Breeding studies on red sea bream Pagrus major: mass selection to genome editing

Red sea bream Pagrus major is extremely important in the aquaculture of Japan. The breeding of red sea bream, which has a history of more than half a century, can be divided into three main categories. The first is selective breeding. The Aquaculture Research Institute, Kindai University, began breeding fast-growing red sea breams in the early 1960s through mass selection. I summarized the results of the selection and characteristics of the fish in the 1990s. The second is breeding through chromosome manipulation and sex control. Two types of gynogenetic diploids (G2N) were artificially induced. Although growth and survival of G2N were inferior, the growth of heterozygous clones produced by mating G2N was comparable to normal diploids. The sex differentiation process and the induction condition of all-male groups by 17α-methyltestosterone treatment were clarified, and the percentage of females could be increased to more than 85%. The third is breeding through genetic manipulation. We first developed the microinjection method, and succeeded in increasing the percentage of edible ratio by deleting the function of myostatin gene by genome editing. Combining new technologies, such as genome editing, with the existing breeding methods mentioned above would result in faster and more effective breeding.


Introduction
Red sea bream Pagrus major is a fish with a maximum length of 1 m, and is characterized by an oval-shaped body with lateral flattening, a raised septum, strong and sharp dorsal fin spines, a red body with scattered cobalt-colored patches on the upper half of the body, and a black posterior margin of the caudal fin. In Japan, the red sea bream is distributed throughout the country except in eastern and northern Hokkaido and Okinawa (Ochiai and Tanaka 1986). In the island nation of Japan, the most beloved and familiar fish of all time is the red sea bream. Its graceful appearance, beautiful red color, and elegant taste have captivated people, and it has become an indispensable part of celebratory occasions. The history of consumption of red sea bream by the Japanese people dates back to the Jomon period, more than 7000 years ago. The most frequently found fish species in Jomon shell middens is the red sea bream, and since these middens contain traces of food that was baked or cooked, it is thought that red sea bream was eaten raw, boiled, or baked for food (Kato 2021).
Currently, red sea bream is mainly caught by boat seine nets, small bottom trawl nets, and fishing fisheries other than longlines, and the catch remained stable at 14,000 to 16,000 tons over 40 years from the late 1970s to the present without significant fluctuations. In Japan, the aquaculture of red sea bream began in earnest after 1970, and by the end of the 1970s, when the aquaculture production of yellowtail had plateaued, the aquaculture production of red sea bream increased and exceeded 10,000 tons in 1978.
In the 1980s, the cultivation of red sea bream using artificial seedlings (juveniles) began in earnest, and its production rapidly overtook that of wild caught fish, exceeding 50,000 tons in 1990 and reaching 87,232 tons in 1999, the largest amount ever produced. Since 2000, the aquaculture production of red sea bream has been declining, falling temporarily to the 50,000-ton level, but recovering to the 60,000-ton level since 2014 (Kato 2021). The red sea bream is an extremely important aquaculture fish species in Japan, with the second largest aquaculture production and the largest number of fish in the production of seedlings for aquaculture.
The Aquaculture Research Institute, Kindai University (ARIKU) began selective breeding of red sea bream in the early 1960s and has continued to the present (Muarta et al. 1996). The ARIKU began shipping artificial seedlings to aquaculture farms in 1973, and from the 1980s onward, the high growth rate of its juveniles was highly acclaimed. I have considered the importance of the greatly improved growth of the "Kindai University strain" in the aquaculture industry, and have been conducting research to further improve the breeding of these seedlings. This review will introduce research on red sea bream breeding, focusing on studies conducted at ARIKU, which has a long history of research.
In Japan, the ARIKU was globally the first to begin mass selection of saltwater fish in the early 1960s and that has continued for more than half a century. In the 1990s, when the high growth rate of the fish was already highly regarded and orders for seedlings were flooding in from aquaculturists, I decided that the process of selection and breeding, growth improvement effects, and other aquacultural characteristics should be recorded in a comprehensive manner, and old field books and shipping records were examined and reorganized. The following is a summary of the contents (Muarta et al. 1996).

