Phylogeography of the brackish water clam Corbicula japonica around the Japanese archipelago inferred from mitochondrial COII gene sequences.

We investigated the phylogeography of the Asian brackish water clam, Corbicula japonica, to clarify its demographic history using partial mitochondrial COII gene sequences (990 bp) from 283 individuals collected from around the Japanese archipelago and adjacent areas. Phylogenetic analyses revealed the presence of two major groups within our samples: monophyletic Group I comprising Lineages A-E of C. japonica and paraphyletic Group II consisting of Corbicula sp. Lineages A-C were distributed in Japan and Sakhalin Island, and Lineages D, E, and Corbicula sp. were distributed in the Korean Peninsula. Nested clade analysis (NCA) revealed that Lineage A-the dominant lineage in Japan-consisted of Pacific and Japan Sea lineages, the latter comprising southern and northern Japan Sea groups. Genetic diversity indices of the southern group were higher than those of the northern group, suggesting historical range expansion in the Sea of Japan from southwest to northeast. Geographical distribution of these genetic groups appears to have been influenced by major ocean currents around the Japanese archipelago. Dominant haplotypes in the star-shaped haplotype network of Lineage A were distributed throughout the entire distribution range of each genetic group, implying rapid range expansion of this species. The results of mismatch distribution analysis and molecular clock estimation suggest that expansion of lineage A occurred during the late Middle or Late Pleistocene. In contrast, restricted or past gene flow suggested by NCA and the many unique haplotypes (110/123; 89.4%) present in Lineage A suggest that gene flow among extant populations is rather limited.

Mitochondrial DNA transcription levels during spermatogenesis and early development in doubly uniparental inheritance of the mitochondrial DNA system of the blue mussel Mytilus galloprovincialis.

In some species of bivalve, there are two highly diverged mitochondrial genomes, one found in all individuals (F type) and the other normally in males only (M type). In Mytilus, a maternally-dependent sex ratio of the progeny has been reported. Some females almost exclusively produce daughters, while others produce a high proportion of sons. We previously reported that in M. galloprovincialis, M type mtDNA copy number may be maintained during spermatogenesis and the development of larvae of male-biased mothers to sustain the doubly uniparental inheritance system. In this study, we investigated transcription levels of M type mtDNA before and after fertilization to understand its function in the germ line. First, we quantified transcription levels of M type mtDNA in testicular cells dissected using laser-capture micro-dissection. The transcription levels of M type mtDNA were not significantly different between spermatogonia and spermatocytes versus spermatids and spermatozoa. Next, we examined differences in transcription levels of M type mtDNA between larvae from male-biased and female-biased mothers. The transcription levels of M type mtDNA significantly increased 24 and 48 h after fertilization in male-biased crosses. By contrast, transcription levels significantly decreased in female-biased crosses. These results suggest M type mtDNA may play a role in early germ line formation.

Relationship between two androgenetic clam species, Corbicula leana and Corbicula fluminea, inferred from mitochondrial cytochrome b and nuclear 28S rRNA markers.

Two shell color types, yellow (type I) and brown (type II), of hermaphrodite Corbicula fluminea clams from Ritto, Shiga Prefecture, Japan, are sympatric with both male and hermaphrodite Corbicula leana. In the present study, the mitochondrial DNA (mtDNA) cytochrome b and nuclear 28S rRNA genes of C. fluminea were sequenced to construct a haplotype network in order to investigate the genetic relationship with C. leana. Ninety C. fluminea samples revealed only two cytb haplotypes; the majority (97.8%) were CB7, while the remainder were CB1. In C. leana, only CB1 was detected in hermaphrodites, but both CB1 and CB7 were detected in males. Nuclear 28S rRNA haplotypes of C. fluminea type I individuals were divergent from those of hermaphrodite C. leana. However, C. fluminea type I clams shared haplotypes with male C. leana individuals, whereas C. fluminea type II individuals shared haplotypes with both hermaphrodite and male C. leana samples. These results suggest that it may be difficult to define a clear genetic border between these species.

28S rDNA haplotypes of males are distinct from those of androgenetic hermaphrodites in the clam Corbicula leana.

