ABSTRACT
Cultured blastocysts homozygous for the t6 mutation lose their inner cell mass within a few days of attachment to the culture dish. At about the same time it becomes apparent that putative t6-mutant trophoblast cells and their nuclei fail to enlarge to the degree of their normal counterparts. These abnormalities in mutant embryos are reflected by an abrupt drop on the seventh equivalent gestation day in the rate of increase of β-glucuronidase activity. The failure of t6/t6 trophoblast nuclei to enlarge normally appears to be due partially to endoreduplication at a slower rate than normal trophoblast nuclei and partially to pre-mature cessation of DNA synthesis. Analyses indicate that this abnormality is not reversed when mutant embryos are placed in chimeric association with normal ones. Trophoblast outgrowths from mutant and normal trophectodermal vesicles are similarly distinguishable by differences in outgrowth and nuclear size as well as DNA content and synthesis. Despite the fact that t6/t6 embryos and trophectodermal vesicles are phenotypically different from normals from early times in culture, the trophoblast cells in the mutant structures acquire and continue to produce two enzymes characteristic of trophoblast differentiation, Δ 5,3 β -hydroxysteroid dehydrogenase and plasminogen activator.
On the basis of these and previous observations, we propose that the primary effect of the t6 mutation is to cause a metabolic lesion which kills inner cell mass cells relatively quickly but which has a more gradual effect upon trophoblast cells. The fact that phenotypically recognizable t6/t6 trophoblast cells can survive for several days before dying makes this a potentially useful system in which to search for the t6 gene product(s).
INTRODUCTION
Recessive lethal t mutations in the mouse fall into six ‘complementation’ groups which are characterized by the time and stage of embryonic development at which lethality is observed (see Sherman & Wudl, 1977). Previously we have presented evidence for the ‘generalized cell-lethal’ nature of at least three of the six types of lethal t mutations (Wudl & Sherman, 1976; Wudl, Sherman & Hillman, 1977). That is, all cells in the embryo are adversely affected by the mutation, even under conditions which minimize the need for embryonic organizational integrity. We also observed that homozygous t12-, t6-and tw5-mutant embryos behave in culture as they do in vivo, with disruption of develop-ment occurring abruptly and at a time and stage determined by the individual mutation. This, therefore, suggests that the lethal effects of t mutations are linked rather closely to the ‘developmental clock’. However, these observations do not allow us to decide whether t mutations disrupt the developmental pro-gramme directly or indirectly by their effects upon some basic metabolic process. Consequently, we are attempting to address this question by studies on early acting t mutations (causing death between the 4th and 8th days of gestation) aimed at closer scrutiny of the events occurring at or near the time at which abnormalities first become apparent.
Homozygous t6 embryos in vivo are grossly abnormal by the 7th day of gestation (Nadijcka & Hillman, 1975; see also Gluecksohn-Schoenheimer, 1940, for morphological studies on the related mutation, t°). Recently, we described the lethal effects of this mutation on embryos grown in an ectopic site and in vitro (Wudl et al. 1977; Sherman & Wudl, 1977). We found that in culture, putative t6/t6 blastocysts hatch and subsequently attach to the sub-stratum in a manner indistinguishable from control ( + / + or + / t6) blastocysts. Thereafter, however, the behaviour of presumptive t6/t6 embryos departs from that of their normal counterparts: soon after trophoblast cells begin to grow out along the culture dish, necrotic cells are apparent in the mutant inner cell masses (ICMs) and by the 9th EGD (equivalent gestation day, i.e. their age had they remained in utero), these ICMs have completely disintegrated (Wudl et al. 1977; see also Erickson & Pedersen, 1975). Furthermore, unlike tropho-blast cells from +/+ and + / t6 embryos, even exceptional ones that ‘spon-taneously’ lose their ICMs, trophoblast cells in presumptive mutant embryos fail to enlarge normally as the cultures develop; by the 11th EGD, some of the mutant trophoblast cells have died, and many of the remaining cells appear to be in the process of degeneration (Sherman & Wudl, 1977). That the presence of ICM cells in control embryos and their absence in mutant embryos is not responsible for the differential appearance of normal and mutant trophoblast cells could be shown conclusively by constructing trophectodermal vesicles (TVs), structures which lack ICM cells. Trophoblast cells in outgrowths from t6/t6 TVs are distinguishable from those of +/+ and + / t6 TVs by the same features observed for trophoblast cells outgrowing from blastocysts (Wudl et al. 1977).