History and methods of selection breeding
The progenitor population of the selective breeding consisted of three populations of wild fish at different times of introduction, which were reared after the collection of juvenile fish between 1963 and 1966. From these populations, 10 or more high-growing fish were selected to serve as broodstock, and the 1st generation (GI) was produced by natural spawning in land-based tanks from 1968 to 1970. The next generation was produced by using 10-150 high-growth 3-to 6-year-old fish as broodstock, and individuals from different generations were crossbred. The succession process from the first broodstock to the 6 th and 7 th generation mixtures (VI·VII) is shown in Fig. 1. The basic selection procedure from fertilized eggs to broodfish is shown in Fig. 2. Each selection was performed mainly based on the size of fish: about 50% in the 1st selection, 12.5% in the 2nd selection, 2.4-4.8% in the 3rd selection, and 5.0-30.0% in the 4th selection of the fish having large bodies were selected from the population, respectively. Figure 3 shows a plot of the red sea bream body weights of the selected generations after hatching and the days required for attaining the commercially available size (1 kg). The average number of days required to attain a body weight of 1 kg, which was determined by the growth curve of each lot, of 1st and 2nd generations was 953 days (lots number, n = 1) and 985 ± 37 (n = 4), respectively. In contrast, that of 5th and 6th generations was 770 ± 61 (n = 6), and that of 6th and 7th generations was 738 ± 61 (n = 11): the days required to reach 1 kg have been getting noticeably shorter. Janssen et al. (2017) reported the estimated cumulative genetic gain in harvest weight and thermal growth coefficient (TGC) by selective breeding of rainbow trout, Atlantic salmon, European sea bass, gilthead sea bream, and turbot in Europe (Table 1). The genetic growth improvement effect of selective breeding was clearly observed in these fish species, and was particularly high for rainbow trout and Atlantic salmon with the higher number of generations of selective breeding. Cumulative genetic gain in TGC that corrected for temperature effects on growth performance was + 200% for rainbow trout and + 80% for Atlantic salmon. In the red sea bream (Murata et al. 1996), assuming that the temperature effect is small because they were kept in the same area (bay), although not completely in the same environment, fish weight at 1000 days of age, estimated from growth curves, was 896 g and 1726 g for 1 3 Fig. 1 Schematic showing the history of selective breeding in red sea bream in Kindai University  adapted from Murata et al. (1996). Numerals on arrows show the number of broodstock used for production of the next generation. Roman numerals in parentheses show generation of the populations Fig. 2 Flowchart of a typical cycle of selection procedure of red sea bream in Kindai University. Adapted from Murata et al. (1996). TL total length the wild fish and 6th and 7th generations, respectively, indicating a cumulative genetic gain of more than + 90%.

Characteristics of the selected strain
The question arose as to whether the selected strain grows better because it eats more, or because it can utilize the diet more efficiently. The results of a feeding experiment using selected and non-selected red sea breams showed that the selected fish consumed higher amount of the diet and used it more efficiently, resulting in superior growth (Kato et al. 1998). Another question was whether there was a difference in body shape (external morphology) between selected and non-selected fish. The proportions of the selected and nonselected fish with equal mean standard length (103.5 mm) were compared, and the ratios of total length, caudal-fin length, and pectoral-fin length to the standard length in the selected fish were significantly lower than those in the non-selected fish (p < 0.01), indicating that the caudal and pectoral fins of the selected fish were smaller than those of the non-selected fish (Table 2). In fishes, the caudal and pectoral fins are used for propulsion and control of swimming, respectively (Lindsey 1978). The selected red sea bream, which were propagated under aquaculture conditions that required little swimming behavior such as foraging and migration, had smaller caudal and pectoral fins, suggesting that their swimming ability was inferior to that of nonselected fish.
