|Home | About | Journals | Submit | Contact Us | Français|
Low membrane permeability is one of the major obstacles to the successful cryopreservation of zebrafish embryos. The aim of the present study was to explore if this could be overcome by yolk modification with different cryoprotectants by micro-injection. Initial investigation of two cryoprotectants, methanol and sucrose, was undertaken to determine their suitability for micro-injection supplementation of the yolk mass. Intact zebrafish embryos at 50% epiboly stage were injected with Hanks' solution, 5.2 M methanol or 1.3 M sucrose yielding approximate final concentrations of 2.0 and 0.5 M of the cryoprotectants within the yolk sac respectively. After micro-manipulation, the embryos were cultured at 28°C for three days and their survival assessed at the hatching stage. All micro-manipulations performed in the present study resulted in a significant decrease in embryo survival (P<0.05). Embryos micro-injected with methanol or sucrose were also subjected to a cooling procedure. They were placed in 3M methanol + 0.5 M sucrose at room temperature for 30 min and then cooled from 20°C to 0°C at 2°C/min, from 0°C to −7.5°C /min at 1°C/min, seeded at −7.5°C and held for 10 min, before cooling at 0.3°C/min to − 20°C or until full crystallization in all embryos. The processes of extra- and intracellular crystallization were studied by cryomicroscopy. The temperature of intracellular crystallization did not differ significantly between control and injected embryos. However, it was found that intracellular crystallization did not always happen instantly after extracellular crystallization.
Whereas the embryos of many mammals have been successfully cryopreserved, there is still no protocol for the successful freezing and thawing of fish embryos. There are several factors that are believed to obstruct their successful cryopreservation. These are the large size of embryos, their low permeability to water and cryoprotectants, the large amount of yolk, the multi-compartmental nature of the embryo and their chilling sensitivity. Fish embryo membrane permeability parameters are among the lowest to be reported for embryos (5,30,31). Studies on the chorion permeability (8,10,26) of zebrafish embryos demonstrated it to be a porous membrane freely permeable to water, electrolytes and a range of cryoprotectants such as methanol, dimethyl sulfoxide (DMSO), ethylene glycol and propane-1,2-diol, but the permeability of the vitelline membrane to water and most cryoprotectants was shown to be low (24,30,31). The yolk syncytial layer (YSL) has also been reported to be responsible for the observed low cryoprotectant permeability of the embryos (5).
These low permeabilities result in insufficient influx of cryoprotectants and efflux of water. In order to solve the problems associated with low membrane permeability, some studies have been carried out on yolk modification in zebrafish embryos with the help of microinjection (9,27). The effects of propylene glycol, dimethyl sulfoxide and anti-freeze protein injections on the freezability of fish embryos were reported. Although no improvement after freezing was observed, the studies demonstrated a need for further investigation of yolk modification. Many factors can affect successful cryopreservation protocol design, namely stage of development, pre-treatment of embryos (with chorion or dechorinated), cryoprotectants, and their concentration, cooling and thawing regimes, etc. In the present study, investigations were carried out on the effects of microinjected cryoprotectants on the processes of crystallization in intact embryos during controlled slow cooling.
The formation of intracellular ice is one of the major contributors to cell death during freezing and thawing (17). Depression of the ice-nucleation temperature is normally achieved by introducing cryoprotectants that change the processes of crystallization as well as enhance the efflux of intracellular water. Depression of intracellular crystallization (ICC) temperature below that of extracellular crystallization (ECC) is believed to be essential to successful cryopreservation using a slow cooling protocol (7). For example in mouse embryos, in order to survive the freezing procedure the temperature of intracellular crystallization needs to be as low as −40° C (25). In zebrafish embryos intracellular crystallization takes place at much higher temperatures ranging from −13° C to −21° C, depending on embryo development stage (7,15). Embryos without chorion are reported to have lower ICC temperatures than those of intact embryos (15). The propagation of ice intracellularly is reported to happen almost instantly (7) after ECC. This is considered a crucial obstacle in designing a slow cooling protocol. In this study methanol and sucrose were injected into the yolk of intact zebrafish embryos because of their benefiting effects reported previously (30,31). The impact of microinjections on the physical process of crystallization within zebrafish embryos were studied as well as the toxic effect of the injected cryoprotectants. Cryomicroscopy was used to register the crystallization events that take place during fish embryo freezing.
