Introduction

Sperms are one of the most important cells to transport genetic information during reproduction. A sperm mainly consists of three parts—a head, a midpiece, and a flagellum. The morphology of sperm and the length of these three parts are extremely diverse among animals (Pitnick et al. 2009). To determine the factors underlying variations in sperm morphology, several evolutionary/behavioural ecological studies on sperm characteristics have been conducted so far (e.g. Fitzpatrick et al. 2009; Pitnick et al. 2009; Immler et al. 2011; Tourmente et al. 2011; Simpson et al. 2014). Accurate sperm morphological measurements, such as flagellum and head length, are required for these studies. However, spermatozoa, particularly their flagella, are extremely fragile with respect to their structures, and their forms change immediately after the death of an animal (Pursel and Johnson 1974). Since tissue autolysis occurs quickly in fish (Lougovois and Kyrana 2005), their sperms are also assumed to degenerate faster than the sperms of other animals. Thus, sperms are preferably observed immediately after picking up their living forms. However, microscopes are often not available when wild animals, including fish, are studied in the field. In such cases, no other ways can be used to measure sperm morphology, except for fixing the sperms and measuring the fixed sperms later in the laboratory.

Of the various kinds of fixative solutions used for observing sperms, formalin and glutaraldehyde are the most common. In many cases, 4-10% formalin fixative is used for studying sperm morphology (e.g. fishes: Gage et al. 1998; Balshine et al. 2001; Lahnsteiner et al. 2004; Alavi et al. 2012, birds: Lüpold et al. 2009, mammals: Seed et al. 1996; Joseph et al. 2010; Kawai et al. 2006). For the electron microscopic observation of tissues and cells, including sperms, an aldehyde-type fixative is used for pre-fixation processes. Combinations of glutaraldehyde and paraformaldehyde have been used in many cases (fishes: Koya et al. 2002; Hara 2009; Hara et al. 2013, mammals: Tanghe et al. 2002, mollusks: Li et al. 2000). Solely paraformaldehyde-type fixative (fishes; Hara 2007) or solely glutaraldehyde-type fixative (echinoderms; Chia et al. 1975) has also been utilized for the pre-fixation process. We generally use 5% formalin fixative diluted by natural seawater or freshwater for fixing the sperms of marine (Ito and Awata, unpublished data) or lake (Ota et al. 2014) fishes, respectively. However, the sperm flagellum is often severely damaged, resulting in a shorter fixed sperm than a living sperm. In addition, the agglutination of sperm was also observed. Thus, it is necessary to improve the fixation methods to measure the accurate length and morphology of the fixed fish sperm. To our knowledge, however, only two reports have investigated the effect of fixation on morphology of cells including sperms and the optimal concentration and composition of the fixative used for sperm fixation. Akatsuka (1995) has suggested that the HEPES buffer is suitable for the electron microscopic observation of a free cell. Pursel and Johnson (1974) have reported that a low concentration of glutaraldehyde solution (0.1–2%) damaged pig sperms less than a relatively higher concentration (4%).

Therefore, we attempted to explore the optimal fixation methods to observe fish sperms with optical microscopes, particularly a differential interference contrast microscope. In this study, we mainly used the sperms of Cottidae that comprises both non-copulatory and copulatory species (Abe and Munehara 2009). Since most marine fishes, including non-copulatory species of Cottidae, are external fertilizers, their sperms should be activated at least in seawater. However, the males of some teleost families, such as the copulatory species of Cottidae, Embiotocidae, and Aulorhynchidae, copulate with females, and thus their sperms are motile in an isotonic solution, but not in seawater (Koya et al. 1993; Petersen et al. 2005; Abe and Munehara 2007; Ito and Awata, unpublished data). Since sperm motility is regulated by changes in osmotic pressure, which in turn are caused by changes in pH, and Na+ and Ca2+ concentrations (Morisawa 1994; Cosson 2004), it was predicted that the effect of a fixative on the sperm cells should vary depending on the differences in the osmotic pressures of the fixative solutions. Furthermore, the sperm head morphology is different between the non-copulatory and copulatory species of Cottidae (Koya et al. 2011; Buser et al. 2017). Hence, it is necessary to consider not only the effect of a fixative solution on the sperm flagellum, but also on the head morphology of a sperm.