Population genetic analysis of the selected strain of red sea bream was conducted in the 1990s using isozyme marker analysis Kato et al. 1998), DNA fingerprinting (Takagi et al. 1995), and restriction fragment length polymorphism (RFLP) analysis of the mitochondrial DNA (Tabata et al. 1997). All three methods showed a reduction in genetic variability in the selected strain of red sea bream compared with the natural population, but no symptoms indicating inbreeding were observed. As the number of generations of selective breeding increases, genetic variability decreases, and inbreeding depression effects may appear if the selection intensity becomes too high (Falconer 1989). It is extremely important to monitor genetic variability in a selective breeding population. In recent years, based on paternity testing using microsatellite DNA markers, Fig. 3 Plots illustrating the growth curves of each generation of the red sea bream produced by selective breeding. Adapted from Murata et al. (1996). *The average days required for attaining a body weight of 1 kg which was determined by the growth curve of each lot. 1 Days from introduction of wild juvenile fish to reaching 1 kg body weight broodfish that produce high-or low-growth individuals and those that produce the red sea bream iridoviral diseaseresistant individuals have been identified Takagi 2015, 2017). The use of paternity testing will greatly increase the accuracy of phenotype-based selection and is expected to contribute significantly to both the fixation of superior traits and the exclusion of undesired traits. The selective breeding of red sea bream in the ARIKU has been practiced for close to 60 years and continues to this day. With more than 13 generations of selection, the growth is greatly improved, and the seedlings (juveniles) are still being shipped to aquaculture farms throughout Japan (Kato 2021). As mentioned above, selective breeding shows significant genetic improvement in growth, but its critical problem is that it requires a long period of time. Therefore, the following are some of the efforts aimed at shortening the breeding period.

Chromosome manipulation and sex control in red sea bream
Chromosome manipulation has attracted attention as an alternative breeding method to selective breeding and has been actively studied since the 1980s. Chromosome manipulation is a technical system that controls the number and combination of chromosomes, suppressing the first cell cleavage or retaining the second polar body and genetic inactivation of gametes. In fish, polyploidy, gynogenetic diploids, and androgenetic diploids have been produced (Arai 2001). In red sea bream, triploids were produced with the expectation of improved growth associated with sterilization, prevention of a decline in commercial value due to the development of secondary sexual characteristics (darker body color), and increased body weight due to increased cell volume. However, no improvement in growth was observed. Although sterile, most of them became male and developed secondary sexual characteristics, and their usefulness as an aquaculture strain has not been recognized (Kitamura et al. 1991;Murata 1997). Gynogenetic diploids were produced for the purpose of all-female production and the creation of inbred lines in a short period of time. They are produced by suppressing the first cell cleavage (mitotic-G2N) or retaining the second polar body (meiotic-G2N) after the eggs are inseminated by genetically inactivated sperm. In mitotic-G2N, one set of chromosomes from the female parent is doubled by inhibition of somatic cell division, resulting in individuals that are homozygous at all loci (double haploid), and clones can be obtained in only two generations. In contrast, in meiotic-G2N, the recombination between locus and the centromere during first meiotic division results in individuals that are partially heterozygous, and these individuals cannot be used as parental fish for cloning. However, by repeating the meiotic gynogenesis for two or three generations, the genotypes are not completely identical, but they are fixed, and a practical inbred line can be obtained (Arai 2001). Gynogenetic diploids have been produced in many fish species (Ihssen et al. 1990;Arai 2001;Manan et al. 2022). In saltwater aquaculture species, gynogenetic diploids have been produced in red sea bream (Kato et al. 2001a(Kato et al. , b, 2002, Japanese flounder Paralichthys olivaceus (Tabata 1991;Yamamoto 1999Yamamoto , 1995, European sea bass Dicentrarchus labrax (Peruzzi and Chatain 2000;Peruzzi et al. 2004;Bertotto et al. 2005), turbot Scophthalmus maximus (Piferrer et al. 2004;Cal et al. 2006;Xu et al. 2008;Meng et al. 2016;Wu et al. 2020), and gilthead sea bream (Gorshkova et al. 1998), although few reports examined the growth and survival rate of gynogenetic diploids based on long-term rearing (Tabata 1991;Yamamoto 1999Yamamoto , 1995Peruzzi et al. 2004;Bertotto et al. 2005;Cal et al. 2006;Wu et al. 2020).
Sex control is necessary for the mass production of improved strains with fixed superior traits by chromosome manipulation, especially by the gynogenesis technique (Piferrer 2001;Haffray et al. 2009). Since male red sea bream darken in body color due to secondary sexual characteristics during and around the spawning season, reducing its commercial value (Adachi et al. 2010), an all-female population is considered effective for red sea bream aquaculture. Establishment of a method of masculinization, which is a form of sex control, will allow mating of normal female Table 2 Comparison of proportional measurements between selected and non-selected strains of red sea breams expressed as a ratio to the standard length * Mean ± SD (n = 10). a Significantly different from the values of selected strain (p < 0.01).