Adult zebrafish (Danio rerio) 12-14 weeks old were obtained from aquatic suppliers. They were maintained in 45 liter (28 × 28 × 58 cm) glass fish tanks at 28°C with approximately 30 fish per tank. One quarter of the tank water was replaced twice a week. Tap water aged for at least 2 days was used as tank water. The tanks were aerated and filtered constantly. Fish were fed three times per day with dry flake and brine shrimps. A controlled day-night cycle was used (14 hr light/10 hr dark). Females and males were kept together in the ratio of 1:2-3. A glass tray covered with a plastic net and plastic grass was placed at the bottom of the tank to trap embryos. The embryos were removed the following morning since spawning was induced at first light. Embryos were then placed in tank water at 28°C and kept incubated until the 50% epiboly stage.
Intact zebrafish embryos at 50% epiboly stage were injected with Hanks' solution, 5.2 M methanol or 1.3 M sucrose yielding final estimate concentrations of 2.0 and 0.5 M within yolk sec respectively. Flaming brown type micropipette puller (Sutter Instrument Co., USA) was used to pull capillary tips for injections. Solutions were loaded with fast green dye (0.24 mg/ml) into pipette tips. A pressure injector (World Precision Instruments Inc., USA) was used to deliver the solutions into yolk of the embryos. The required volume was achieved by adjusting pressure and duration of injection and diameter of the tip of the micro-capillary. The volume injected was determined by measuring the diameter (D) of drop injected into a large drop of oil on a micrometer, and the volume (V) calculated using the equation V=(4/3)(D/2)3π. The estimated volume of injected cryoprotectants was 33.5 nl. The embryos were lined up in rows on the injection pad, and injection was performed directly into the yolk from the vegetal pole. After injection the embryos were incubated in 6-well tissue culture dishes in clean tank water with 0.1 mg/ml streptomycin and 100 U/ml penicillin (Sigma Chemical Co) at 28.5° C until hatching. Embryos were inspected daily, dead embryos were removed and the incubating media was refreshed. The total survival of embryos at hatching was determined.
Two or twenty four hours after injection the embryos were placed in 3M methanol with 0.5 M sucrose for 30 minutes at room temperature and then cooled using the following regime: 20°C to 0°C at 2°C/min, 0°C to −7.5°C at 1°C/min, seeded at −7.5°C (by moving a carrier towards the seeding point of the stage), held for 10 min, then −7.5°C to −20°C at 0.3°C/min. Embryos were kept at −20° C for 10min before thawing was performed at 130° C/min. The embryos were then washed with Hanks' solution and left at room temperature for 3 hours.
Only embryos having good morphological characteristics with a visual trace of dye inside the yolk were used for freezing. After equilibration in cryoprotectant medium described above 3-4 embryos were placed into 200 μl of cryoprotectant media in a quartz crucible on the stage of cryomicroscope (BCS196, Linkam Scientific Instruments, Surrey, UK). Embryos were cooled and thawed according the protocol described above. Photomicrographs of embryos during the cooling regime were recorded.
For the embryos at heart-beat stage (24 h after injection) the following parameters were recorded during cooling: temperature of extracellular crystallization, temperature of the beginning of intracellular crystallization (first signs of embryo darkening) and temperature of full intracellular crystallization (darkening of the whole embryo).
In all experiments, 10-20 embryos were used for each treatment and the experiments were repeated at least 3 times. Differences between treatments were analyzed using Student's t test. Values of P < 0.05 were considered to be statistically significant. All data presented were means with standard error of the mean (SEM).
All micro-manipulations performed in the present study resulted in a significant decrease in the survival of the embryos (P<0.05). Injection of Hanks' solution, 5.2 M methanol or 1.3 M sucrose resulted in embryo survivals of 22.7 ± 6.7%, 15.2 ± 9.1% and 18.7 ± 9.1% respectively in comparison with 76.6 ± 5.9% in control (Fig. 1). Puncture of embryos without injecting solutions also decreased embryo survival significantly, but to a lesser extent (51.0 ± 12.8%).
