spermatozoon motility such as carp, a significant spermatozoon volume change is involved in the motility activation process at natural spawning temperature (Bondarenko et al., 2013).

To better visualize differences in spermatozoon swelling, the ROD₆₀ was assessed in each medium at different temperatures as well as at each temperature in different tested media (Figure 5). The trends of ROD in Suc 270, Suc 270 + Ca²⁺, and AM did not differ with temperature (Figure 4). The results in media in which ROD changed over time included Suc 100, Suc 100 + Ca²⁺, and Tris and Tris + Ca²⁺ are presented in Figure 5.

The level of spermatozoon swelling in Suc 100 and Suc 100 + Ca²⁺ was significantly lower at 4°C (0.14 ± 0.04) compared to other temperatures (Figure 5). The spermatozoa activated in Tris and Tris + Ca²⁺ exhibited similar levels of swelling at 4 (Tris: 0.27 ± 0.05 and Tris + Ca²⁺: 0.24 ± 0.02) as well as at 10°C (Tris: 0.33 ± 0.03 and Tris + Ca²⁺: 0.29 ± 0.04) (Figure 5). A similar trend in Tris and Tris + Ca²⁺ was observed at both 20 and 30°C (Figure 5). However, the ROD₆₀ at 4 and 10°C was significantly lower than that observed at 20 and 30°C in both mentioned media, showing a higher swelling rate at high temperatures.

Lower ROD₆₀ was revealed in sperm diluted in Suc 100 and Suc 100 + Ca²⁺ compared to that diluted in Tris and Tris + Ca²⁺ at all studied temperatures, showing an increase of spermatozoon volume with exposure to low environmental osmolality (Perchec-Poupard et al., 1997).

Spermatozoon swelling at 60 s was greater than at 30 s in Tris (Figure 4). A short motility duration (∼20–30 s) was observed in hypotonic media at 30°C and the results of changes of ROD showed that swelling continued even after cessation of motility. It is assumed that the decrease of motility duration in hypotonic media might be accompanied by high degree of cell swelling and probably structural changes. However, species-specific modes of spermatozoon motility signaling might affect the role of spermatozoon swelling in hypotonic media.

Swelling of burbot spermatozoa during motility in hypotonic conditions appears to be a temperature-dependent mechanism. No swelling was observed during 60 s motility in ionic and non-ionic isotonic media (Suc 270, Suc 270 + Ca²⁺, and AM) at any studied temperature, suggesting that spermatozoon swelling is not involved in signaling of spermatozoon motility activation and spontaneous activation of burbot.