The clam Corbicula leana exists in two forms, hermaphrodites and males. Our previous study on mitochondrial DNA suggested that the male nuclear DNA might have derived from hermaphrodite C. leana relatively recently. To clarify the origin of males in the clam, sequences of the nuclear 28S rDNA divergent domain (which is 441-444 bp long) in androgenetic hermaphrodites and males and dioecious (bisexual) species were analyzed. Unexpectedly, the nuclear 28S rDNA haplotypes of males and hermaphrodites were distinct. Haplotype network analysis indicated that males and hermaphrodites are reproductively isolated from each other without sharing the same nuclear haplotype. These results support a hypothesis that the egg nuclear genome of androgenetic hermaphrodites is replaced by the male sperm genome, and only males develop after fertilization by a male spermatozoon.

Different transcriptional ratios of male and female transmitted mitochondrial DNA and tissue-specific expression patterns in the blue mussel, Mytilus galloprovincialis.

In some bivalve species, paternal mitochondrial DNA (mtDNA) from sperm is transmitted to the offspring. This is called "doubly uniparental inheritance" (DUI). Under DUI, male offspring receive both paternal (M type) and maternal (F type) mtDNA. Females predominantly receive F type. Expression levels of M and F type mtDNA and mitochondrial RNA localization have not been studied extensively. In this study, we quantified M and F type mtDNA and their expression levels in male and female somatic tissues and gonads with real-time polymerase chain reaction (PCR) in the blue mussel, Mytilus galloprovincialis. M and F type expression patterns were studied with in situ hybridization, using probes specific to M and F type mtDNA in the cytochrome b region. We found that (i) F type mtDNA was expressed in somatic tissues and female gonads, while M type was not expressed in these tissues; (ii) M type expression in male gonads was limited, but strong expression was observed during early spermatogenesis; and (iii) F type expression ratios were significantly lower in female gonads than in somatic tissues and lower than both M and F type expression ratios in male gonads. We propose (i) different systems for M and F type tissue-specific transcriptional regulation; and (ii) different functions for F and M type mtDNA, with F type being functional in somatic tissues and female gonads and M type functioning only in spermatogenetic cells.

Mitochondrial DNA copy number is maintained during spermatogenesis and in the development of male larvae to sustain the doubly uniparental inheritance of mitochondrial DNA system in the blue mussel Mytilus galloprovincialis.

Doubly uniparental inheritance (DUI) of mitochondrial (mt) DNA has been reported in the blue mussel Mytilus galloprovincialis. In DUI, males inherit both paternal (M type) and maternal (F type) mtDNA. Here we investigated changes in M type mtDNA copy numbers and mitochondrial mass in testicular cells by real-time polymerase chain reaction and flow cytometry. The ratios of M type mtDNA copy numbers to nuclear DNA content were not different between haploid (1n), diploid (2n) and tetraploid (4n) spermatogenic cells. The mitochondrial mass decreased gradually during spermatogenesis. These results suggest that mtDNA and mitochondrial mass are maintained during spermatogenesis. We then traced M type mtDNA in larvae after fertilization. M type mtDNA was maintained up to 24 h after fertilization in the male-biased crosses, but decreased significantly in female-biased crosses (predicted by Mito Tracker staining pattern). These results are strikingly different from those reported for mammals and fish, where it is well known that the mitochondria and mtDNA are reduced during spermatogenesis and that sperm mitochondria and mtDNA are eliminated soon after fertilization. Thus, the M type mtDNA copy number is maintained during spermatogenesis and in the development of male larvae to sustain the DUI system in the blue mussel.

Origin and possible role of males in hermaphroditic androgenetic Corbicula clams.

Hermaphroditic Corbicula leana clams reproduce by androgenesis and have been regarded as simultaneous hermaphrodites. To date, there has been no report on the occurrence of male clams in hermaphroditic Corbicula. In an irrigation ditch in Shiga Prefecture, we found that 78.2% of C. leana specimens were males and 21.8% were hermaphrodites. Microfluorometric analysis revealed that males were diploids and hermaphrodites were triploids. All males produced nonreductional and biflagellate spermatozoa. The sequence analysis of mitochondrial DNA (cytochrome b, 621 bp) for 31 specimens of C. leana showed that four male and nine hermaphrodites shared the same H2 mtDNA haplotype; H1 was detected from 17 males and H3 was detected from one hermaphrodite. Coexisting C. fluminea clams also have haplotypes H1 and H2. Phylogenetic tree by a neighborjoining method based on the partial sequence of cytochrome b revealed that the haplotypes (H1- 3) of C. leana were evidently different from those of dioecious C. sandai (S1 and S2) and C. japonica (J1 and J2). These results suggest that males may be derived from hermaphrodite C. leana clams. The role of males in hermaphroditic populations is unknown. However, if the spermatozoon from a male is able to fertilize an egg from a hermaphrodite and the nuclear genome of the egg is expelled as polar bodies, the sperm nucleus could form a zygote nucleus. This mode of reproduction would allow the replacement of the nuclear genome.