In this report, we describe the results of more extensive studies of t6 embryos in vitro, investigations dealing with both the nature and time of expression of this mutation. We have analysed the levels in mutant embryos of an enzyme, β-glucuronidase, that rises dramatically during early development (Wudl & Chapman, 1976; Wudl & Sherman, 1976) in order to get some idea of when cellular metabolism is first interrupted; we have also looked for the expression by t6-mutant embryos and TVs of plasminogen activator (Strickland, Reich & Sherman, 1976; Sherman, Strickland & Reich, 1976) and Δ 5,3 β -hydroxysteroid dehydrogenase (3 β -HSD) (Chew & Sherman, 1975; Sherman & Atienza, 1977), enzyme markers that are characteristic of trophoblast differentiation. Finally, the small size of t6/t6 trophoblast nuclei suggested to us that they were failing to polyploidize normally (Sherman & Wudl, 1977; Wudl & Sherman, 1977); we have therefore compared the DNA content of, and incorporation of thymidine into DNA by, nuclei of phenotypically mutant v. normal trophoblast cells.
The t6 mutation lends itself to such studies because its critical period coincides with the time of marked differentiation of trophoblast cells and, while the mutation affects these cells adversely, it is not immediately lethal to them. This decreases the chance of observing alterations which are artifacts resulting from cell death. A further advantage of the t6 system is that mutant trophoblast cells which have never been in contact with other embryonic cell types can be identified and obtained. Therefore, we can largely eliminate any adverse effects of improper cell-cell interaction as being causitive of abnormalities that are observed. In short, we are attempting to reduce the complexity of the Z-mutant experimental system from one involving observations on dying embryos in vivo to one dealing with a single living cell type in vitro.
METHODS AND MATERIALS
Materials
[3H]thymidine, [3H]uridine and [3H]progesterone were purchased from New England Nuclear, Inc., Boston, Mass. NCTC-109 culture medium and fetal calf serum were purchased from Microbiological Associates, Bethesda, Md. Progesterone and pregnenolone were purchased from Steraloids, Inc., Pawling, N.Y., and recrystallized prior to use. Antiserum to progesterone (lot S257, no. 2) was purchased from Dr Guy Abraham, Harbor General Hospital, Torrance, Calif. 4-Methylumbelliferyl-β -D-glucuronic acid was purchased from Sigma Chemicals, St Louis, Mo. Basic fuchsin for the Feulgen reaction was purchased from Allied Chemical, Morristown, N.J., and adsorbed with activated charcoal to give a straw-yellow color prior to use. Fibrinogen, bovine, fraction I, was obtained from Calbiochem, La Jolla, Calif., and was purified by (NH4)2SO4 and ethanol precipitations as described previously (Beers, Strickland & Reich, 1975). Thrombin (topical) was purchased from Parke-Davis, Avon, Conn. Pregnant mare serum (Gestyl) and human chorionic gonadotrophin (Pregnyl) were purchased from Organon, South Orange, N.J. Coomassie brilliant blue R-250 was purchased from Inolex Corp., Glenwood, Ill. NTB-2 Nuclear Track Emulsion was obtained from Eastman Kodak, Rochester, N.Y. G. T. Gurr neutral mounting medium was purchased from BioMedical Specialties, Santa Monica, Calif. Culture vessels were from Falcon, Oxnard, Calif., except for 8-chamber, fixed gasket tissue culture slides which were obtained from Lab-Tek Products, Napierville, Ill.
Mice
T/ t6 tailless mice were obtained originally from Dr N. Hillman, Temple University, Philadelphia, Pa. Males were mated with SWR/J (wild-type for T-complex genes) females (Jackson Laboratory, Bar Harbor, Maine). The off-spring were checked for tail size and from these observations, it was observed that the male transmission frequency for the t6 allele was approximately 80% (Wudl et al. 1977). Normal-tailed offspring ( + / t6) were used in the experiments to be described. In the experimental cross, +/ t6 mice were mated inter se; the expected proportion of genotypes of the blastocysts obtained was + / + = 10%; + / t6 = 50%; t6/t6 = 40%. In the control cross, + / + (SWR/J) females were mated with + /t6 males. The expected proportion of genotypes was: + / + = 20%; + /t6 = 80%. Females were superovulated (Runner & Palm, 1953) prior to mating. The day of observation of the sperm plug is considered the first day of pregnancy.
