Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
949 Development 110, 949-954 (1990) Printed in Great Britain © The Company of Biologists Limited 1990 The development of handed asymmetry in aggregation chimeras of situs inversus mutant and wild-type mouse embryo NIGEL A. BROWN1 *, AFSHAN McCARTHY1 and LEWIS WOLPERT2 l MRC Experimental Embryology and Teratology Unit, Saint George's Hospital Medical School, Cranmer Terrace, London, SW17 ORE, UK Department of Anatomy and Developmental Biology, University College and Middlesex Hospital School of Medicine, Windeyer Building, Cleveland Street, London, W1P 6DB, UK 2 * Author for correspondence Summary Mutant iv/iv mice develop as if they have no sense of left and right, so the development of asymmetry is random: half normal, half as a mirror-image of normal, situs inversus. We have made aggregation chimeras of 8-cell stage iv/iv and + / + embryos, transferred them into pseudopregnant mice, and examined their phenotype on day 10 of gestation. The contribution of mutant and wild-type cells to tissues of the embryo was estimated by strain-specific isozyme (GPI-1) analysis. We have also performed reciprocal embryo transfers, iv/iv blastocysts into + / + mice, and vice versa. These transfers show that the development of handed asymmetry is determined by embryonic genotype, and is unaffected by the maternal environment (at least after day 3), or by the procedures of embryo collection, culture and transfer. Our observations on the development of 21 viable chimeric embryos show that neither iv/iv nor + / + cells are dominant. All embryos (12) with less than 50% contribution of iv/iv cells to the heart developed with normal situs. Of 9 embryos with greater than 50 % iv/iv cells, only 2 developed with inverted situs. These findings suggests that there was partial 'rescue' of embryos by some influence of normal over mutant cells. However, we cannot, statistically, exclude an alternative interpretation that cells are behaving autonomously. Interestingly, the embryos that developed with inverted situs were unique in having greater than two thirds contribution of iv/iv cells to both the heart and the visceral yolk-sac. Introduction of left and right, so their asymmetry is not handed, but random (Layton, 1976). In our terms, the conversion or biasing process is faulty and does not interact with the random generation of asymmetry; thus, half the embryos develop with normal visceral situs, half with a mirror-image pattern: situs inversus (Hummel and Chapman, 1959). There are also other abnormalities in iv/iv mice, including isomerisms, heterotaxias, and heart malformations, and we have discussed these in relation to our model (Brown and Wolpert, 1990). There is good evidence that each organ has its asymmetry specified independently. In this study, we have attempted to answer two questions: does the maternal uterine environment play a role in the development of handed asymmetry in the mouse? And, primarily, would the presence of some normal cells in an iv/iv embryo 'rescue' a chimeric embryo from the effects of the mutation? To examine the first question, we performed a series of reciprocal embryo transfer experiments, moving iv/iv blastocysts into + / + recipients, and vice versa. These studies also allowed us to evaluate whether embryo collection, Handed asymmetry is the consistent difference between left and right sides of the body. For example, the position of the stomach on the left side; differing numbers of lung lobes on the two sides; aortic loop to the left; and so on. Clearly, the embryo must have a method to distinguish its left from right, in order to create this body pattern. There is virtually nothing known about the mechanisms involved, but we have recently proposed a model, and a conceptual framework (Brown and Wolpert, 1990). The model proposes that there are three processes: conversion of a molecular asymmetry to the multicellular level, which biases a system for the random generation of asymmetry, and finally an interpretation step for individual organs. The effects of the mouse mutant gene, iv (situs inversus viscerum), were important to the genesis of our model, as they have been to the thinking of others (see Brown and Wolpert, 1990). The embryos of homozygous iv/iv mice develop as if they have lost their sense Key words: asymmetry, handedness, chimeras, situs inversus. 