Selected strain
Non-selected strain eggs with sex reversal male sperm, and analysis of the sex ratio of the resulting fish will reveal whether it is possible to determine the genetic sex-determination mechanism of this species (Haffray et al. 2009). If the genetic sex-determination system is female homozygous (XX/XY), once a sex reversal male matures, fertilization of eggs from a normal female with the sperm results in an all-female population. Generally, sex change in gonochoristic fish is induced by the administration of sex steroids during and after sex differentiation (Piferrer 2001). Therefore, it is necessary to know the basics of the sex differentiation process in order to establish sex control technology. Matsuyama et al. (1988) investigated gonad development, sexual maturation, and sex ratio in a wild population of red sea bream and reported that the fish are gonochoristic with a hermaphroditic juvenile stage, but did not investigate the details of sexual differentiation during growth from hatching. As mentioned above, the growth of red sea bream under rearing conditions is significantly faster in selected strains than in the wild population, but sex differentiation in these selected strains had not yet been investigated.
The following is my work on the gynogenetic diploids and clonal populations produced by chromosome manipulation and sex control in red sea bream.

Gynogenetic diploids (Kato et al. 2001a)
Two types of gynogenetic diploids were artificially induced in the red sea bream either by suppressing the first cell cleavage (mitotic-G2N) or by retaining the second polar body (meiotic-G2N). The eggs of red sea bream were inseminated with UV-irradiated (3000 erg/mm 2 ) sperm of Japanese parrot fish Oplegnathus fasciatus, and hydrostatic pressure shock of 700 kg/cm 2 for 5.5 min at 46 min after insemination (mitotic-G2N) and cold shock at 1 °C for 30 min at 3 min after insemination (meiotic-G2N) were applied to the eggs, respectively. The total hatching rate, hatching rate of normal larvae of the normal diploid, and meiotic-and mitotic-G2Ns were 86.5 and 94.9%, 38.1 and 45.8%, and 12.8 and 35.0%, respectively (Table 3). These results may be due to the physical treatment for chromosome manipulation and inbreeding depression effect. The standard deviations, variances and coefficients of variation of the body weight, and the standard length and body depth in 91-day-old juveniles were always larger in mitotic-G2N, smaller in normal-2N, and intermediate in meiotic-G2N (Table 4). The incidences of deformities were highest in the mitotic-G2N group and the survival rate of both G2Ns were lower than that in the normal-2N. The increased variance of quantitative traits in meioticand mitotic-G2Ns was also reported in ayu Plecoglossus altiveli (Taniguchi et al. 1990). In the current experiment, the variance of body weight in the meiotic-and mitotic-G2Ns was 3.6-4.0 times greater than that in the normal-2N (Table 4). Taniguchi et al. (1990) suggested that the increased variance in meiotic-and mitotic-G2Ns may be due to the segregation of polygenes involved in quantitative traits, as well as the effects of recessive genes and developmental instability in inbred populations. Bongers et al. (1997) investigated the causes of increased variation in gynogenetic and androgenetic diploids using common carp, and found that the variation was increased by embryo damage caused by chromosomal manipulation such as cold or heat sock treatment. The growth of the meiotic-and mitotic-G2Ns were significantly lower than those of normal-2N (Table 5). Tabata and Gorie (1988) also reported inferior survival and growth in mitotic-G2N in Japanese flounder. Therefore, utilization of the first generation of gynogenetic diploids of red sea bream as a practical strain seems to be difficult. Both G2Ns survived for 3 years to the adult stage.

Clones (Kato et al. 2002)
Clones are important experimental animals in immunology, endocrinology, developmental biology, molecular biology, genetics, etc., because they are a population of individuals with exactly the same genome. Additionally, the use of clones as aquaculture strains in fish is attracting attention (Yamamoto 1995;Müller-Belecke and Hörstgen-Schwark 2000). Two types of clones of red sea bream were produced from mitotic-G2N broodstock. Eggs from a mitotic gynogenetic diploid (mitotic-G2N) red sea bream were inseminated either with sperm from a mitotic-G2N male to produce a heterozygous clone (hetero-clone), or with UV-irradiated sperm of Japanese parrot fish and the second meiotic division suppressed by cold shock to produce a homozygous clone (homo-clone). Normal diploids (normal-2N) were also produced from one male and female as a control. The clonality of both cloned fish populations was examined by DNA fingerprinting (Kato 2003) and, globally, the first cloned red sea bream was confirmed (Figs. 4, 5).