Collection and culture of embryos
Late morulae and early blastocysts were collected on the fourth day of pregnancy and cultured individually in wells of Falcon Microtest 1 dishes con-taining NCTC-109 medium supplemented with antibiotics and 10% heat-inactivated fetal calf serum as described by Wudl & Sherman (1976). For Feulgen staining and autoradiography, embryos were cultured in groups of five in each compartment of an 8-chamber Lab-Tek slide. After the incubation period, the chambers were removed and the gaskets were scraped off with a scalpel blade.
Trophectodermal vesicles (TVs) were prepared according to the procedure of Tarkowski & Wroblewska (1967) with modifications described by Sherman (1975 a). Generally, embryos were removed at the two-cell stage, cultured for one day in preimplantation culture medium (Goldstein, Spindle & Pedersen, 1975) and then the blastomeres were disaggregated and cultured individually (from 4-cell embryos) or in pairs (from 8-cell embryos) in the same medium. After 2 days the resulting structures were inspected under an inverted phase optics microscope and TVs were collected, transferred to supplemented NCTC-109 medium and cultured as were blastocysts. Phenotyping of the resultant trophoblast outgrowths has revealed that TVs from t6/t6 embryos survive initially as well as do their +/+ or +/t6 counterparts: in three experiments a total of 130 experimental cross TVs were analyzed and 41% had a t6/t6 pheno-type, very close to the expected value of 40%.
Chimeric embryos were made at the 8-cell stage between pairs of embryos obtained from the experimental cross (Wudl & Sherman, 1976).
Embryonic development in culture is referred to in terms of equivalent gestation day (EGD), i.e. the age of the embryos had they been left in utero. EGD terminology is also used for TVs. It should be noted that TVs of a given EGD are developmentally retarded by approximately 1 full day when compared with blastocysts on the same EGD. This is because it takes about a day longer overall in comparison with blastocysts for disaggregated blastomere growths to pass through the initial developmental stages, i.e. cavitation, attachment to the culture dish, and outgrowth. Normal and mutant blastocysts could be distin-guished phenotypically on the 7th EGD, but TVs could only be phenotyped on the 8th EGD.
Nuclear and outgrowth measurements
Blastocyst or TV outgrowths were photographed (165 × magnification) with a Wild inverted phase-optics microscope equipped with a Polaroid camera attachment. Average nuclear diameters were measured from the photographs according to the formula , where dl and ds are the largest and smallest diameters, respectively. Outgrowth areas were measured on the photo-graphs with a planimeter.
DNA analyses
Blastocysts or TVs were fixed and stained by the Feulgen reaction as described by Barlow & Sherman (1972). Coverslips were affixed to the slides with neutral mounting medium. DNA contents were determined by microfluorometry with the aid of a Leitz MPV 1 microfluorometer. Excitation and emission wave-lengths were 560 and 627 nm, respectively. Machine units were converted into ploidy (C) values by measurements of liver nuclei, which contain 2, 4 and 8 times the haploid amount of DNA (Barlow & Sherman, 1972).
A utoradiography
Blastocyst or TV cultures were incubated in 0·5 ml volumes in chamber slides in the presence of [3H]thymidine (24 h) or [3H]uridine (4 h) at 1 μCi/ml and a final specific activity of 10 mCi/mmole. After the incubation period the cultures were washed, fixed, treated with trichloroacetic acid and autoradio-graphed as described previously (Sherman & Atienza, 1975). Slides were exposed for 7 days at 4°C, developed with Kodak Dektol developer for 2 min, and stained with Giemsa.