950 N. A. Brown, A. McCarthy and L. Wolpert handling, culture or transfer had any effect on the development of asymmetry. Chimeric mouse embryos have often been used to study the interaction of normal and developmentally mutant cells. Rescue of mutant cells by aggregation of normal and mutant embryos has been reported for (1) lethal t12 and jpmsd mutations (Mintz, 1964; Eicher and Hoppe, 1973), (2) mutations of eye development (LaVail and Mullen, 1976; Muggleton-Harris et al. 1987) and (3) some murine trisomies (see Cox et al. 1984). It is almost certain that iv is a loss-of-function mutation, and it is likely that the normal allele codes for some component of the left/right conversion system. If the result of the iv mutation is the absence of a functional molecule, then the presence of non-mutant cells may be able to restore normality. We hypothesize that the development of iv/iv plus wild-type chimeras will follow one of two possibilities: if the overall determination of left/right positional information is a cell-interactive process then interaction between cells, mutant and normal, may be able to rescue iv/iv cells. Alternatively, cells may behave autonomously with respect to left/right position. Then, in a chimeric organ, the overall sidedness may be a 'majority decision' - whichever genotype contributes more than half the cells will dominate. Thus, any organ primordium with more than 50 % iv/iv cells would be phenotypically mutant, and half of these would develop with situs inversus. We have made chimeras by aggregation of 8-cell iv/iv and + / + mouse embryos. These were allowed to develop in utero to day 10 of gestation, the earliest stage at which visceral situs can be determined. Embryonic tissues were examined for chimerism using an isozyme marker. We focused, initially, on the heart because there is evidence that the development of its asymmetry is an intrinsic property of the heart, itself (Stalsberg, 1970), and looping of the heart is also the first observable manifestation of handed asymmetry. Materials and methods Mice The Si/Col inbred strain of mouse is homozygous for the iv gene, has a pale chinchilla coat colour, and is homozygous for the Gpi-lsb allele, which codes for the fast-migrating form of glucose phosphate isomerase (GPI-1B). The MF1 ( + / + ) outbred strain is albino, and homozygous for the Gpi-lj1 allele for slow-migrating GPI-1A. The inbred strain C57BL/ 6JLac ( + / + ) is pigmented, and homozygous for Gpi-lsb. B6CBF1 ( + / + ) mice are Fi hybrids (C57BL/6JxCBA/Ca), also homozygous for Gpi-lsb). Males of the T145H strain are genetically sterile and were used to induce pseudopregnancy. Collection of embryos Female + / + (MF1 or B6CBF1) and iv/iv (SI/Col) mice (3 to 4 or 6 to 8 weeks old) were superovulated by intraperitoneal injection of 7.5i.u. or 2.5i.u. of pregnant mares' serum gonadotrophin, respectively, followed 48h later with 5 i.u. of human chorionic gonadotrophin. Mice were paired overnight with their respective males. The presence of a vaginal plug the following morning was taken as an indication of successful mating, and this was designated day 1. On day 3, the oviduct and anterior portion of the uterine horn were flushed with warm (about 37 °C) Medium 2 (M2, Fulton and Whittingham, 1978). Fragmenting, lysed, 1-, 2- or 4-cell stage embryos were discarded, and only 8-cell or early morula stage embryos were collected. Chimera production A well-established technique was used to produce chimeras (Tarkowski, 1961; Mintz, 1962). The zona pellucida was removed by brief exposure of embryos to warm acid Tyrode's solution (Nicholson et al. 1975). Zona-free embryos were washed three times in M2, followed by three washes in warm Medium 16 (M16, Whittingham, 1971) in small drops on a 60x15 mm plastic Petri dish under paraffin oil, which had been previously equilibrated with medium. One iv/iv and one + / + embryo were placed in a drop of M16 under paraffin oil, the two embryos were brought into contact with the aid of a glass pipette, and incubated at 37°C in an atmosphere of 5 % CO2 in air. Embryos were examined 2 to 3h later and, if needed, contact was once again established, prior to incubation overnight. The following day, fully aggregated chimeras, either at the late morulae or early blastocyst stage, were transferred into M2, and 4 to 6 were transferred into one horn of the uterus of a day 3 pseudopregnant B6CBF1 recipient female. To help establish pregnancy, the other horn was supplemented with C57BL/ 6JLac embryos, which had been collected at the 8-cell stage and cultured, as above, without removing the zona pellucida. Examination of chimeras The B6CBF1 recipients were killed on day 10, uteri removed and embryos were dissected out, keeping those from each horn (potential chimeras and C57BL/6JLac embryos) separate. The gross morphology of all embryos was examined and recorded, including the two major signs of situs visible at this time: (a) the side to which the