The survival rates at the embryonic stage (ES), total hatching rates (HR), and hatching rates of normal larvae (HRN) of homo-and hetero-clones and normal-2N are shown in Table 6. The ES, HR, and HRN of homo-clones were significantly lower than those in the others. Survival rates from hatching to 548 days after hatching for each genetic group are shown in Table 7. Homo-clones tended to have lower survival rates, while hetero-clones were comparable with the normal-2N. The body weight and fork length at 50, 100, 200, 365, 456, and 548 days after hatching for each genetic group are shown in Table 8. At all stages, those of the homo-clones were significantly lower than those of the others (p < 0.05), while hetero-clones showed no significant  Fig. 4 Image showing detected DNA fingerprinting with a 33.6 probe and Hinf I restriction endonuclease of normal-2N, hetero-clones, and homo-clones, and the mother of the clones (M, mitotic-G2N). Adapted from Kato et al. (2002) differences from the normal-2N except for significantly lower body weight at 200 and 365 days than found in the normal-2N. The ES, HR, HRN, survival rate, and growth of homo-clones were all significantly lower than those in the others. This may be due to an inbreeding depression effect in homozygous individuals, or physical effects of cold treatment to prevent the release of the second polar body. In contrast, the ES, HR, HRN, survival rate, and growth of the hetero-clones were similar to those of the normal-2N. Such results were also reported for the Japanese flounder (Yamamoto 1999(Yamamoto , 1995. Since hetero-clone populations are genetically uniform while maintaining heterozygosity, they are expected to reduce phenotypic variance and improve developmental stability (Komen et al. 1993;Yamamoto 1999Yamamoto , 1995. As Yamamoto (1999Yamamoto ( , 1995 stated, crosses between homozygous clones of different strains with fixed superior traits will produce an optimal heterozygous clonal population for aquaculture; however, to date, it has not been used commercially.

Sex differentiation (Kato et al. 1999) and control (Kato et al. 2001b, 2003)
The process of gonadal sex differentiation of red sea bream of the selected strain was histologically studied using fish reared in the laboratory (Fig. 6). Central cavities were observed in gonads in all of the fish aged 3 and 4 months (Fig. 6a, b). The gonads in 34 of the 40 fish aged 7 months had oocytes at the peri-nucleolus stage (Fig. 6c). From 8 to 12 months after hatching (Fig. 6d, f), about half of the fish had bisexual gonads. Gonads of fish from 1 year to 1 year and 4 months old were ovaries or bisexual gonads (Fig. 6f, g), while those of 1-year-old fish after September and 2-year-old fish were ovaries, bisexual gonads, or testes. Bisexual gonads were not seen in 3-year-old fish (Fig. 6h,  i). Therefore, the pattern of gonadal sex differentiation in this selected strain may be summarized as follows. Differentiation to ovaries progresses in all fish until 7 months after hatching. Thereafter, half the fish continue to develop their ovaries, while the rest start to develop testis instead until 1 to   2 years old. Matsuyama et al. (1988) studied the gonads of a wild population of red sea bream and reported that gonads of 12-and 18-month-old fish were ovaries or bisexual gonads, while those of 2-year-old fish were ovaries, bisexual gonads or testes, and those of 3-to 8-year-old fish had ovaries or testes, except for a few bisexual gonads found in 3-and 4-yearold fish. The sampling frequency and number of samples in their study are different from those in the current study.  