Risk of equine infectious anemia virus disease transmission through in vitro embryo production using somatic cell nuclear transfer

Prevention and regulation of equine infectious anemia virus (EIAV) disease transmission solely depend on identification, isolation, and elimination of infected animals because of lack of an effective vaccine. Embryo production via the somatic cell nuclear transfer (SCNT) technology uses oocytes collected mainly from untested animals, which creates a potential risk of EIAV transmission through infected embryos. The current review examines the risk of EIAV disease transmission through SCNT embryo production and transfer. Equine infectious anemia virus is a lentivirus from the family Retroviridae. Because of a lack of direct reports on this subject, relevant information gathered from close relatives of EIAV, such as human immunodeficiency virus (HIV), bovine immunodeficiency virus (BIV), feline immunodeficiency virus (FIV), and small ruminant lentiviruses (SRLVs), is summarized and used to predict the biological plausibility of EIAV disease transmission through transfers of the equine SCNT embryos. Based on published information regarding interaction of oocytes with lentiviruses and the sufficiency of oocyte and embryo washing procedures to prevent lentivirus transmission from in vitro-produced embryos of various species, we predicted the risk of EIAV transmission through SCNT embryo production and transfer to be very small or absent.     

Effects of vascular endothelial growth factor on porcine preimplantation embryos produced by in vitro fertilization and somatic cell nuclear transfer

This study examined the effects of vascular endothelial growth factor (VEGF) on porcine embryos produced by in vitro fertilization (IVF) and somatic cell nuclear transfer (SCNT) at different developmental stages. Four sets of experiments were performed. In the first, supplementation of the in vitro culture medium with 5 ng/mL VEGF was suitable for porcine IVF embryo development, and the blastocyst formation rate was significantly higher than the control and other groups (57.73 ± 6.78% (5 ng/mL VEGF) vs. 43.21 ± 10.22% (control), 42.16 ± 10.24% (50 ng/mL VEGF) and 41.91 ± 11.74% (500 ng/mL VEGF); P < 0.05). The total cell number after supplementation with 5 ng/mL VEGF was significantly higher than the control and other groups (151.85 ± 39.77 (5 ng/mL VEGF) vs. 100.00 ± 34.43 (control), 91.2 ± 31.51 (50 ng/mL VEGF), and 112.53 ± 47.66 (500 ng/mL VEGF); P < 0.05). In the second experiment, when VEGF was added at different developmental stages of IVF derived embryos (early stage, days 1-3, late stage, days 4-7), the blastocyst formation rate and total cell number were significantly higher at the late stage (47.71 ± 9.13% and 131.5 ± 20.70, respectively) than in the control (34.32 ± 7.44% and 85.50 ± 20.41, respectively) and at the early stage (33.60 ± 5.78% and 86.75 ± 25.10, respectively; P < 0.05). There was no significant difference in the blastocyst development rate or total cell number between the whole culture period (days 1-7) and the late stage culture period after supplementation with 5 ng/mL VEGF (P > 0.05). In the third experiment, the cleavage rate was significantly higher when SCNT embryos were cultured with VEGF during the whole culture period than in the late stage (63.56 ± 15.52% vs. 39.72 ± 4.94%; P < 0.05), but there was no significant difference between the control and the early stage culture period (P > 0.05). The blastocyst formation rate was significantly higher at the late stage culture period with VEGF than at the early stage culture period (34.40 ± 15.06% vs. (16.07 ± 5.01%; P < 0.05). There was no significant difference in the total cell number between the groups (P > 0.05). In experiment 4, using real-time PCR, VEGF mRNA expression was detected in all the developmental stages of IVF and SCNT embryos, but the expression level varied according to the developmental stage. VEGF receptor, KDR mRNA was detected in all stages IVF and SCNT embryos. However, flt-1 mRNA was not expressed in all embryonic stages of IVF and SCNT embryos. These data suggest that VEGF supplementation at the late embryonic developmental stage might improve the developmental potential of both IVF and SCNT preimplantation porcine embryos through its receptors.     