In this study, we investigated the effect of fixation on the sperms fixed with formalin diluted with seawater or isotonic solution, or two fixation methods of electron microscopy, using the sperms of two non-copulatory species, wherein the sperms are activated in seawater—Icelus mororanis (Cottidae) and Repomucenus beniteguri (Callionymidae)—and two copulatory species, wherein the sperms are motile in an isotonic solution—Pseudoblennius marmoratus (Cottidae) and Radulinopsis taranetzi (Cottidae). We compared the total length and morphology of the sperms fixed with different methods to those of live sperms and explored appropriate fixing methods that caused relatively smaller changes in the sperm length and deformation of the sperm head, midpiece, and flagellum. In addition, we determined the duration for which the sperm morphology was maintained in each fixative solution.

Materials and methods

Preparation of fixative solution. Eight fixative solutions were prepared (Table 1). One hundred percent formalin (37% formaldehyde) was diluted with natural seawater (FSW) or isotonic HEPES solution (FIS) that resembled the internal environment of a female ovary (150 mM NaCl, 10 mM HEPES, pH 8.0; Koya et al. 1993), so that the final concentrations became 2.5, 5, and 10%. In addition, two types of prefixed solutions used for electron microscopy were prepared: 2% paraformaldehyde, 2% glutaraldehyde, and 30 mM HEPES (2%PA2%GA, Hitachi High-technology information [https://www.hitachi-hightech.com/jp/products/science/tech/em/sem/technique/chapter1_2.html; accessed 16 May 2016]); and 2.5% glutaraldehyde, 0.45 M glucose, and 60 mM HEPES (2.5%GA, modified from Liakatas et al. 1982; Gage et al. 2002). To assess the temperature effect on sperm preservation, eight samples were dispensed into two sets of 2-mL tubes and placed at room temperature (13.2–33.4 °C) and approximately 4 °C in the refrigerator (16 tubes in total).

Table 1 Fixative solutions used in this study
Full size table

Collection of fish and their semen. Two non-copulatory fish species, Repomucenus beniteguri (Callionymidae) and Icelus mororanis (Cottidae), and two copulatory fish species, Pseudoblennius marmoratus (Cottidae) and Radulinopsis taranetzi (Cottidae) were chosen for the experiment. Repomucenus beniteguri was a substitute for non-copulatory sculpins. Six Rep. beniteguri males were collected by line-fishing at 1–2 m depth at Tassya Fishing Port, Sado Island, Japan, from June to July 2016. One P. marmoratus male was collected by hand net using SCUBA at 5 m depth in the rocky area of Tassya in November 2016. Four I. mororanis males and one Rad. taranetzi male were collected by hand net using SCUBA at 8 and 12 m depths, respectively, in the rocky area and sandy bottom of Usujiri, southern Hokkaido, Japan, in April 2017, respectively. All species were collected during their reproductive seasons. The specimens were anesthetized with 2-phenoxyethanol, and their testes were carefully dissected on ice so as not to contaminate them with urine, mucus, and seawater. After incision, the semen was taken and diluted with 0.8–2.0 µL/mL of each fixative solution. There were a few semen samples in I. mororanis testes, and thus a sample of testicular tissue (c.a. 2 mm2) was placed in 1 mL of fixing solution. One set of eight samples was kept in a refrigerator at 4 °C, and the other set of eight samples was kept in the dark at room temperature (13.2–33.4 °C). After fixation, the living sperms collected from the testes were immediately observed with a differential interference contrast microscope (BX53; Olympus, Tokyo, Japan). In this study, the room temperature was 26.0–33.4 °C in the experiments using Rep. beniteguri, 20.4–24.5 °C in the experiments using I. mororanis and Rad. taranetzi, and 13.2–21.5 °C in the experiments using P. marmoratus. The refrigerator temperature was 2.4–6.8 °C.