Therefore, it is difficult to simply compare their data with the results of this study but, in both cases, ovaries, bisexual gonads, and testes were found in 2-year-old fish, and almost no bisexual gonads were found in 3-year-old fish. The process of sex differentiation in the selected strain of red sea bream is considered to be basically similar to that in the wild population. In the case of gonochoristic fish, in which sex differentiation occurs prior to the young stage, hormone treatment for sexual control is performed during and around the period of sex differentiation (Yamazaki 1983), but there are few reports on appropriate timing of hormone treatment when sex differentiation occurs, which is over a relatively long period of time in case of red sea bream. Therefore, the appropriate hormone dosage and timing were examined. The necessary dose and growth stage for oral administration of 17α-methyltestosterone (MT) suitable for induction of all-male groups were investigated in red sea bream. Oral administration of MT [0.01-1.0 mg MT/kg body weight (BW)/day] for 16 weeks to 281-day-old meiotic-G2N resulted in 100% functional males in the following spawning season (Table 9). MT treatment (0.1 mg/kg BW/day) of fish of different ages (55, 141, and 893 days after hatching) for 16 weeks induced males, and testicular tissue was observed in the gonads of all MT-treated fish. While functional sperm were obtained from the fish treated with MT from 141 and 893 day-old-fish, no sperm was produced in treated 55 dayold-fish (Table 10). Additionally, when eggs from normal female were inseminated with sperm from MT-treated gynogenetic diploid males and reared, the percentage of female offspring was 85.2%, which is significantly higher than that in normal fish. Red sea bream has no heterochromosomes (Murofushi 1981) and its sex-determination system is not known. Matsuyama et al. (1988) reported that the sex ratio in a wild population of red sea bream was about 1:1. As mentioned above, the ratio of 1-year-old fish with ovaries to those with bisexual gonads or testes was about 1:1. However, high female ratios were observed in the gynogenetic diploids and/or fish produced by crossing normal females and gynogenetic diploid males. These facts show the possibility of a genetic sex determination system existing in this species. In some fish species, the sex is determined by not only genetic factors but also environmental factors (Devlin Table 9 Effects of oral administration of 17α-methyltestosteron (MT) on the gonadal sexes of gynogenetic diploid red sea bream at the end of the feeding experiment (the following spawning season) *1 Mean ± s.d. *2 100 × gonad weight/body weight.  and Nagahama 2002). Therefore, further research is required to clarify the sex determination system of red sea bream by means of progeny testing using MT-treated males. Identification of a sex-linked DNA marker and/or sex-determining gene is also needed to understand the genetic sex determination system and to control sex in this species (Li et al. 2022).

Gene manipulation
Gene manipulation is a method that can be used for rapid breeding with great effect. Over the past 30 years, within genetic manipulation, transgenesis has been studied in a wide variety of fish species and traits (e.g., Osmond and Colombo 2019;Cebeci et al. 2020;Chen and Chen 2020), starting with the addition of economically useful traits to fish, such as growth enhancement (Zhang et al. 1990), cold tolerance (Hew et al. 1992), and disease resistance (Sarmasik et al. 2002). Methods of transgenesis in fish include microinjection (Meng et al. 1999;Tanaka and Kinoshita 2001), electroporation (Inoue et al. 1990), use of a retroviral vector (Lu et al. 1997), and particle guns (Kinoshita et al. 2003;Yazawa et al. 2005). The most reliable method of gene transfer is microinjection, in which a nucleic acid solution is injected directly into cells or fertilized eggs. The microinjection method is a technique developed primarily for experimental fish species such as zebrafish and medaka (Meng et al. 1999;Tanaka and Kinoshita 2001). Zebrafish and medaka differ from red sea bream, i.e., they are freshwater versus saltwater fish, they have sinking versus floating eggs, the number of eggs obtained from one female parent fish is greater in red sea bream, and rearing from a small number of eggs is relatively more difficult in red sea bream. Therefore, we developed a microinjection method suitable for red sea bream. Next, we used this method in the breeding of red sea bream by genome editing, which has been the focus of attention in recent years (Zhu and Ge 2018;Gratacap et al. 2019).