Embryos with good morphology and obvious trace of dye in their yolk were subjected to a cooling procedure to −20°C. No survival was observed under these experimental conditions. However, when individual embryos were studied under the cryomicroscope the morphology of embryos that were cooled 2 hours after micro-injection (60-70%-epiboly stage) was seen to deteriorate almost immediately after thawing (Fig. 2), whereas embryos that had been allowed to develop to heart-beat stage showed preserved morphological structure after thawing (Fig. 3e). No difference in morphology was established between injected and non-injected embryos in this investigation for heart-beat stage.
Studies on crystallization in heart-beat stage embryos showed that there were no significant differences in temperatures of intracellular crystallization (ICC) between the control, methanol and sucrose injected embryos (Figure 4). It was observed that ICC did not always occur first in the embryos that made initial contact with the extracellular ice (Figure 5b) and ICC did not always develop from a single point. It was clear that ICC could have more than one sources of initiation (Figure (Figure5c5c and and6d).6d). The morphology of thawed embryos was preserved quite well (Figure 6f) until the addition of washing media. After contact with washing media, the embryo tissues became opaque. On one occasion an interesting observation was made. A sucrose injected embryo did not reach full ICC at −20°C (Figure 3) (usually ICC occurs at above −20°C), it was therefore cooled down to −23°C to ensure that full intracellular crystallization took place. The embryo turned completely dark at −20.7°C which was the lowest temperature for ICC recorded in all experiments. The embryo was then thawed and treated following the normal experimental procedures. This embryo retained an intact morphology (Figure 3e) 3 hours after thawing and had spontaneous movements. This was the single observation of an embryo which had clearly registered intracellular crystallization but showed the signs of life after thawing.
It is generally believed that low permeability of fish embryo vitelline membrane and the yolk syncytial layer is one of the obstacles that affects the movement of cryoprotectants (5, 30) and water and this consequently affects the freezability of the embryos. In this present study, direct microinjection of cryoprotectants into the yolk was undertaken. The results indicated that microinjection procedure itself is harmful for the embryos and the introduction of Hanks' media without cryoprotectant affects the embryo survival significantly.
Biochemical toxicity of cryoprotectants could also result directly from the interaction between cryoprotectant and cell enzymes or indirectly by altering the environment of the cellular bio-molecules, which in turn may result in changes in the dielectric constant, redox potential, ionic strength, pH, and surface tension (1,2,4). A desirable high concentration of cryoprotectant within embryo yolk during freezing might have negative effects on embryo development subsequently.
Studies reported for turbot embryos (Scophthalmus maximus) (27) also showed low or zero early stage embryo survival after injection of embryo media or even simple puncture, although embryo survival after microinjection appeared to be increased at tail bud stage. The embryo survival after injection of high concentration of cryoprotectants (33nl 5.2M methanol or 1.3M sucrose) obtained in the present work are comparable with Janik et al's report, where propylene glycol and dimethyl sulfoxide were injected into zebrafish embryos (9). It was shown that even though injected cryoprotectants could freely diffuse through the yolk, they would not move through the yolk syncytial layer to the blastoderm. This phenomenon was also observed in turbot embryos (27). The results obtained in this study seem to indicate that decreased survival may have been caused by physical distortion compromising anatomical integrity of the embryos during introducing of extra volume rather than by toxic interaction.
After overcoming the impact of cryoprotectant injection, the embryos are likely to be more sensitive to further stress. The physical distortion of the embryos has been shown to affect their further capability to resist stress such as cooling (14). It was shown that punctured zebrafish embryos could resist cooling procedure much better if they were incubated as long as 24 hours after puncture in comparison with those incubated only 2 or 6 hours. It is believed that some resealing takes place that might restore membrane integrity (20, 28). Therefore, the injected embryos in the present study were incubated for 24 hours after injection to make sure that restoration of membrane integrity takes place before they are subjected to cooling when membrane integrity is of paramount importance. Any attempt to cool embryos soon after microinjection resulted in immediate destruction in morphology of the embryos. The observed destruction of embryos immediately after cooling could be a result of membrane damage caused by manipulation as well as a higher sensitivity of earlier stage embryo to cooling. Therefore 24 h incubation of the injected embryos may contribute into membrane healing and also the embryos progressively develop to a stage that is known to be more chilling resistant (3,13). This was confirmed in this work, the embryos incubated for more then 24 hours retained a better morphology after cooling-thawing than those having a short post-injection incubation period.