Meta-analysis of melatonin treatment and porcine somatic cell nuclear transfer embryo development

Porcine somatic cell nuclear transfer (SCNT) plays an important role in many areas of research. However, the low efficiency of SCNT in porcine embryos limits its applications. Porcine embryos contain high concentrations of lipid, which makes them vulnerable to oxidative stress. Some studies have used melatonin to reduce reactive oxygen species damage. At present there are many reports concerning the effect of exogenous melatonin on porcine SCNT. Some studies suggest that the addition of melatonin can increase the number of blastocyst cells, while others indicate that melatonin can reduce the number of blastocyst cells. Therefore, a meta-analysis was carried out to resolve the contradiction. In this study, a total of 63 articles from the past 30 years were analyzed, and six papers were finally selected. Through the analysis, it was found that the blastocyst rate was increased by adding exogenous melatonin. Melatonin had no effect on cleavage rate or the number of blastocyst cells, but did decrease the number of apoptotic cells. This result is crucial for future research on embryo implantation.     

Anti-apoptotic effect of insulin-like growth factor (IGF)-I and its receptor in porcine preimplantation embryos derived from in vitro fertilization and somatic cell nuclear transfer

Insulin-like growth factor (IGF)-I is a receptor-mediated autocrine and/or paracrine growth and/or survival factor for mammalian embryo development. It is known to promote the growth and development of mouse preimplantation embryos. The present study was designed to investigate the effects of IGF-I (50 ng/ml), anti-IGF-I receptor antibody (50 ng/ml) and their combination on porcine preimplantation embryo development. Furthermore, the mechanism underlying the embryotropic effects of IGF-I was evaluated by monitoring the incidence of apoptosis and expression of apoptosis-related genes. In both in vitro fertilized (IVF) and somatic cell nuclear transfer (SCNT) embryos, culturing with IGF-I increased the rate of blastocyst formation and this embryotrophic effect was neutralized by culturing with IGF-I along with anti-IGF-I receptor (IGF-IR) antibody. Culturing IVF and SCNT embryos with IGF-I significantly increased the number of total cells in blastocysts and decreased the number of apoptotic nuclei. These effects of IGF-I were also neutralized by culturing with IGF-I along with anti-IGF-IR antibody. Expression of the anti-apoptotic Bcl-2 gene was increased, while expression of the pro-apoptotic Bax was decreased in both IVF and SCNT embryos cultured with IGF-I. In both IVF and SCNT embryos, anti-IGF-IR antibody along with IGF-I neutralized the effect of IGF-I on expression of Bcl-2 and Bax genes. In conclusion, the present study demonstrated that IGF-I through its specific receptors improved the developmental competence of IVF and SCNT embryos by decreasing the incidence of apoptosis and regulating apoptosis-related genes in porcine preimplantation embryos.     

Predictive variables of in vitro fertilization and pre-implantation embryo development in the mouse

The present study aims to analyze the cause-effect relationships among several in-vitro fertilization and pre-implantation embryo development variables in the mouse. Superovulation of hybrid (C57Bl/6JIco female X CBA/JIco male) female mice of 4-6 weeks of age was induced by a priming injection of pregnant mare's serum gonadotropin at the estrus stage of the estrous cycle followed after a 48-hr interval by human chrorionic gonadotropin. Ovulated cumulus-enclosed oocytes were inseminated with sperm from hybrid males of 12-16 weeks of age. The multiple linear regression analyses performed indicated that (a) total number of ovulated oocytes is a good predictor of both fertilization frequency and total number of cells in day-5 blastocysts; (b) fertilization frequency predicts percentage of day-5 blastocysts; (c) total number of cells in day-5 blastocysts is predicted by percentage of day-5 blastocysts; and (d) total number of cells in day-5 blastocysts predicts percentage of apoptotic cells, number of inner cell mass (ICM) and trophectoderm (TE) cells, and ICM/TE ratio in day-5 blastocysts. Mitotic index in day-5 blastocysts was positively correlated with total number of ovulated oocytes, percentage of ovulated cumulus-enclosed oocytes, fertilization frequency, percentage of day-5 blastocysts and total number of cells in day-5 blastocysts. On the contrary, it was negatively correlated with percentage of apoptotic cells in day-5 blastocysts.     