Microinjection
Although a reliable method of microinjection into a small number of eggs was developed for small laboratory fish that can be reared from eggs in small numbers or individually, we aimed to develop a method for rearing several hundred microinjected eggs in red sea bream. Although tens to hundreds of thousands of unfertilized eggs can be obtained from a single mature female red sea bream, if they are inseminated at once, the chorion of the egg hardens over time after insemination, making microinjection difficult, and only a portion of the eggs can be injected. Therefore, about 1000 to 2000 unfertilized eggs were inseminated. Unfertilized eggs of red sea bream could be inseminated up to 2-3 h after egg collection (Kishimoto et al. 2019). A minute after insemination, the eggs were washed in seawater and immersed in Leibovitz's L-15 culture medium and arranged with a pipette in a groove on an acrylic plate with the L-15. There are two reasons for using L-15: the eggs of the red sea bream are buoyant eggs, which float in seawater, therefore it takes time to arrange them on the acrylic plate, and I thought that the damage caused by piercing with the glass needle would be lessened in L-15. While microinjection on agarose gels was the conventional method for zebrafish and medaka, the use of acrylic plates for red sea bream significantly shortened the time and improved the number of eggs processed by microinjection. After 5-10 min of insemination, the chorion of the eggs hardened and it became difficult for the glass needle to penetrate, so we repeated insemination every 5 min to repeat microinjection of eggs before hardening. This made it possible to perform microinjection on 1000-3000 eggs per person per day (Fig. 7) (Kato et al. 2007;Kishimoto et al. 2019).
Using this method, an expression vector was constructed by linking the green fluorescent protein gene (GFP) to the β-actin promoter region of red sea bream and microinjected into their eggs, which resulted in the observation of clear GFP expression (Table 11). Thus, the microinjection method developed here suggests that microinjected expression vectors can be transported into cells and synthesize proteins in red sea bream embryos.

Genome editing
Genome editing technology uses targetable nucleases, including zinc-finger nucleases (ZFNs), transcription activator-like effector nuclease (TALENs), and clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPRassociated nuclease (CRISPR/Cas), that introduce DNA double-strand breaks (DSBs) into user-defined sites (genes) in cells, and use the repair process of the DSBs site to modify the genes (Ochiai and Yamamoto 2015). In recent years, the application of genome editing technology to breeding in the fields of agriculture, livestock, and aquaculture has been attracting much anticipation worldwide (Ricroch 2019). Many gene knockout mutants and some gene replacement and insertion mutants have been produced through the use of genome editing technologies in a wide variety of plants, and many of these mutants have been shown to be useful for crop improvement (Zhang et al. 2018). Progress in genome editing has also been made in livestock, such as pigs, cattle, sheep, and goats (Lillico et al. 2013;Tan et al. 2013;Proudfoot et al. 2015), for instance to confer resistance to diseases (Whitworth et al. 2016), introduce polledness (absence of horns) (Carlson et al. 2016), or increase muscle mass (Proudfoot et al. 2015). The time required for breeding by genome editing is expected to be shorter than for cross  0 ± 0 ac 0 ± 0 ac 0 ± 0 a breeding, mutation breeding, or transgenic breeding in crops (Chen et al. 2019) and also shorter than for genomic selection in livestock (Bastiaansen et al. 2018;Jenko et al. 2015).
We have worked on breeding by genome editing (clustered regularly interspaced short palindromic repeats/ CRISPR-associated nuclease 9, CRISPR/Cas9 method) (Jinek et al. 2012) using the microinjection method described above. While the yellowtail and salmon have more than 60% of their weight in edible parts, the red sea breams have only about 40% (Kato 2021;Fry et al. 2018). This makes the cost of processing them high, and we considered ways to increase the edible portion of the red sea bream; therefore, we focused our attention on the myostatin (mstn) gene, which is responsible for suppressing excessive skeletal muscle growth. Spontaneous mutants of the mstn gene are known in many animals. A dysfunctional mstn gene in the Belgian Blue and Piedmontese breeds of beef cattle, resulting in a much greater amount of muscle mass than found in normal cattle (Kambadur et al. 1997), has been reported. If there was loss of function of this gene due to genome editing, the percentage of edible parts would increase.