The effect of cryoprotectant injection on intracellular crystallization was also investigated. Several theories have been put forward to explain the mechanism of intracellular crystallization. Muldrew and McGann (21) suggested that ICC is caused by the damage in plasma membrane that allows external ice to penetrate the cell. Other theories suggest that the external ice either grows through pre-existing membrane pores (18) causing ICC or initiates conformational change in membrane (29) triggering ICC formation. All data obtained for fish embryos so far suggests that intracellular crystallization is strongly dependent on extracellular crystallization (7,15). Some authors have suggested that high nucleation temperature of zebrafish embryos makes a slow cooling regime a highly unlikely protocol for the achievement of any success in fish embryo cryopreservation (15,19). Although the mechanism related to ice spreading is not entirely clear, there are strong connections between ICC and ECC in fish embryos. The possibility of ICC occurrence is determined by a susceptibility of intracellular media to an initiation of ice formation by extracellular ice crystals and by a barrier property of membrane (12,22). This susceptibility depends on the degree of water bonding, viscosity of intracellular medium and capability to inhibit ice formation. It has been shown that, in weather loach (Misgurnus fossilis) embryos, increasing the degree of media supercooling during cryopreservation significantly decreases the range of temperatures within which intracellular crystallization takes place (23). The more supercooled the media, the faster intracellular crystallization occurs in fish embryos. The increase of cooling rate also decreases the time over which embryos become completely crystallized. It was suggested that cooling rate less than 0.6°C/min along with minimal supercooling of external solution have the potential to decrease the risk of ICC in weather loach embryos at relatively high temperatures (23). In the present study, a cooling rate of 0.3°C/min together with seeding at −7.5°C demonstrated a relatively slow propagation of intracellular crystallization within embryos, contrary to a previous study (7), where ICC was reported to occur almost instantly after ECC when dechorinated embryos were cooled in media without seeding. It is clear that intracellular crystallization is a complex process that depends on a range of factors, but the degree of supercooling of the cryoprotectant media appears to be one of the most important factors affecting the propagation of ice within embryos. In the present study we also observed an “explosive” type of ice formation across the whole cryoprotectant medium and the embryos in the absence of seeding or when seeding was performed at lower temperatures (data not shown). This suggests that even if ICC is initiated by ECC, the rate of intra-embryonic ice growth might be controlled. The maximal possible impediment to ice propagation within the embryo might potentially have a positive effect.
Changing intra-embryonic properties by the injection of cryoprotectants did not lead in this study to the desired effect. This could be due to the fact that there was still insufficient cryoprotectant in the yolk. However, providing fish embryos with even higher cryoprotectant might not solve the problem of high temperature ICC. The evidence to this can be found in another aquatic model, the starfish oocyte, which also has a relatively high ICC temperature (11) and large size. Contrary to zebrafish embryos, starfish oocytes do not have such a low membrane permeability and can easily be dehydrated or saturated with cryoprotectants. However, none of these features helped to suppress temperature of intracellular crystallization and ICC was reported to happen almost immediately after ECC, as in the case with zebrafish embryos. Koseoglu et al (11) suggested that there must be some effective ice nucleator in the oocyte cytoplasm that is responsible for the ice propagation. This might also be the case with zebrafish embryos.
In the present study it was observed that during the course of ICC it was always the cellular embryonic mass that showed first signs of darkening even though embryos were only injected with embryo media and exposed to cryoprotectant before freezing. If the permeability of the yolk syncytial layer is low and blastoderm cells are considered to be quite permeable to cryoprotectants (5), it is unclear why they darken first due to ICC well before the yolk compartment? It could be the case that after puncture or microinjection, embryo vitelline membrane sensitivity to damage was increased by external ice (21) or the ability of the cell membrane to induce heterogeneous nucleation when external ice is present (29). If it is the later case zebrafish embryos possess features that will have to be addressed before they can be successfully cryopreserved using slow-cooling. More studies are needed on the biology and physics of ice formation in zebrafish embryos. There are some indications from this work to suggest that it might be possible to adjust the media and cooling rates to delay ICC in relation to ECC. However, modification of the yolk by injection of cryoprotectants may not have a role to play in this development.
This research is funded by the Welcome Trust (GR069889RP)