Comparison of expression patterns of shell matrix protein genes in the mantle tissues between high- and low-quality pearl-producing recipients of the pearl oyster, Pinctada fucata.

The production of a cultured pearl is the result of a complex interplay between the donor and recipient oysters. However, there is a paucity of information on the relationship between donor and recipient oyster gene expression patterns and pearl quality. Shell matrix proteins affect not only the formation of the shell, but also that of the pearls. We compared the gene expression patterns of five shell matrix proteins (msi60, nacrein, msi31, prismalin-14, and aspein) in the mantle edge (ME), which forms the prismatic layer, and the mantle center (MC), which forms the nacreous layer, between high- (HP) and low quality pearl- (LP) producing recipient oysters. After culturing for about two months, ME and MC tissues were collected from nine recipient oysters: four with HP, five with LP. In the ME, the average threshold cycle (ΔC(T)) for aspein was higher in HP than in LP (t-test, p = 0.03). Additionally, in the MC, the average ΔC(T) for msi60 was lower in HP than in LP (p = 0.06). This means the relative expression level of msi60 in the mantle of HP was higher than that of LP, and expression level of aspein in the mantle of HP was lower than that of LP. Pearl quality was closely related to the expression patterns of shell matrix protein genes of recipient oysters.

The proliferation and migration of immature germ cells in the mussel, Mytilus galloprovincialis: observation of the expression pattern in the M. galloprovincialis vasa-like gene (Myvlg) by in situ hybridization.

In bivalve, the distribution of primordial germ cells can be traced from early embryogenesis to the veliger larva by the expression of the vasa ortholog. However, the distribution of germ cells from metamorphosis to maturation in bivalves has not been examined extensively. In this study, we used in situ hybridization to observe expression of the Mytilus galloprovincialis vasa-like gene (Myvlg). The distribution of germ cells was clarified in immature mussels. We observed germ cells in adult mussels during the non-reproductive and reproductive seasons. Myvlg was specifically expressed in germ cells. Gametogenesis occurs in acini surrounded by connective tissue. Myvlg expression was detected in spermatogonia, spermatocytes, oogonia, and oocytes. In the non-reproductive season, gametes were not observed in the acini, but Myvlg was expressed in germinal stem cells along the acini. The expression intensity in the non-reproductive season, however, was much weaker than that in the reproductive season. Myvlg-positive cells proliferated during the non-reproductive season. In immature mussels, a pair of germ cell clumps was distributed laterally in the connective tissue between the nephric tubules and posterior byssal retractor muscle. Germ cells were also observed along pericardium. When immature mussels grew, a pair of germ cell clumps migrated anteriorly in the connective tissue along the outer epithelium at the dorsal region of the mantle base between the mantle and gill. The number of germ cells increased significantly as the mussels grew. This is the first report to observe the proliferation and migration of germ cells in immature mussels.

Mitochondrial DNA transmitted from sperm in the blue mussel Mytilus galloprovincialis showing doubly uniparental inheritance of mitochondria, quantified by real-time PCR.

Doubly uniparental inheritance (DUI) of mitochondrial DNA transmission to progeny has been reported in the mussel, Mytilus. In DUI, males have both paternally (M type) and maternally (F type) transmitted mitochondrial DNA (mtDNA), but females have only the F type. To estimate how much M type mtDNA enters the egg with sperm in the DUI system, ratios of M type to F type mtDNA were measured before and after fertilization. M type mtDNA content in eggs increased markedly after fertilization. Similar patterns in M type content changes after fertilization were observed in crosses using the same males. To compare mtDNA quantities, we subsequently measured the ratios of mtDNA to the 28S ribosomal RNA gene (an endogenous control sequence) in sperm or unfertilized eggs using a real-time polymerase chain reaction (PCR) assay. F type content in unfertilized eggs was greater than the M type in sperm by about 1000-fold on average. M type content in spermatozoa was greater than in unfertilized egg, but their distribution overlapped. These results may explain the post-fertilization changes in zygotic M type content. We previously demonstrated that paternal and maternal M type mtDNAs are transmitted to offspring, and hypothesized that the paternal M type contributed to M type transmission to the next generation more than the maternal type did. These quantitative data on M and F type mtDNA in sperm and eggs provide further support for that hypothesis.