The sgRNA used in the CRISPR/Cas9 system was designed and synthesized from the mstn gene sequence of red sea bream, mixed with separately synthesized Cas9 RNA, and microinjected into 2365 artificially inseminated selected strain (12th and 13th generations) of red sea bream eggs in April 2014. The hatching rate was about 50%, and the survival rate after 2 months was more than 40%, producing about 500 juveniles. After about 6 months of rearing, the fish were implanted with ID tags and genomic DNA was extracted from a portion of their fins to check for mutations in the mstn gene. A heteroduplex mobility assay revealed that about 42% of the individuals were highly mutated, and they were reared continuously. In April 2016, when those fish were 2 years old, they spawned spontaneously in the tank and the second generation of genome edited fish was successfully obtained. The results of genotyping those fish in captivity showed that some fish had mutations in a homozygous type (HM) that inherited the mutations from both female and male parents. The mutants were clearly different from wild-type (WT) fish in body shape, 5 months after hatching, and the increase in muscle mass was easily visible to the naked eye (Fig. 8) (Kishimoto et al. 2018).
Furthermore, in 2018, a third generation of genome edited fish were produced and examined for various characteristics considered necessary for use as an aquaculture strain, such as growth, feed intake, feed efficiency, muscle mass (percentage of edible parts), and external morphology (Ohama et al. 2020;Washio et al. 2021). HM (mean body weight 41.1 ± 0.3 g) and WT (42.7 ± 0.3 g) were reared for 8 weeks to compare daily feed intake, specific growth rate (SGR), weight gain, and feed efficiency. At the end of the trial, weight gain, specific growth rate, and feed efficiency were significantly higher in the HM than in the WT group (p < 0.05), and there was no significant difference in the daily feed intake (Table 12). The protein efficiency and apparent protein retention were significantly higher in the HM than in the WT (p < 0.05). These results suggest that HM feed similarly to WT fish during the juvenile stage. However, HM fish have a higher ability to convert feed efficiently and accumulate ingested protein than WT, resulting in better overall growth (Ohama et al. 2020). The processing yields were then compared between HM (635 ± 15.2 g) and WT (632.5 ± 22.7 g). For the edible part, i.e., the weight and its ratio of fillet A (with skin, scales, and ribs) and B (without skin, scales, and ribs) to body weight, the HM fish (58.7 and 47.9%, respectively) showed significantly higher values than those of the WT fish (51.5 and 39.7%, respectively; p < 0.05) (Table 13). For the weight and the ratio of non-edible parts to body weight, including the head, collar lumber with paired fins, backbone with median fin, internal organs, and blood, we observed a trend in which each mean value was higher for the WT and lower for the HM fish. A significant difference between the WT and HM fish was observed in terms of the non-edible parts for the head weight, ratio of head weight to body weight, and weight of backbone with median fins (Washio et al. 2021). Based on the growth performance and processing yield, we concluded that the mstn-deficient red sea bream would be a beneficial breed for future aquaculture. This mstn-deficient red sea bream was reported to the Ministry of Agriculture, Forestry and Fisheries and the Ministry of Health, Labor and Welfare through the venture company, and is now available for distribution in Japan (Kato 2022). In Japan, the world's first genome-edited tomatoes were launched in the market in September 2021 (Ezura 2022). This mstn-deficient sea bream is the second genome-edited food to be marketed after this, and is the world's first genome-edited farmed fish approved for marketing by the government.

Conclusion
I have been working on red sea bream breeding for many years. It is always important to rear them well for practical use. Hence, it was fortunate that red sea bream was chosen as the species to breed, as it is easy to keep among the saltwater cultured fishes, and it spawns spontaneously and actively in the land-based tank. Over 50 years of selective breeding has more than doubled the growth rate of red sea bream, and the roots of cultured red sea bream in Japan can be found at ARIKU, which has made significant contributions to the aquaculture industry of red sea bream. This selective breeding of red sea bream was used to further improve the breed through genome editing, and the mstn-deficient red sea bream globally became the first government-approved aquaculture product for distribution. As for future issues, most saltwater farmed fish, including red sea bream, take more than 2 years to reach maturity, and even with genome editing, mass production will not be possible until the third generation. Shortening the time to maturity is an important issue for breed improvement in the future. Gene transfer by genome editing technology is expected to have greater effect, although it falls under the designation of genetically modified organism (GMO). To make industrial use of breeding through genetic engineering, I would like to consider the use of practical sterility/sterility technology, including the use of chromosome manipulation to produce triploids and sex control. By combining the various breeding methods developed to date with the latest genetic manipulation techniques, I hope to achieve aquaculture production that is adaptable to future global population growth.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.