Aberrant DNA methylation in porcine in vitro-, parthenogenetic-, and somatic cell nuclear transfer-produced blastocysts

Early embryonic development in the pig requires DNA methylation remodeling of the maternal and paternal genomes. Aberrant remodeling, which can be exasperated by in vitro technologies, is detrimental to development and can result in physiological and anatomic abnormalities in the developing fetus and offspring. Here, we developed and validated a microarray based approach to characterize on a global scale the CpG methylation profiles of porcine gametes and blastocyst stage embryos. The relative methylation in the gamete and blastocyst samples showed that 18.5% (921/4,992) of the DNA clones were found to be significantly different (P < 0.01) in at least one of the samples. Furthermore, for the different blastocyst groups, the methylation profile of the in vitro-produced blastocysts was less similar to the in vivo-produced blastocysts as compared to the parthenogenetic- and somatic cell nuclear transfer (SCNT)-produced blastocysts. The microarray results were validated by using bisulfite sequencing for 12 of the genomic regions in liver, sperm, and in vivo-produced blastocysts. These results suggest that a generalized change in global methylation is not responsible for the low developmental potential of blastocysts produced by using in vitro techniques. Instead, the appropriate methylation of a relatively small number of genomic regions in the early embryo may enable early development to occur.     

Pregnant mares' serum gonadotropin source influences fertilization and fresh or thawed embryo development,but the effect is genotype specific

The influence of the source of pregnant mares' serum gonadotropin (PMSG) on the num ber, quality, and in vitro development of mouse embryos before and after freezing was evaluated among three genotypes: N:NIH(S), C57BL/6N, and C3H/HeN-MTV^?. Immature females were given PMSG from one of five commercial sources. Following col lection ( 116 hr later), embryos were evaluated for stage of development, and four-to eight-cell embryos were pooled within genotype and assigned to standardized fresh or freeze-thaw culture trials. Different PMSG sources stimulated the production of different num bers of total embryos (P < 0.05) but not necessarily more embryos suitable for freezing. Differences in embryo production among genotypes indicated that absolute embryo num bers using a single mouse genotype may not accurately reflect the potency of a specific gonadotropin source. The PMSG source also affected the ability of an embryo to survive in culture either immediately after collection or after frozen storage. The effect, however, was genotype specific, with some mouse strains being relatively insensitive to PMSG source, whereas gonadotropin source played a major role in determining in vitro viability in others. Development rates for freshly collected embryos differed, often inconsistently, from those of thawed embryos regardless of the PMSG source used, demonstrating that fresh embryo development cannot be used to estimate expected post-thaw survival. In vitro development of thawed embryos is influenced not only by genotype, but also the source of the gonadotropin used to promote follicular development and oocyte maturation. These findings may explain, in part, the wide variation in embryo viability and culture rates reported among laboratories and intraspecies animal populations.     

Presence of epidermal growth factor during in vitro maturation of pig oocytes and embryo culture can modulate blastocyst development after in vitro fertilization

The present study examined the effect of epidermal growth factor (EGF) during in vitro maturation (IVM) and embryo culture on blastocyst development in the pig. In experiment 1, cumulus oocyte complexes were cultured in North Carolina State University (NCSU) 23 medium containing porcine follicular fluid, cysteine, hormonal supplements, and with or without EGF (0–40 ng/ml) for 20–22 hr. They then were cultured for an additional 20–22 hr without hormones. After maturation, cumulus-free oocytes were co-incubated with frozen-thawed spermatozoa for 5–6 hr. Putative embryos were transferred to NCSU 23 containing 0.4% BSA and cultured for 144 hr. In experiment 2, oocytes were matured in medium containing 10 ng/ml EGF, inseminated, and putative embryos were cultured in the presence of 0–40 ng/ml EGF. In experiment 3, oocytes were cultured in the presence of 0, 10 and 40 ng/ml EGF to examine the kinetics of meiotic maturation. In experiment 4, 2- to 4-cell and 8-cell to morula stage embryos derived from oocytes matured with 10 ng/ml EGF were transferred to the oviduct and uterus, respectively, of each of three recipient gilts (3 and 4 days post-estrus, respectively). The presence or absence of EGF during IVM did not affect cumulus expansion, nuclear maturation, fertilization parameters, or cleavage rate. However, compared to no addition (21%), presence of 1 (33%) and 10 ng/ml EGF (42%) during IVM increased (P < 0.01) the rate of blastocyst development in a concentration-dependent manner. Compared to 10 ng/ml EGF, higher concentrations (20 and 40 ng/ml) reduced (P < 0.01) blastocyst development in a concentration-dependent manner (35% and 24%, respectively). No difference was observed between no addition and 40 ng/ml EGF (22%). Compared to no addition and 10 ng/ml EGF, a significantly (P < 0.001) higher proportion (25% vs. 55%) of oocytes reached metaphase II stage 33 hr after IVM with 40 ng/ml EGF. However, no difference was observed at 44 hr. Transfer of embryos to six recipient gilts resulted in three pregnancies and birth of 18 piglets. The results show that EGF at certain concentrations in IVM medium can influence the developmental competence of oocytes. However, addition of EGF during the culture of pig embryos derived from oocytes matured in the presence of EGF is without effect. Birth of piglets provides evidence that embryos derived from oocytes matured in a medium containing EGF are viable. Mol. Reprod. Dev. 51:395–401, 1998. © 1998 Wiley-Liss, Inc.     