heart tube had looped; and (b) the direction of embryo turning, which is manifest from two features: the side of exit/entry of the vitelline vessels; and the side of the head to which the tail-bud lies. In normal, situs solitus, embryos the heart tube loops to the right and the embryo turns to its right, resulting in the vitelline vessels lying to the left, and the tail bud to the right. The ectoplacental cone and amnion were removed and discarded and the embryo and yolk sac washed thoroughly in isotonic saline. The embryo was dissected into heart (the pericardium was discarded); head (up to and including the first branchial arch); body (up to the fore limb bud); and tail (remaining embryo). The separated tissues were transferred into 1.5 ml microcentrifuge tubes and stored at minus 20 °C. In the case of resorptions, any visible conceptual tissue, divided into the embryo and yolk sac if possible, was thoroughly washed and stored as above. Analysis of tissue genotype Tissue genotype was determined from GPI-1 type by electrophoretic analysis. First, a sample of heart tissue was analyzed. If this was found to be chimeric, then the remaining tissues, as well as the heart for a second time, were analyzed to assess the degree of chimerism throughout the embryo and yolk sac. GPI-1 analysis was carried out using a standard electrophoretic method. Tissues were homogenised by sonication in 30//1 50 mM Tris-HCl buffer pH8.0, using a Branson ultrasonic generator, duty cycle 30%, output 3, for 6s. 2 to 3 fA of this suspension was applied to cellulose acetate plates Development of handedness in chimeric embryos which had been soaked in 25 mM Tris/192mM glycine buffer pH8.5 for lOmin at room temperature. The running buffer was also 25 mM Tris/l92mM glycine pH8.5, and the isoenzymes were separated at 120 volts for 90min. In this system, the isoenzymes move towards the cathode with the BB form moving much further than the A A form. The isoenzymes were visualised by applying to the cellulose acetate plate a mixture containing: 2 ml 1.5% agarose; lml 50mM Tris-HCl buffer pH8.0; lml NADP (lmgmr 1 ); 3 drops MgCU (5.41g/l00ml); 3 drops fructose-6-phosphate (100mgmr f ); 3 drops 3-(4,5-dimethythiazol-2-yl)-2,5- diphenyltetrazolium bromide (lOmgml"1); 3 drops phenazine methosulphate (2.5mgml~1) and 9 units glucose-6-phosphate dehydrogenase. The colour was allowed to develop in the dark for about 5 min after which the reaction was stopped by soaking the plates in 5% acetic acid. The relative amounts of AA and BB isozymes were estimated from the intensity of staining of the two bands, measured by an LKB Ultrascan laser densitometer with integrator. These amounts were taken as a direct measure of the relative contributions of iv/iv and + / + genotype cells to the tissue analyzed. Preliminary experiments showed that equal amounts of iv/iv and + / + embryonic tissue gave the same intensity of staining on electrophoresis plates. In addition, electrophoretic analysis of mixed samples of iv/iv and + / + tissue showed that relative staining intensities were proportional to amounts of tissue (data not shown). Reciprocal transfers Si/Col (iv/iv); MF1 ( + / + ) and B6CBF1 ( + / + ) embryos were collected on day 3, cultured for 24 h, and transferred into day 3 pseudopregnant recipients, as described above, without removal of the zona pellucida. iv/iv embryos were transferred into B6CBF1 recipients, and + / + embryos into iv/iv (SI/Col) recipients. Recipient dams were killed on day 10, and the embryos examined, as described above. In some cases, recipients were killed on day 18 and fetuses were examined, as previously described (Brown et al. 1989). Statistical analyses The analyses of the proportions of normal and inverted embryos of different genotypes were by chi-square, with Fisher's exact or Yates correction where appropriate, using the SPSS/PC programme. The correlations of iv/iv contributions between different tissues was by Pearson product moment correlation coefficient and analysis of variance, using the Minitab/PC programme. Results Reciprocal embryo transfer +/+ embryos into iv/iv recipients A total of 147 + / + embryos (100 B6CBF1, 47 MF1), were transferred into 14 iv/iv recipients (5 inverted, 9 normal phenotype). Eight recipients (3 inverted, 5 normal), containing 81 embryos, did not become pregnant. Of the remaining 66 embryos, 40 (61%) implanted, and of these, 26 (65 %) were viable fetuses when examined on day 18. All 26 (100%) fetuses had normal visceral situs. iv/iv embryos into +/+ recipients 228 iv/iv embryos were transferred into 16 B6CBF1 + / + recipients. Eight recipients, containing 67 em- 951 bryos, did not become pregnant. Of the remaining 161 embryos, 90 (56%) implanted, and there were 46 (51 %) viable embryos upon examination on day 10. 