Enzyme analyses
β -glucuronidase assays were carried out by measuring the release of 4-methyl-umbelliferone from 4-methylumbelliferyl-β -D-glucuronic acid, as described by Wudl & Sherman (1976). Plasminogen activator assays were done by the fibrin-agar overlay procedure (Strickland et al. 1976), i.e. blastocysts or TVs cultured in 35 mm culture dishes were covered with a mixture of agar, medium, fibrinogen, thrombin and acid-treated fetal calf serum. The thrombin converts the fibrinogen to a fibrin clot. Plasminogen activator, secreted by the cells, will locally convert plasminogen in the serum to plasmin. The plasmin in turn breaks down the fibrin, in. the area. The result is a zone of lysis surrounding plasminogen activator positive cells; when the undegraded fibrin is stained with Coomassie brilliant blue, the lysis zones are detectable as clear halos around the cells. Δ 5,3 β -Hydroxysteroid dehydrogenase (3 β -HSD) activity was determined on single blastocysts cultured in wells of a Microtest dish in NCTC-109 medium supple-mented with serum absorbed with dextran-norit to remove endogenous steroids (Salomon & Sherman, 1975). On the 9th EGD the cultures were washed and each blastocyst was fed with 15 μl of this medium containing pregnenolone at a concentration of 1 μ g/ml. Twenty-four hours later 10 μl of culture medium were collected and monitored for the presence of progesterone by radioimmuno-assay (Marcal, Chew, Salomon & Sherman, 1975).
RESULTS
β -Glueuronidase activity in t6-mutant and normal embryos
Embryos derived from control ( + /+ × +/t6) and experimental ( + /t6 × + /t6) crosses were cultured singly and individual enzyme activities were measured from the 4th through the 9th EGD. The average enzyme activities for embryos from both the experimental and control crosses were determined for each day (Table 1). It is apparent that the overall average enzyme activities in both crosses rise markedly throughout most of the assay period. Since, due to transmission frequency distortion, approximately 40% of embryos in the experimental population are homozygous t6 mutants (Wudl et al. WIT), both the experimental and control populations were divided into subpopulations containing 60% of the embryos with the highest enzyme activity and the remaining 40% with the lowest activity. So long as enzyme activities of the total population represent a unimodal distribution about the mean activity, the ratio of average activities of the two subpopulations should remain constant. That this is the case for the embryos in the control cross is illustrated in Table 1 and Fig. 1. On each of the days studied, the average of the individual β -glucuronidase activities in the ‘high 60%’ subpopulation is approximately twice that found for the Tow 40%’ subpopulation (Table 1). Thus, the rates of increase in enzyme activity are the same for both groups (closed and open circles, Fig. 1).
β -glucuronidase activities in cultured blastocysts from control and experimental crosses. The values are taken from the data in Table 1. ●, Control cross, high 60%; ○, control cross, low 40%; ▴, experimental cross, high 60%; ▵, experimental cross, low 40%.
β -glucuronidase activities in cultured blastocysts from control and experimental crosses. The values are taken from the data in Table 1. ●, Control cross, high 60%; ○, control cross, low 40%; ▴, experimental cross, high 60%; ▵, experimental cross, low 40%.
The situation for embryos in the experimental cross is initially the same as that for the control cross, but only to the 6th EGD. Up to and including that time, the enzyme activities in the high 60% group are approximately twice those of the low 40% subpopulation (analyses of the distributions of individual enzyme activities in the total control and experimental crosses indicate that on the 6th EGD, the variances are not significantly different; P > 0·1). Thus, 4th to 6th EGD homozygous mutant embryos cannot be distinguished from heterozygous and wild-type counterparts on the basis of β -glucuronidase activity.
After the 6th EGD, the high 60%/low40% ratio in the experimentalpopulation increases continuously (Table 1). Fig. 1 illustrates that the high 60% sub-population, continues to show increased enzyme activities with a rate paralleling that of embryos in the control cross (the absolute enzyme activities in the experimental cross are greater because these embryos were somewhat advanced over those in the control cross at the beginning of the experiment). The rate curve for the low 40% population in the experimental cross, however, drops off between the 6th and 7th EGD and remains subnormal through the 9th EGD. Thus, the abrupt departure from normality of β -glucuronidase levels in putative t6-mutant embryos correlates with the time at which mutant ICM cells are visibly degenerating and mutant trophoblast cells can first be distinguished from normals by virtue of the smaller size of their nuclei.