TSA和VPA处理对食蟹猴-猪异种体细胞核移植胚胎早期发育的影响

食蟹猴-猪异种体细胞核移植(Interspecies somatic cell nuclear transfer,iSCNT)研究旨在由iSCNT胚胎建立具有与人类相似遗传背景的胚胎干细胞(ESCs),作为医学和基础科学研究的实验材料。文章探讨了两种组蛋白脱乙酰化酶抑制剂(HDACi)Trichostatin A(TSA)和Valproic acid(VPA)处理浓度、时间与培养液(PZM-3和HECM-10)组合对食蟹猴-猪iSCNT胚胎早期发育的影响。结果表明,在PZM-3中添加10 nmol/L TSA处理48 h组的囊胚率显著高于对照组(22.78%vs 9.86%,P<0.05)。但是,不管在PZM-3或是HECM-10中,添加2～10mmol/L VPA处理均不能提高iSCNT胚胎早期发育能力。文章证明了TSA处理可以提高食蟹猴-猪iSCNT胚胎早期发育能力。     

影响猪体细胞核移植重构胚体外发育的若干因素

以卵丘细胞为核供体细胞组成重构胚,卵裂率达到56．7％,发育至桑椹胚达11．7％、孵化囊胚率为6．7％,显著高于成纤维细胞组成的重构胚（P＜0．05）。我们研究了卵母细胞的采集方法,激活方法和卵龄对卵丘细胞核移植重构胚体外发育的影响。以血清饥饿法将卵丘细胞诱导至G0或G1期,抽吸法／解剖法采集卵母细胞,体外培养33或44h,将卵丘细胞置于去核卵母细胞的卵周隙中,重构胚以钙离子载体A23817或电泳冲结合6－DMAP激活处理,体外培养6天,结果表明,卵 母细胞采集方法、激活液中细胞松弛素（CB）并不影响重构胚的发育（以卵龄44h的卵母细胞为受体）;而以电脉冲结合6－DMAP激活处理能提高重构胚发育能力（以卵龄33h的卵母细胞为受体）（P＜0．05）。本研究显示,以电脉冲结合6－DMAP激活卵丘细胞重构胚,能在体外发育至囊胚。     

Production of porcine cloned transgenic embryos expressing green fluorescent protein by somatic cell nuclear transfer

In the present study, nuclear transferred embryos (NTEs) were reconstructed by using pig fetal fibroblasts as donors and in vitro matured oocytes as recipients. The effects of G418 selection on donor cells, duration of IVM of prepubertal gilt oocytes and oxygen tension in IVM of oocytes were investigated. The results were as follows: (i) When G418 selected cells expressing GFP were used as donors, the cleavage rate of NTEs decreased drastically in comparison to NTEs derived from donors without antibiotic selection (47.5% vs. 71.6%, p<0.05). For the blastocyst rate, no significant difference was observed between two groups (10% vs. 10.4%, p>0.05). (ii) The rate of nuclear maturation of oocytes increased significantly when IVM duration time was extended from 36 to 42 h (83.6% vs. 96.7%, p<0.05). However, no statistical difference was observed between NTEs derived from oocytes of 36 h IVM group and NTEs from oocytes of 42 h IVM group in the rates of cleavage (59.3% vs. 73.6%, p>0.05) and blastocyst formation (9.3% vs. 13.2%, p>0.05); (iii) no significant difference was observed between NTEs reconstructed from oocytes matured under lower oxygen (7% O₂) tension and NTEs derived from oocytes matured under higher oxygen tension (20% O₂) in cleavage rate (70.6% vs. 67.1%, p>0.05) and blastocyst rate (11.8% vs. 12.3%, p>0.05). These results suggest that: (i) G418 selection does not have a significant effect on cleavage rate of NTEs expressing GFP. (ii) Nuclear maturation is greatly improved by prolonging IVM duration from 36 to 42 h, while no significant differences were observed for developmental potential of transgenic embryos. Thus IVM 42 h is the better choice in order to obtain maximum number of MII oocytes as recipients. (iii) Lower oxygen tension and higher oxygen tension in IVM have no significant effect on development of cloned embryos.     