22 (48 %) had normal situs, 20 (43 %) were situs inversus, and 4 (9%) were heterotaxic. All the heterotaxic embryos had inverted heart loops, but normal cephalocaudal turning. Chimeras The superovulation and mating of 472 iv/iv mice produced 201 (43 %) plugged animals, from which 1201 embryos were flushed. Of these, 658 (55 %) were at the 8- cell stage. 502 of these iv/iv embryos were aggregated with 8-cell or early morulae + / + embryos. After 24 h of culture, 384 (76%) of the aggregates had formed morulae, and were transferred into day 3 B6CBF1 pseudopregnant recipients. When recipients were killed on day 10 and their uteri examined, 135 implantation sites, containing 137 embryos or resorptions (36% of the number transferred), were collected and examined. Observations on these 137 conceptuses are summarised in Fig. 1. 55 (40 %) of the conceptuses were classified as resorptions as they did not contain a viable embryo. In 24 of these, some resorbing embryonic material was recovered, and 6 of these embryonic samples were chimeric, but in no case could embryonic situs be determined, so these are not informative. Of the 82 viable embryos, only 7 showed the situs inversus phenotype. Five of these appeared not to contain any + / + tissue, based on GPI-1 analysis, so are not chimeric, or contain only a very small proportion of normal cells. Both other situs inversus embryos contained mostly iv/iv tissue. One embryo was actually heterotaxic, with normal turning but inverted heart, and came from an implantation site containing a second, resorbing, embryo which was also chimeric (about 50:50 chimerism). A majority of the embryos with normal situs (47, 64%) appeared not to contain any iv/iv tissue, while some (9, 12%) were 100% iv/iv. A total of 19 werechimeric, 10 of which were mostly + / + tissue. The data in Fig. 1 show the poor developmental potential of iv/iv embryos. Of the 137 conceptuses, 54 were wholly + / + derived, but only half as many (25) were wholly of iv/iv origin. Similarly, only 6 of the 27 chimeric conceptuses were mostly iv/iv, whereas 14 were mostly + / + , and 7 equal. This suggests a selective advantage of + / + embryos, but this may be unrelated to the iv locus since the strains used differ at many loci. The genotype of the 31 completely resorbed conceptuses is unknown, but of the partially resorbed embryos 11 were iv/iv, only 7 + / + . This suggests a higher rate of embryonic death for iv/iv embryos. The degree of chimerism in different tissues of the 21 chimeric viable embryos is shown in Table 1. In general, the degree of chimerism was similar in the heart (H), head (D), body (B) and tail (T) portions of the embryos, and the percentage iv/iv contribution was highly correlated for all these tissues (for the 17 chimeras with complete data, the correlation coefficient 952 N. A. Brown, A. McCarthy and L. Wolpert Total Conceptuses 137 Resorptions 55 (40%) Viable Embryos 82 (60%) Situs Solitus 75 (91%) Situs Inversus 7 (9%) 100% iv/lv 5 (71%) Chimeras 2 (29%) 100% iv/iv 9 (12%) 100% • / • 47 (64%) Some conceptus 24 (44%) No conceptus 31 (56%) Chimeras 19 (26%) 100% • / • 7 (29%) 100% iv/iv 11 (46%) Mostly • / • 10 (53%) Mostly Iv/lv ' 2 (100%) Chimeric 6 (25%) _ Mostly • / • 4 (67%) Mostly iv/iv 3 (16%) Mostly iv/iv ' 1 (17) Equal 6 (32%) Equal 1 (17%) Fig. 1. Summary of the fate of iv/iv/'+/+ embryos, aggregated at the 8-cell/early morula stage, cultured overnight, transferred into day 3 pseudopregnant recipients, and examined on day 10. Table 1. Genotype of tissues in iv/iv + / + chimeric day 10 embryos Percentage contribution of iv/iv to tissue No. Heart Head Body Tail YSac Situs 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 79 71 68 68 66 55 54 54 53 49 43 39 28 28 28 25 23 21 15 15 2 57 51 53 46 62 42 37 33 54 63 50 58 82 67 47 40 45 39 52 50 54 75 60 50 40 43 44 46 73 71 6 55 61 52 50 54 41 52 34 20 37 43 36 21 60 45 38 31 24 16 16 19 25 20 18 13 24 18 18 7 43 67 30 70 36 49 58 21 38 N I H N N N N N N N N N N N N N N N N N N — = Tissue not analyzed. * Situs, N=Normal; I=Inverted; H=Heterotaxic. Table 2. Summary of tissue genotype and embryo phenotype in iv/iv plus + / + aggregated embryos examined on day 10 of gestation Tissue genotype iv/iv contribution to heart tissue All >50% <50% None Embryo phenotype Number of embryos Normal Inverted 9 7 12 47 Embryos aggregated at the 8-cell/early morula stage and examined on day 10. Genotype determined by GPI analysis. more than two thirds contribution of iv/iv tissue to both the heart and the visceral yolk sac. A summary of