Maternal inheritance of mitochondrial DNA (mtDNA) in the Pacific oyster (Crassostrea gigas): a preliminary study using mtDNA sequence analysis with evidence of random distribution of MitoTracker-stained sperm mitochondria in fertilized eggs.

In many bivalve species, paternal and maternal mitochondrial DNA (mtDNA) from sperm and eggs is transmitted to the offspring. This phenomenon is known as doubly uniparental inheritance (DUI). In these species, sperm mtDNA (M type) is inherited by the male gonad of the offspring. Egg mtDNA (F type) is inherited by both male and female somatic cells and female gonadal cells. In Mytilidae, sperm mitochondria are distributed in the cytoplasm of differentiating male germ cells because they are transmitted to the male gonad. In the present study, we investigated maternal inheritance of mtDNA in the Pacific oyster, Crassostrea gigas. Sequence analysis of two mitochondrial non-coding regions revealed an identical sequence pattern in the gametes and adductor muscle samples taken from six males and five females. To observe whether sperm mitochondria were specifically located in the cytoplasm of differentiating germ cells, their distribution was recorded in C. gigas fertilized eggs by vital staining with MitoTracker Green. Although the 1D blastomere was identified in the cytoplasm of differentiating germ cells, sperm mitochondria were located at the 1D blastomere in only 32% of eggs during the 8-cell stage. Thus, in C. gigas, sperm mitochondria do not specifically locate in the germ cell region at the 1D blastomere. We suggest that the distribution of sperm mitochondria is not associated with germ cell formation in C. gigas. Furthermore, as evidenced by the mtDNA sequences of two non-coding regions, we conclude that mitochondrial DNA is maternally inherited in this species.

A hypothesis of ploidy elevation by formation of a female pronucleus in the androgenetic clam Corbicula fluminea in the Tone River estuary, Japan.

We propose a hypothesis of ploidy elevation in the androgenetic clam Corbicula fluminea, based on an abnormal process of fertilization in clams collected at the Tone River, Ibaraki Prefecture, Japan. Most eggs showed androgenesis, that is, extrusion of all maternal chromosomes as two polar bodies during the first meiotic division. Most eggs did not form a female pronucleus, but only a male pronucleus. However, some eggs proceeded to the second meiosis and formed both a female and a male pronucleus. The formation of the female pronucleus suggests the hypothesis that ploidy elevation in androgenetic clams may have occurred by aberrant meiosis due to an altered orientation of the meiotic spindle.

Ultrastructural observations of eu- and paraspermiogenesis in the cottid fish Hemilepidotus gilberti (Teleostei: Scorpaeniformes: Cottidae).

The developmental process of eu- and paraspermatozoa in the cottid fish, Hemilepidotus gilberti, was observed by electron microscopy. Euspermatozoa of H. gilberti consist of a thin disk-like sperm head (about 3 microm in length), a short middle piece, and a long flagellum, but lack an acrosome. On the other hand, during spermiogenesis, aberrant spermatids, rich in cytoplasm and possessing binuclei, develop into cysts containing spermatids. The developing aberrant spermatids connect with normal spermatids and euspermatozoa by intercellular bridges. The early phase of chromatin condensation in aberrant spermatids is almost identical to that in normal spermatids, but the nuclei in the later phase develop into a mass of highly electron-dense globules. Since the aberrant spermatids complete karyokinesis but not cytokinesis at telophase of the second meiotic division, they are considered to develop into hyperpyrenic cells due to incomplete cytokinesis of the second meiotic division. These spermatids are oval in shape (5-7 microm in diameter) and lack a flagellum. The aberrant spermatids of H. gilberti are shed along with euspermatozoa and amount to about 50% of semen in volume. Judging from their form and developmental process, aberrant spermatids produced in H. gilberti are considered hyperpyrenic paraspermatozoa.