Measurements of sperm morphology. The sperm samples were observed with a differential interference contrast microscope, equipped with a digital colour CCD camera (DP73; Olympus, Tokyo, Japan) and a recorder (CellSens Standard software ver. 1.9; Olympus, Tokyo, Japan) at five time points: day 0 (before fixation), day 3, day 7, day 14, and day 28 after fixation. The total sperm length (TSL, µm) was measured using an image analysis software (Image J ver. 1.50i; US, National Institutes of Health, Bethesda, MD, USA). Because the sperm tail (end- and principal piece) and head, as well as the midpieces of I. mororanis, Rad. taranetzi, and P. marmoratus, were observed to be deformed by several fixative methods, we calculated the percentages of coiled tails and looped heads in the total number of sperms observed. According to the WHO criteria (WHO 1999), a sperm tail is classified as a coiled tail if its end and principal piece are coiled. A looped head was categorized into a “complete loop” and a “half loop” (midpiece and head region bent more than 90 degrees, but not completely looped).

Statistical analyses. Two trials were conducted with Rep. beniteguri, one examining the TSL of 680 spermatozoa from three males and the other examining the TSL of 925 spermatozoa from three males. One trial was conducted with the other three species. We examined the TSL of 1,004 spermatozoa from four I. mororanis males, that of 991 spermatozoa from one P. marmoratus male, and that of 996 spermatozoa from one Rad. taranetzi male. The upper and lower outliers were excluded from the data at day-0 measurement to obtain a more accurate mean value. After 3 days of fixation, however, only the upper outliers were excluded from the data to take the broken and contracted flagella into account. All data were analysed using R 3.4.1 statistical software (R Core Team 2016). The one-way analysis of variance (ANOVA, results not shown) followed by the Dunnett’s post hoc test was used to compare the means of the TSL of fixed sperms at different days against the TSL of a living day-0 sperm. All data were shown as mean ± SD. The P value of < 0.05 was considered statistically significant. All bar graphs showed a relative TSL, calculating the mean ratio of the TSL at different sampling dates to the mean value of the TSL of a living day-0 sperm, although the statistical analyses were performed using the original data. The sperm tail-coiling rates were calculated using 1,990, 1,262, 1,291, and 1,008 fixed sperms from the males of Rep. beniteguri, I. mororanis, P. marmoratus, and Rad. taranetzi, respectively. The head complete-looping and head half-looping rates were calculated for the two copulatory species with elongated sperm heads: P. marmoratus (N = 2,292) and Rad. taranetzi (N = 2,398).

Results

The observations of a live sperm revealed that the TSL was quite different among the four species studied. The average TSL was 69.9 ± 2.2 μm in Repomucenus beniteguri (N = 94 sperms of 6 males); 38.8 ± 2.1 μm in Icelus mororanis (N = 75 sperms of 4 males); 110.8 ± 2.0 μm in Pseudoblennius marmoratus (N = 23 sperms of 1 male); and 62.0 ± 1.5 μm in Radulinopsis taranetzi (N = 23 sperms of 1 male). The head and midpiece morphologies were also different among the four species (Fig. 1a-d). The fixation treatment often causes the deformation of a sperm (Fig. 1e-h). Overall, the degree of damage (the degree of sperm shortening and the head looping rate) to a sperm was significantly related to the strength of fixatives and the type of solvent (i.e. seawater or isotonic HEPES solution), and weakly related to the period of time, but unrelated to the storage temperature (Figs. 2, 3; Tables 2, 3; Electronic Supplementary Materials [ESM] Figs. S1–S5).