the tissue genotype and embryo phenotype of all the 82 viable embryos produced in the chimera study is given in Table 2. Discussion The maternal environment did not play a role in the development of handed asymmetry in these experiments. The results of the reciprocal embryo transfer r=0.91 B v T; 0.90 H v B; 0.86 H v D; 0.83 H v T; 0.83 experiments were clear-cut. Normal + / + embryos D v B; and 0.82 D v T). In contrast, the visceral yolk-sac transferred into iv/iv recipients developed with normal visceral situs, in all cases. Similarly, iv/iv embryos (Y) was often quite different, and was not correlated transferred into + / + recipients developed as they with the other tissues (r=0.23 Y v D; 0.17 Y v H; -0.10 would have done had they not been transferred: half Y v B; -0.03 Y v T). The mean contribution of iv/iv cells to the yolk sac in these 17 chimeras was 50%, with normal situs, half with situs inversus. This suggests that the maternal environment plays no role in the apparently greater than for any tissue of the embryo determination of the left/right axis, unless the influence proper (39 to 46%), but these are not statistically is irreversibly exerted during the first 2 days after significant differences (f>0.05, ANOVA). The two fertilization, prior to embryo collection in these embryos with abnormal situs were unique in having Development of handedness in chimeric embryos experiments. It is unlikely that left/right position is encoded within maternal cytoplasm since mutant (iv/iv) oocytes fertilized with + / + sperm develop completely normally (Layton, 1976; Brown, unpublished). Thus, left/right positional sense seems intrinsic to the embryo. These observations also show that the manipulations involved in collection, culture and transfer of iv/iv embryos did not have a detectable influence on the 50:50 ratio of normal:inverted visceral development. Thus, any effects on this ratio in the chimera studies can not be attributed to such artifacts. Three categories of viable embryos were produced in the chimera study, based on heart tissue genotype analysis: (a) all-+/+; (b) all-/v/iv; and (c) chimeras. The phenotype of the all—1-/ + embryos was, as expected, 100% normal visceral situs. Of the 14 alliv/iv embryos, 9 were normal, 5 were situs inversus. This does not differ, statistically (P>0.05, chi-square), from the expected 7:7 ratio, but the sample is small, in statistical terms. However, if the apparent shift towards a larger proportion of normal embryos were real, it could be due to rescue by the presence of a small amount of + / + tissue, either in the heart below the limit of detection of our analysis (about 1%), or in other tissues, which were not analyzed in those embryos that did not show apparent chimerism of the heart. The proposition of a rescue of iv/iv cells by normal cells in a chimera appears to be supported by our observations, although rescue is clearly not complete. There were 21 viable chimeric embryos in this study, but only 2 had abnormal situs. No embryo with less than two thirds iv/iv tissue (17 in total) showed abnormal situs, including 7 embryos with approximately equal (43 to 66 % iv/iv) proportions of normal and mutant tissue. This suggests that + / + cells exert some correcting influence over mutant tissue. Of course, these studies cannot suggest any mechanism for such an influence, although it is compatible with the possibility that + / + cells provide a molecule that the mutant cells lack. We cannot, however, reject the alternative possibility of cell-autonomy; indeed, our observations can be interpreted as generally consistent with this hypothesis. If cells were autonomous, we might expect that an embryonic organ with more than 50 % + / + cells would have the + / + (normal) phenotype, whereas one with more than 50% iv/iv cells would develop as a mutant, and thus half would be situs inversus. Of 12 chimeras with more than 50% + / + contribution to the heart, all had normal situs. Of the 9 chimeras with more than 50% iv/iv contribution to the heart, 2 had inverted heart loops. Although this 7:2 ratio appears to favour normal development, it is not statistically different (P>0.05, Fisher's exact) to the ratio of 4:5, or 5:4, that would be expected if the hypothesis were correct. Again, the numbers are small, and would have to be increased four-fold for the difference in the observed ratios to achieve statistical significance. It is possible that the asymmetry of an organ is determined by some critical sub-population of cells, which would be masked by our whole-organ analyses. Use of a cell-based 953 marker for chimerism would address this possibility (see below). While we cannot