Effects of the t6 mutation on trophoblast nuclear and outgrowth size and DNA content
For these studies, we have analyzed trophoblast cells derived either from blastocysts or from TVs. To demonstrate that all TVs from a single normal or mutant embryo would behave similarly when cultured further, and that we could dependably distinguish between mutant and normal outgrowths, we followed the development of TVs derived from the same embryo. Although disaggregated blastomeres can develop as miniblastocysts and non-integrated forms as well as TVs (Tarkowski & Wroblewska, 1967; Sherman, 1975 a), we observed a number of cases in which all four pairs of 8-cell blastomeres developed into TVs. Nuclear diameters were measured in 11th EGD outgrowths from two such sets, one wild-type, the other phenotypically mutant. The data in Fig. 2 demonstrate first that the nuclear diameters in the wild-type outgrowths are, on average, substantially larger than those in the presumptive mutants (30·2 v. 19·8 μm); second, the range of nuclear diameters from one TV to another within the same set is very similar. Thus, each TV behaves in a manner consistent with the genotype of the original embryo.
Diameters of trophoblast nuclei from pbenotypically t6-mutant and normal trophectodermal vesicles. The different symbols in A represent values from each of four phenotypically mutant TVs on EGD 11 derived from a single embryo. Those in B are from TVs from a control cross embryo of the same age.
Diameters of trophoblast nuclei from pbenotypically t6-mutant and normal trophectodermal vesicles. The different symbols in A represent values from each of four phenotypically mutant TVs on EGD 11 derived from a single embryo. Those in B are from TVs from a control cross embryo of the same age.
We have also measured total culture areas of individual TVs from control and experimental crosses on consecutive days from the 7th to 11th EGD. We found differences between the two populations of TVs beginning on the 8th EGD (Fig. 3, open v. filled circles). If we once again group the data from the experi-mental population into high 60% and low 40% subpopulations it is apparent that beyond the 7th EGD, the average culture areas for the high 60% resembles the total control population, whereas the low 40% population undergoes only a slight increase in average culture area over the period of study. This difference is not due to continuous cell death in presumptive mutant TVs: comparisons between nuclear numbers in control and experimental cross outgrowths show no substantial differences over the 5-day period of analysis (Table 2).
Outgrowth area measurements of trophectodermal vesicles derived from control and experimental cross embryos. After individually cultured TVs had begun to grow out along the culture dish, they were photographed for five successive days. The area of each outgrowth was measured with a planimeter and corrected for the magnification factor of the photograph. ○, Average area of 16 – 17 TV outgrowths from the control cross. ●, Average area of 40 – 45 TV outgrowths from the experimental cross. The experimental cross TV outgrowth areas have also been averaged for the high 60% (▵) and low 40% (▴) values beyond the 7th EGD. Separation into the two classes was not carried out on the 7th EGD because TVs are only beginning to grow out at this stage and no difference is observed in the average outgrowth area of experimental and control cross TVs.
Outgrowth area measurements of trophectodermal vesicles derived from control and experimental cross embryos. After individually cultured TVs had begun to grow out along the culture dish, they were photographed for five successive days. The area of each outgrowth was measured with a planimeter and corrected for the magnification factor of the photograph. ○, Average area of 16 – 17 TV outgrowths from the control cross. ●, Average area of 40 – 45 TV outgrowths from the experimental cross. The experimental cross TV outgrowth areas have also been averaged for the high 60% (▵) and low 40% (▴) values beyond the 7th EGD. Separation into the two classes was not carried out on the 7th EGD because TVs are only beginning to grow out at this stage and no difference is observed in the average outgrowth area of experimental and control cross TVs.