Fig. 1
figure 1

Examples of the living normal (a–d) and fixed damaged (e–f) sperms of the four species used in this study. (a) Repomucenus beniteguri; (b) Icelus mororanis; (c) Pseudoblennius marmoratus; (d) Radulinopsis taranetzi; (e) day-14 sperm of Rep. beniteguri fixed with 10%FSWRT; (f) day-28 sperm of I. mororanis fixed with 10%FIS4°C; (g) day-28 sperm of P. marmoratus fixed with 2.5%FISRT; (h) day-28 sperm of Rad. taranetzi fixed with 5%FISRT. White scale bars show 10 µm

Full size image
Fig. 2
figure 2

Changes in the total sperm length after fixation with different fixatives in the first trial of Repomucenus beniteguri (a–f) and Icelus mororanis (g–l). (a) 2.5%FSWRT; (b) 10%FSWRT; (c) 2.5%FISRT; (d) 10%FISRT; (e) 2%PA2%GART; (f) 2.5%GART; (g) 2.5%FSWRT; (h) 10%FSWRT; (i) 2.5%FISRT; (j) 10%FISRT; (k) 2%PA2%GART; (l) 2.5%GART. For the definition of abbreviations, see Table 1. ***P < 0.001, **P < 0.01, *P < 0.05, compared to the control at each measurement date (Dunnett’s post hoc test)

Full size image
Fig. 3
figure 3

Changes in the total sperm length after fixation with different fixatives in Pseudoblennius marmoratus (a–f) and Radulinopsis taranetzi (g–l). (a) 2.5%FSWRT; (b) 10%FSWRT; (c) 2.5%FISRT; (d) 10%FISRT; (e) 2%PA2%GART; (f) 2.5%GART; (g) 2.5%FSWRT; (h) 10%FSWRT; (i) 2.5%FISRT; (j) 10%FISRT; (k) 2%PA2%GART; (l) 2.5%GART. For the definition of abbreviations, see Table 1. ***P < 0.001, **P < 0.01, *P < 0.05, compared to the control at each measurement date (Dunnett’s post hoc test)

Full size image
Table 2 Sperm tail-coiling rates of four species. The percentages and number of sperms with coiled tails/all sperms examined in parenthesis are shown
Full size table
Table 3 Head looping rates of the two copulatory species. The percentage and number of head complete-looped sperms/all sperms and head half-looped sperms/all sperms examined in parenthesis are shown
Full size table

Non-copulatory species 1:Repomucenus beniteguri. The first and the second trials showed similar results, although the statistical results were slightly different between the trials (see ESM Figs. S1 and S2 for details). Thus, we have only shown the results of the first trial here. The TSL fixed with all three different concentrations of the FSW was shortened over time (Fig. 2a, b). In particular, 10%FSW damaged the flagella of a large number of sperms, shortening their lengths by a maximum of 20% or more probably due to flagella breakage and shrinkage (Figs. 1e, 2b). Almost all sperms fixed with 10%FSW agglutinated 14 and 28 days after the onset of the fixation treatment. Therefore, the morphology of only a few sperms and of no sperm could be measured on days 14 and 28, respectively (Fig. 2b). In contrast, when sperms were fixed with FIS, the formalin concentration and retention period did not strongly affect the TSL (Fig. 2c, d), and the flagellum remained flexible (easily bent flagella like a fishing rod when a water current hits the sperm in the space between the slide and the coverslip) and suffered slight damage. 2%PA2%GA made the sperm flagellum rigid and broken three or more days after the onset of fixation (Fig. 2e), whereas the flagella of the sperms fixed with 2.5%GA showed flexibility and slight damage, although they were shorter than those of the living sperms (Fig. 2f).

Non-copulatory species 2:Icelus mororanis. The sperms of I. mororanis were more susceptible to fixatives compared to the other three species used in this study (Fig. 2g-l; see ESM Fig. S3 for details). 2.5%FSW relatively maintained the sperm morphology (Fig. 2g). The FIS significantly and negatively affected the TSL (Fig. 2i, j). The flagella of the sperms fixed with a high concentration of formalin (10%) were considerably shrunk and/or broken, and thus many sperms were not measurable (Figs. 1f, 2h, 2j). In particular, the TSL was shortened by nearly 30% of the original living sperm in 10%FSW (Fig. 2h). The tail-coiling rates of the fixed sperms in I. mororanis increased with the increase in the concentrations of formalin and were higher in FIS than in FSW (Table 2). However, the flexibility of sperms was higher in FIS than FSW. The TSL in 2.5%GA underwent less shrinkage than that in the mixture of 2%PA2%GA (Fig. 2k, l). The tail-coiling rates were also higher when 2%PA2%GA was used as a fixative than when 2.5%GA was used (Table 2).