Development of vitrified-thawed bovine oocytes after in vitro fertilization and somatic cell nuclear transfer

Cryopreservation could be a useful technique for providing a steady source of oocytes for nuclear transfer and in vitro embryo production. The purpose of this study was to develop a method for cryopreservation of bovine oocytes while maintaining the developmental potential following subsequent in vitro fertilization (IVF) or somatic cell nuclear transfer (SCNT). Following vitrification-thawing, the surviving oocytes were (a) used for parthenogenetic activation, (b) examined for pronuclear formation after IVF, (c) examined for embryo development after IVF, and (d) used for SCNT employing fetal fibroblasts transfected with green fluorescent protein (GFP) gene. While most of the oocytes survived vitrification when the microdrop method was used (92.50%), the cleavage and blastocyst formation rates after parthenogenetic activation were lower (46.5% and 11.1%) than that in the non-vitrified control (86.6% and 13.5%). After IVF, the pronuclear formation (2PN) of fertilized embryos was lower in the vitrified group than in the control (21.7% and 59.9%). After SCNT, fusion rates were similar in control (58.33%) and vitrified-thawed oocytes (53.19%). However, the cleavage (73.1% and 46.3%) and blastocyst formation rates (22.2%, 7.4%; p<0.05) differed between control and vitrified-thawed oocytes. In vitrified-thawed or control oocytes, all embryos reconstructed using fetal fibroblasts transfected with GFP gene showed GFP expression. To evaluate the complete developmental potential, embryos derived from vitrified-thawed and fresh control oocytes were non-surgically transferred to 27 recipients (16 for control and 11 for vitrified-thawed). In the vitrified-thawed group, two pregnancies were detected at day 60, and one of them lasted until day 222. While in the fresh group, one pregnancy maintained to term. In conclusion, vitrified-thawed bovine oocytes could support development into the subsequent stages after IVF and SCNT. In addition, this study showed the possibility of the vitrified-thawed bovine oocytes in the production of transgenic cloned animals. In addition, further studies are required to increase the efficiency of oocyte vitrification for the practical uses and production of live offspring.     

Production of porcine cloned transgenic embryos expressing green fluorescent protein by somatic cell nuclear transfer

In the present study, nuclear transferred embryos (NTEs) were reconstructed by using pig fetal fibroblasts as donors and in vitro matured oocytes as recipients. The effects of G418 selection on donor cells, duration of IVM of prepubertal gilt oocytes and oxygen tension in IVM of oocytes were investigated. The results were as follows: (i) When G418 selected cells expressing GFP were used as donors, the cleavage rate of NTEs decreased drastically in comparison to NTEs derived from donors without antibiotic selection (45.7% vs. 71.6%, p<0.05). For the blastocyst rate, no significant difference was observed between two groups (10% vs. 10.4%, p>0.05). (ii) The rate of nuclear maturation of oocytes increased significantly when IVM duration time was extended from 36h to 42h (83.6% vs. 96.7%, p<0.05). However, no statistical difference were observed between NTEs derived from oocytes of 36 h IVM group and NTEs from oocytes of 42 h IVM group in the rates of cleavage (59.3% vs. 73.6%, p>0.05) and blastocyst formation (9.3% vs. 13.2%, p>0.05); (iii) no significant difference was observed between NTEs reconstructed from oocytes matured under lower oxygen (7% O₂) tension and NTEs derived from oocytes matured under higher oxygen tension (20% O₂) in cleavage rate (70.6% vs. 67.1%, p>0.05) and blastocyst rate (11.8% vs. 12.8%, p>0.05). These results suggest that: (i) G418 selection does not have a significant effect on cleavage rate of NTEs expressing GFP. (ii) Nuclear maturation is greatly improved by prolonging IVM duration from 36 to 42 h, while no significant differences were observed for developmental potential of transgenic embryos. Thus IVM 42 h is the better choice in order to obtain maximum number of MII oocytes as recipients. (iii) Lower oxygen tension and higher oxygen tension in IVM have no significant effect on development of cloned embryos.     