In order to determine whether the differences in nuclear size of trophoblast cells in normal v. mutant embryos reflect differences in their DNA content, we have measured DNA levels in trophoblast nuclei from blastocysts and TVs by means of a quantitative Feulgen stain coupled with microfluorometry. In an initial series of experiments, we cultured blastocysts from a ( + /t6 × + /t6) cross and measured DNA contents in trophoblast nuclei on the 11th EGD. Mutant blastocysts were recognized by their absence of ICM cells and by their small trophoblast nuclear and culture areas. For comparative purposes, we selected phenotypically wild-type blastocysts that had relatively small ICMs (but normal trophoblast nuclear and culture areas) in an attempt to minimize any effects of ICM cells. The results are recorded in Fig. 4 and Table 3. We found that all blastocysts contained trophoblast cells with polyploid nuclei. However, the distribution of DNA contents in trophoblast nuclei of phenotypically mutant and normal blastocysts is visibly different: in general, normal trophoblast nuclei undergo about 2·5 more endoreduplicative cycles than mutant ones as reflected by the average and largest DNA contents in the two groups. In other studies (not shown), the DNA contents of the five largest trophoblast nuclei from each of fifteen phenotypically normal and mutant bastocysts were measured. In this case, the largest DNA content in a normal nucleus (655 C) was about three endoreduplicative cycles beyond that found in a mutant nucleus (80 C).
DNA contents of trophoblast cells from phenotypically normal and t6-mutant blastocysts. Blastocysts from the experimental cross were cultured to the 11th EGD. They were then fixed and stained as described in Methods. Embryos were phenotyped using a low power microscope objective prior to the determination of DNA contents in trophoblast cells. A total of 150 presumptive normal and 150 presumptive t6-mutant nuclei was measured. Machine units were converted to C values, i.e. multiples of the haploid amount of DNA, by standardization with liver nuclei that were identically fixed and stained. (A) Trophoblast nuclear DNA contents from presumptive t6/t6 blastocysts; (B) trophoblast nuclear DNA contents from pre-sumptive + /t6 and + / + blastocysts.
DNA contents of trophoblast cells from phenotypically normal and t6-mutant blastocysts. Blastocysts from the experimental cross were cultured to the 11th EGD. They were then fixed and stained as described in Methods. Embryos were phenotyped using a low power microscope objective prior to the determination of DNA contents in trophoblast cells. A total of 150 presumptive normal and 150 presumptive t6-mutant nuclei was measured. Machine units were converted to C values, i.e. multiples of the haploid amount of DNA, by standardization with liver nuclei that were identically fixed and stained. (A) Trophoblast nuclear DNA contents from presumptive t6/t6 blastocysts; (B) trophoblast nuclear DNA contents from pre-sumptive + /t6 and + / + blastocysts.
We carried out similar studies with TVs from normal and mutant embryos on the 12th EGD (we measured these one EGD later than blastocysts since TV development is retarded by approximately a day relative to blastocysts). The results were basically the same: normal trophoblast nuclei polyploidize almost to the extent of their counterparts in blastocysts (Fig. 5, Table 3) and to degrees similar to those in TVs from the control ( + /+ × + /t6) cross (Table 3). Poly-ploidization also takes place in most mutant TV nuclei, but not to the extent of normals. In fact, more than one-third, of mutant trophoblast nuclei have DNA values at the level of 8 C or less, whereas almost all normal trophoblast nuclei, both from experimental and control crosses, have DNA contents greater than 8 C. Conversely, four-fifths of the latter nuclei, but none of the former, have DNA contents in the 32 C range or greater (Table 3).
DNA contents of trophoblast cells from phenotypically normal and t6-mutant trophectodermal vesicles. TVs from the experimental cross were cultured to the 12th EGD and treated as were blastocysts (see Methods and legend to Fig. 4). Sixty-two (A) and ninety-four (B) trophoblast nuclei were measured from pheno-typically mutant and normal TVs, respectively.
DNA contents of trophoblast cells from phenotypically normal and t6-mutant trophectodermal vesicles. TVs from the experimental cross were cultured to the 12th EGD and treated as were blastocysts (see Methods and legend to Fig. 4). Sixty-two (A) and ninety-four (B) trophoblast nuclei were measured from pheno-typically mutant and normal TVs, respectively.
We have examined more closely the relationship between nuclear size and DNA content by constructing regression lines for these parameters for the 150 nuclei each from normal and mutant blastocysts analyzed in Fig. 4. The regression lines for the normal and mutant nuclei, shown in Fig. 6, have similar, but not identical, slopes. The difference in slopes indicates that, on average, mutant nuclei of the same diameter as normal ones contain slightly less DNA, suggesting, perhaps, that mutant nuclei flatten out more extensively on the culture dish surface than do control nuclei.