Copulatory species 1:Pseudoblennius marmoratus. No significant decreases were observed in the TSL fixed in 2.5% and 5%FSW (Fig. 3a; see ESM Fig. S4 for details). The TSL was slightly shortened after fixation with 10%FSW (Fig. 3b). On the contrary, the TSL fixed with three different concentrations of FIS was shortened over time (Fig. 3c, d). Furthermore, the rate of sperm deformation in terms of coiled tail and head complete loop was much greater in FIS than in FSW (Tables 2, 3), and the heads of many sperms formed a circle in FIS (Fig. 1g). The two types of fixatives used for electron microscopy did not significantly affect the TSL (Fig. 3e, f). In the 2%PA2%GA treatment, however, the sperm flagellum became rigid and coiled (Table 2), and the head became looped (Table 3).

Copulatory species 2:Radulinopsis taranetzi. As found in P. marmoratus, 2.5% and 5%FSW did not greatly affect the TSL, although slight decreases in TSL were detected on several days after the onset of the formalin treatment (Fig. 3g; see ESM Fig. S5 for details). A significant decrease in the TSL was observed in the treatment with 10%FSW as compared to the live sperm (Fig. 3h). The TSL was statistically significantly shortened after fixation with all concentrations (especially 10%) of FIS (Fig. 3i, j). Furthermore, the head morphology of the sperms fixed in FIS was quite different compared to that of the normal sperm (Fig. 1h), and more than 86% of sperms showed a completely looped head (Table 3). In the FSW treatment, a sperm with a half-loop head (incomplete loop) was also observed (Table 3). The sperm flagellum became rigid and short in 2%PA2%GA (Fig. 3k), but not in 2.5%GA (Fig. 3l). In addition, the rates of sperm with complete-loop and half-loop heads were higher in 2%PA and 2%GA than in 2.5%GA (Table 3).

Discussion

Concentration of formalin fixative. Our study showed that sperm morphology was severely damaged by a high concentration (10%) of formalin regardless of its diluting solution (seawater or isotonic solution). In particular, the sperms of the non-copulatory sculpin Icelus mororanis were more susceptible to the fixative and were shortened by a maximum of 30%. In contrast, formalin at relatively lower concentrations (2.5% and 5%) did not cause severe damages to the sperms and their fine images were observed as expected for 5%FSW. A previous study has shown no changes in the avian sperm morphology after preservation in 5% formalin solution for one year (Schmoll et al. 2016). However, the present study demonstrated that the sperms fixed with 2.5% formalin had more flexible flagella than those fixed with 5% formalin, and the sperms preserved in 5%FSW gradually deteriorated within one month. Therefore, we suggest that a low concentration (2.5%) would be optimal when the sperms are fixed using a formalin solution.

Fish body and tissues that are formed by cell junctions are generally fixed by 5–10% formalin for preservations and further observations. In contrast, since the sperm cells are far smaller than the fish body and tissues and they are free cells, it is reasonable to suppose that fixatives can more quickly permeate and fix the internal environment of a sperm. Hence, even lower concentrations of formalin can be suitable for fixing the sperm. However, the degrees of sperm damage caused by the formalin fixatives were different among the four species. The sperms of non-copulatory species are considered to be more fragile, because their sperms were more shortened than those of copulatory species. This suggests that the sperms of non-copulatory species are required to be handled more carefully, e.g. do not mix sperms inside the tubes.

During the pre-fixation process for electron microscopy, the fixed sperms are kept at 4 °C or room temperature (Li et al. 2000; Koya et al. 2002; Hara 2009; Relucenti et al. 2010). Electron microscopy aims to observe the membrane structure or inner structure of the sperm. Hence, the storage temperature is important to maintain the structure. In this study, however, the degree of damage to a sperm was unrelated to the storage temperature. This result implies that once a sperm is fixed, it is less influenced by the storage temperature, at least for observing and measuring of the outer part of the sperm with optical microscopes.