Production of porcine cloned transgenic embryos expressing green fluorescent protein by somatic cell nuclear transfer

Somatic cell nuclear transfer (SCNT), combined with genome modification techniques, is a very pow-erful tool for agriculture, medicine and fundamental research on basic biological mechanisms. The effi-ciency of producing transgenic animals is greatly prom…     

Differential cytoplast requirement for embryonic and somatic cell nuclear transfer in cattle

Effective activation of a recipient oocyte and its compatibility with the nuclear donor are critical to the successful nuclear reprogramming during nuclear transfer. We designed a series of experiments using various activation methods to determine the optimum activation efficiency of bovine oocytes. We then performed nuclear transfer (NT) of embryonic and somatic cells into cytoplasts presumably at G1/S phase (with prior activation) or at metaphase II (MII, without prior activation). Oocytes at 24 hr of maturation in vitro were activated with various combinations of calcium ionophore A23187 (A187) (5 microM, 5 min), electric pulse (EP), ethanol (7%, 7 min), cycloheximide (CHX) (10 micro g/ml, 6 hr), and then cultured in cytochalasin D (CD) for a total of 18 hr. Through a series of experiments (Exp. 1-4), an improved activation protocol (A187/EP/CHX/CD) was identified and used for comparison of NT efficiency of embryonic versus somatic donor cells (Exp. 5). When embryonic cells from morula and blastocysts (BL) were used as nuclear donors, a significantly higher rate of blastocyst development from cloned embryos was obtained with G1/S phase cytoplasts than with MII-phase cytoplasts (36 vs. 11%, P < 0.05). In contrast, when skin fibroblasts were used as donor cells, the use of an MII cytoplast (vs. G1/S phase) was imperative for blastocyst development (30 vs. 6%, P < 0.05). Differential staining showed that parthenogenetic, embryonic, and somatic cloned BL contained 26, 29, and 33% presumptive inner cell mass (ICM) cells, respectively, which is similar to that of frozen-thawed in vivo embryos at a comparable developmental stage (23%). These data indicate that embryonic and somatic nuclei require different recipient cytoplast environment for remodeling/ reprogramming, and this is likely due to the different cell cycle stage and profiles of molecular differentiation of the transferred donor nuclei.     

Antiapoptotic and embryotrophic effects of alpha-tocopherol and L-ascorbic acid on porcine embryos derived from in vitro fertilization and somatic cell nuclear transfer

The objective was to determine optimal concentrations of alpha-tocopherol and l-ascorbic acid for development of porcine embryos derived from in vitro-fertilization (IVF) or somatic cell nuclear transfer (SCNT). The frequency of blastocyst formation in IVF embryos was 17.6, 28.6, 32.4 and 21.4% for control, 50, 100 and 200microM alpha-tocopherol, respectively, whereas in SCNT embryos, the frequency was 12.8, 19.0, 24.8 and 17.7% for corresponding concentrations of alpha-tocopherol. For both IVF and SCNT embryos, there were significantly more cells in blastocysts and the embryos had greater developmental competence when the embryo culture medium was supplemented with 100microM alpha-tocopherol. Although either alpha-tocopherol or l-ascorbic acid reduced the proportion of apoptotic cells in blastocysts, in combination they resulted in rates of apoptosis that were similar to the control group. For IVF embryos, the apoptotic index was 0.09 and 0.11 for alpha-tocopherol or l-ascorbic acid at a concentration of 100microM, respectively. Conversely, when these antioxidants were supplemented together, the apoptotic index increased significantly and was similar to the control group (i.e., 0.17 and 0.21 for combined and control group). For SCNT embryos, the apoptotic index was 0.10 for 100microM for both alpha-tocopherol and l-ascorbic acid, whereas the index was 0.23 and 0.17 for control and combined group. In conclusion, we recommend supplementing porcine embryo culture medium with 100microM alpha-tocopherol or 100microM l-ascorbic (as a second choice).