conclusively distinguish between the two hypotheses, it is clear that neither iv/iv nor + / + cells in an embryo are dominant. Only 2 of 21 chimeras developed with situs inversus, which is highly statistically different (P<0.01, chi-square) to the expected 10:11 ratio if iv/iv cells were dominant over +/+• Similarly, + / + cells are not dominant. The two embryos that had inverted situs contained as much as 50% + / + cells in some tissues, and about 30% in the heart. Thus, to answer our original question: the presence of a small proportion of non-mutant tissue is not sufficient to ensure normal development. The number of informative chimeras produced in this study was small, despite aggregating more than 500 iv/iv with + / + embryos. This is partially because of the poor developmental potential of Si/Col iv/iv embryos, as we have previously described (Brown et al. 1989). In addition, these iv/iv embryos seem to be at a disadvantage in aggregation chimeras with MF1 + / + embryos. It does not seem profitable to attempt to increase the numbers with further experiments of the same design. Rather, we intend to examine further chimeras using a cell-based marker, on histological preparations, rather than the whole-tissue based marker used here. In this way, we will gain information of the spatial distribution of chimerism within tissues, in relation to the development of asymmetry. One interesting observation can be made from our current data; the embryos with abnormal development of situs, in particular heart looping, were those with a high proportion of iv/iv cells in the heart, itself, rather than in the surrounding tissue. This supports the suggestion (Stalsberg, 1970) that the control of the direction of heart looping is intrinsic to the heart. In addition, it was only those embryos that had high iv/iv contribution to both heart and visceral yolk-sac that developed abnormally. It may not be coincidence that the early paired cardiac primordia develop in very close proximity to the visceral yolk-sac endoderm. Again, cell-based information on chimerism may allow some conclusions on the relative importance of endoderm and mesoderm components of the heart and visceral yolk sac. We thank the Biological Services Division of the National Institute for Medical Research for breeding the SI/Col mice. This work was supported, in part, by the Wellcome Trust, to whom we are grateful. References BROWN, N. A., HOYLE, C. I., MCCARTHY, A. AND WOLPERT, L. (1989). The development of asymmetry: the sidedness of druginduced limb abnormalities is reversed in situs inversus mice. Development 107, 637-642. BROWN, N. A. AND WOLPERT, L. (1990). The development of handedness in left/right asymmetry. Development 109, 1-9. Cox, D. R., SMITH, S. A., EPSTEIN, L. B. AND EPSTEIN, C. (1984). Mouse tnsomy 16 as an animal model of human trisomy 21 (Down's syndrome): formation of viable trisomy 16 diploid mouse embryos. Devi Biol. 101, 416-424. 954 N. A. Brown, A. McCarthy and L. Wolpert EICHER, E. M. AND HOPPE, P. C. (1973). Use of chimeras to transmit lethal genes in the mouse and to demonstrate allelism of the two X-linked genes jp and msd. J. exp. Zool. 183, 181-184. FULTON, B. P. AND WHITTINGHAM, D. G. (1978). Activation of mammalian oocytes by intracellular injection of calcium. Nature 273, 149-151. HUMMEL, K. P. AND CHAPMAN, D. B. (1959). Visceral inversion and associated anomalies in the mouse. J. Hered. 50, 9-13. LA VAIL, M. M. AND MULLEN, R. M. (1976). Role of the pigment epithelium in inherited retinal degeneration analyzed with experimental mouse chimeras. Expl Eye Res. 23, 227-245. LAYTON, W. M. (1976). Random determination of a developmental process: reversal of normal visceral asymmetry in the mouse. J. Hered. 67, 336-338. MINTZ, B. (1962). Formation of genotypically mosaic mouse embryos. Am. Zool. 2, 432. MINTZ, B. (1964). Formation of genetically mosaic mouse embryos, and early development of 'lethal' (t12/t12) normal mosaics. J. exp. Zool. 157, 273-292. MUGCLETON-HARRIS, A. L., HARDY, K. AND HIGBEE, N. (1987). Rescue of developmental lens abnormalities in chimaeras of noncataractous and congenital cataractous mice. Development 99, 473-480. NICHOLSON, G. L., YANAGIMACHI, R. AND YANAGIMACHI, H. (1975). Ultrastructural localisation of lectin binding sites on zonae pellucidae and plasma membranes of mammalian eggs. J. Cell Biol. 66, 263-274. STALSBERG, H. (1970). Mechanism of dextral looping of the embryonic heart. Am. J. Cardiol. 25, 265-271. TARKOWSKI, A. K. (1961). Mouse chimers developed from fused eggs. Nature 190, 857-860. WHITTINGHAM, D. G. (1971). Culture of mouse ova. J. Reprod. Fen. suppl. 14, 7-21. {Accepted 21 August 1990)