Androgenetic reproduction in a freshwater diploid clam Corbicula fluminea (Bivalvia: Corbiculidae).

Two shell color types of the exotic bivalve Corbicula fluminea were collected in Kyoto city, Japan. DNA microfluorometry revealed that both types were diploids with non-reductional spermatozoa. Maternal chromosomes were found to be extruded as two polar bodies at the first meiosis, and the second meiosis could not be observed. Only the male pronucleus was present in the egg cytoplasm and became metaphase chromosomes at the first mitosis. The present study indicates that the diploid C. fluminea in Japan has the same mode of androgenetic reproduction as the triploid C. leana.

Pearl microstructure and expression of shell matrix protein genes MSI31 and MSI60 in the pearl sac epithelium of Pinctada fucata by in situ hybridization.

Expression patterns of the shell matrix protein genes MSI31 and MSI60 in the pearl sac epithelium were examined by in situ hybridization 38 days after implantation, and related to pearl quality. A pearl sac that produced a nacreous pearl showed very weak expression of MSI31 and strong expression of MSI60. A pearl sac, which yielded a prismatic pearl, strongly expressed MSI31 and very weakly expressed MSI60. In a complex pearl, whose surface consisted of a mosaic of both nacreous and prismatic layers, the expression pattern of MSI31 and MSI60 similarly corresponded to the underlying surface structures of the pearl. A nacreous pearl whose pearl sac showed strong MSI31 expression had an entirely nacreous surface composed of a laminar structure with unusual tablet growth at the corresponding site. MSI31 and MSI60 are the major components of the shell matrix proteins of the nacreous and prismatic layers. Clearly, high expression of MSI31 does not always result in prismatic secretion. These observations cannot be explained solely on the basis of the expression patterns of MSI31 and MSI60. We propose that, in addition to the MSI genes that form the prismatic and nacreous layers, upstream from these genes there are regulatory master genes that determine whether a nacreous layer (aragonite) or a prismatic layer (calcite) is formed.

Gene expression patterns in the outer mantle epithelial cells associated with pearl sac formation.

For pearl culture, nucleus and mantle grafts are implanted into the gonad of the host oyster. The epithelial cells of the implanted mantle graft elongate and surround the nucleus, and a pearl sac is formed. Shell matrix proteins secreted by the pearl sac play an important role in pearl formation. We studied the gene expression patterns of six shell matrix proteins (msi60, n16, nacrein, msi31, prismalin-14, and aspein) in the epithelial cells associated with pearl sac formation. There were differences in the expression patterns of the six genes in the epithelial cells, and the relative expression levels for msi60 and aspein differed between the mantle graft and pearl sac (48 days after implantation). Therefore, the gene expression patterns of the epithelial cells were genetically undetermined, and changed between before and after pearl sac formation. The gene expression patterns of the epithelial cells of the pearl sac may be regulated by the host oysters.

Can the quality of pearls from the Japanese pearl oyster (Pinctada fucata) be explained by the gene expression patterns of the major shell matrix proteins in the pearl sac?

For pearl culture, the pearl oyster is forced open and a nucleus is implanted into the gonad with a mantle graft. The outer mantle epithelial cells of the implanted mantle graft elongate and surrounding the nucleus a pearl sac is formed. Shell matrix proteins secreted by the pearl sac play an important role in the regulation of pearl formation. Recently, seven shell matrix proteins were identified from the pearl oyster Pinctada fucata. However, there is a paucity of information on the function of these proteins and their gene expression patterns. Our study aims to elucidate the relationship between pearl type, quality, and gene expression patterns of six shell matrix proteins (msi60, n16, nacrein, msi31, prismalin-14, and aspein) in the pearl sac based on real-time PCR analysis. After culturing for about 2 months, the pearl sac tissues were collected from 22 individuals: 12 with high quality (HP), nine with low quality (LP), and one with organic (ORG) pearl formation. The surface of each of the 12 HP pearls was composed only of a nacreous layer; in contrast, that of the nine LP pearls was composed of nacreous and prismatic layers. The six target gene expressions were detected in all individuals. However, delta threshold cycle (ΔC(T)) for msi31 was significantly higher in the HP than in the LP individuals (Mann-Whitney's U test, p=0.02). This means that the relative expression level of msi31, which constitutes the framework of the prismatic layer, was higher in the LP than in the HP individuals.