Seawater or HEPES-buffered isotonic solution? The HEPES buffer is generally considered to have an excellent buffering capacity to prevent changes in pH caused by formic acid generated from formalin. In this study, we used a HEPES-buffered isotonic solution to reproduce the osmotic pressure of ovarian fluid. The results show that in copulatory sculpins of both species, the damages caused to sperms were less severe in FSW than in FIS. In addition, looped flagella and looped heads were frequently found in FIS. Thus, despite their sperms being motile in a HEPES-buffered isotonic solution (Abe and Munehara 2007; Ito and Awata, unpublished data), seawater, in which the sperms originally lose motility, is more suitable at least in these copulatory fish.

On the other hand, the HEPES-buffered isotonic solution, rather than the seawater, was optimal in non-copulatory fish Repomucenus beniteguri, whereas opposing results were obtained in the non-copulatory fish I. mororanis. The HEPES buffer is generally considered weakly toxic and optimal for observing the fixed free cells using an electronic microscope (Akatsuka 1995). However, the end and principal pieces of the sperm tail in I. mororanis looped in the HEPES buffer. The coiled flagella and kinks in the midpiece and flagella were also observed in the rat sperm when exposed to a toxic dose of 10 mg/L methylmercury (Rao 1989). Similarly, the spermatozoa of monkeys exposed to a toxic dose of 1 mg/L methylmercury hydroxide showed coiled flagella and kinks at the midpiece junction (Mohamed et al. 1986). Thus, the selection of a diluting solution as well as the concentrations of fixatives is important for measuring a fixed sperm under the microscope.

Two types of fixatives for electron microscope. The sperm flagella became rigid across all tested fish species when 2%PA2%GA was used. Additionally, a large number of sperms with looped head and coiled flagella were observed in the two copulatory species. As 2% paraformaldehyde contained an amount of the formaldehyde corresponding to 5% formalin, and 2% glutaraldehyde is a strong fixative (Kiernan 2000), the flagella became rigid due to the excessive effect of fixation. In contrast, 2.5%GA maintained the sperm morphology of all species used in the present study. Both types of fixative contained the HEPES buffer, and thus the difference between the two fixatives was in their formalin (formaldehyde) contents. Thus, it is speculated that the combination of glutaraldehyde, formaldehyde, and HEPES buffer might have severe negative impacts: rigid and shortened flagella and high rates of half-loop sperm heads. Formaldehyde readily penetrates animal tissues, but the chemical reaction of formaldehyde with protein is slow. In contrast, glutaraldehyde chemically reacts quickly, but has a low permeability (Kiernan 2000). However, because sperms are small single cells, the mixed fixatives might be too strong to fix them and cause severe damages to them. In this study, sucrose was added to the 2.5%GA solution, according to the method described by Liakatas et al. (1982), suggesting that the osmotic regulation by sucrose might protect the surface of sperm in some ways.

Thus far, it has been believed that post-fixation is preferably carried out within a few days after pre-fixation. In practice, the sperms of a Myctophidae species stored for one month with 8% paraformaldehyde fixative containing phosphate buffer did not maintain their microstructure, when observed with an electron microscope (Hara 2007). However, this study has found that the sperm morphology can be maintained for about one month if fixation is performed with 2.5%GA at least for observations with an optical microscope. Similar results have been reported for pigs, wherein the sperms fixed with 0.1–2% glutaraldehyde solution (although types of buffer are unknown) maintained their morphology for 14 days (Pursel and Johnson 1974). We conclude that the low concentrations (2.5%) of glutaraldehyde fixatives buffered in HEPES are optimal to fix fish sperm, maintaining not only the flagellum, but also the head shape for at least one month. The 2.5%GA was also proposed as an optimal fixative for observing cell morphology and surface ultrastructure of bacteria by using atomic force microscopy (Chao and Zhang 2011).

Our study provides a reference for measuring the sperm morphology of wild animals and contributes to understanding the ultimate factors underlying the diversity of sperm morphology. However, further research is required to determine the optical methods to fix sperm of each animal species because of the diverse sperm fragility.