Quantitation of the male and female types of mitochondrial DNA in a blue mussel, Mytilus galloprovincialis, using real-time polymerase chain reaction assay.

The system termed doubly uniparental inheritance (DUI) of mitochondrial transmission to progeny has been reported in Mytilus. Under DUI, it has been thought that males have both paternally (M type) and maternally (F type) transmitted mitochondrial DNA (mtDNA), and females have only F type. However, the presence of M type in females has been reported. To clarify the ratio of M type to F type mtDNA in female and male tissues to further our understanding of mitochondrial transmission, we developed a procedure to measure the copy numbers of the two types of mtDNA in Mytilus galloprovincialis using a real-time polymerase chain reaction assay. The following results were obtained by this method. In females, the copy numbers of M type mtDNA detected in adductor muscle, gonad and eggs were approximately 10 000-fold lower than those of F type. In males, F type dominated in adductor muscle, as in the female tissue. However, copy numbers of M type mtDNA were approximately 1000-fold higher than those of F type in gonad and 100 000-fold higher than those of F type in sperm. We examined the quantity relationship between the two types of mtDNA and the transmission mechanism of mtDNA in M. galloprovincialis.

Inheritance of two M type mitochondrial DNA from sperm and unfertilized eggs to offspring in Mytilus galloprovincialis.

In Mytilus mussels, paternal mitochondrial DNA (mtDNA) from sperm is known to be transmitted to offspring. This phenomenon is called doubly uniparental inheritance (DUI). Under DUI, sperm mtDNA (M type) is inherited only by males. Female mussels receive maternal mtDNA (F type). However, in our previous study, we showed female and unfertilized eggs have both F and M types. We hypothesized that the two M types both from sperm and unfertilized eggs were transmitted to offspring. To test the hypothesis, we examined the number of M type haplotypes in mature M. galloprovincialis. The M type in larvae was compared with those of the parents. Cross experiments were carried out to test the inheritance of M type. In six of 20 mature mussels, two M types were detected by sequence analysis and polymerase chain reaction-restriction fragment length polymorphism. In cross experiments of larval samples from five of 12 crosses, double peak wave was observed by single nucleotide polymorphisms analysis. In these larval samples, the higher peak wave was identical to the parental M type. Larvae received much more paternal M type than the maternal ones. We demonstrated that two M types from sperm and unfertilized eggs were transmitted to offspring in M. galloprovincialis.

Effects of cryopreservation methods on post-thaw motility of spermatozoa from the Japanese pearl oyster, Pinctada fucata martensii.

In order to develop cryopreservation techniques for Japanese pearl oyster spermatozoa, the effects of various cryopreservation conditions on post-thaw motility were examined. Spermatozoa cryopreserved with 10% methanol (MET), dimethylformamide or dimethylacetamide plus 90% diluent comprising 80% seawater and 20% fetal bovine serum (FBS) showed higher percentages of post-thaw motility than those cryopreserved with 10% dimethylsulfoxide or glycerol. When spermatozoa were cryopreserved with various concentrations (0-20%) of MET and 100-80% diluent, 10% MET showed the highest percentages of post-thaw motility. When spermatozoa were cryopreserved with 10% MET and 90% diluent comprising various concentrations (0-100%) of FBS or Ringer solution mixed with seawater, the percentages of post-thaw motility peaked at 20% FBS or Ringer solution, and were significantly higher for 20% FBS than for 20% Ringer solution. The percentages of post-thaw motility increased with increasing dilution ratios from 2.5- to 50-fold. Spermatozoa cooled to -50 degrees C and then immersed in liquid nitrogen (LN) showed higher post-thaw motility than those cooled to -30 degrees C or -40 degrees C. When spermatozoa were cryopreserved to -50 degrees C at various cooling rates by changing the sample height above the LN surface, the post-thaw motilities of spermatozoa cooled at 10 cm (cooling rate: -21.3 degrees C/min) and 12.5 cm (-15.6 degrees C/min) from the LN surface were higher than those at 5, 7.5 or 15 cm. These results indicate that 10% MET plus 90% diluent comprising 80% seawater and 20% FBS is a suitable extender for cryopreservation of Japanese pearl oyster spermatozoa and that samples should be cooled to -50 degrees C at a cooling rate between -15 and -20 degrees C/min for efficient storage.