Relationship between trophoblast nuclear diameters and DNA contents. At the same time as DNA contents in 150 normal and 150 t6-mutant trophoblast nuclei were determined (see Fig. 4), nuclear diameters were also measured. Regression lines between these parameters were constructed by computer analysis. – –, Nuclei from phenotypically normal blastocyst outgrowths; —, nuclei from phenotypically t6-mutant blastocyst outgrowths.
Relationship between trophoblast nuclear diameters and DNA contents. At the same time as DNA contents in 150 normal and 150 t6-mutant trophoblast nuclei were determined (see Fig. 4), nuclear diameters were also measured. Regression lines between these parameters were constructed by computer analysis. – –, Nuclei from phenotypically normal blastocyst outgrowths; —, nuclei from phenotypically t6-mutant blastocyst outgrowths.
The data presented so far do not indicate whether the differences in DNA content between normal and mutant nuclei are due to continuing endore-duplication in the latter, although at a slower rate than in the former, to pre-mature cessation of polyploidization in mutant nuclei after an initial period of endoreduplication at a normal rate, or to a combination of decreased endo-reduplication rates and premature cessation. We approached this question by measuring nuclear diameters in trophoblast cells from normal and mutant TVs on successive culture days and then estimating DNA contents from the regres-sion lines in Fig. 6. Each circle in Fig. 7 represents the average nuclear diameter of trophoblast cells in an individually cultured TV from the control or the experimental cross. The circles in the left-hand panels represent values for trophoblast nuclei in the control cross; there is a steady increase in the average nuclear diameters with age. Similar analyses on the experimental cross (right-hand panels) generally show a wider distribution of average nuclear diameters than those observed in the control cross.
Average trophoblast nuclear diameters of trophectodermal vesicles derived from control and experimental cross embryos. Diameters were determined for tropho-blast nuclei in each of the TVs described in the legend to Fig. 3 on the 7th through 11th EGD. Each circle represents the average diameter of all the nuclei in an indi-vidually cultured TV. ●, TVs from the control cross; ○, TVs from the experimental cross.
Average trophoblast nuclear diameters of trophectodermal vesicles derived from control and experimental cross embryos. Diameters were determined for tropho-blast nuclei in each of the TVs described in the legend to Fig. 3 on the 7th through 11th EGD. Each circle represents the average diameter of all the nuclei in an indi-vidually cultured TV. ●, TVs from the control cross; ○, TVs from the experimental cross.
The average values in Fig. 7 have been further averaged for each day and the results are plotted in Fig. 8. The average nuclear diameters increase in both experimental (filled circles) and control (open circles) populations, although at a somewhat lower rate in the former. When the experimental population is separated into the high 60% and low 40% subpopulations for the reasons described in previous sections, the result is an overlap between the high 60% subpopulation and the control cross population, indicating that nuclear dia-meters of trophoblast cells from presumptive + / + and + /t6 TVs in both crosses are similar. The low 40% subpopulation shows a slight, but continuous, increase in average nuclear diameters with time.
Relationship between average trophoblast nuclear diameter and equivalent gestation age for outgrowths of trophectodermal vesicles from the control and experimental crosses. Each circle on these curves represents an average of the averages illustrated in Fig. 7. ○, Control cross; ●, experimental cross. Also, the lowest 40% of the experimental cross values (▴) from Fig. 7 have been averaged separately from the highest 60% (▵).
Relationship between average trophoblast nuclear diameter and equivalent gestation age for outgrowths of trophectodermal vesicles from the control and experimental crosses. Each circle on these curves represents an average of the averages illustrated in Fig. 7. ○, Control cross; ●, experimental cross. Also, the lowest 40% of the experimental cross values (▴) from Fig. 7 have been averaged separately from the highest 60% (▵).
In Table 4 we have estimated DNA contents from nuclear diameters using the appropriate regression lines in Fig. 6. As expected, the average estimated DNA contents for normal nuclei in the experimental and control crosses rise in parallel with time. The presumptive mutant nuclei show only a doubling in the average estimated DNA contents over the 5-day period, but the rise is continuous.
Nucleic acid synthesis by t6 Nmutant and normal trophoblast cells
In order to obtain a clearer idea of the ability of t6-mutant embryos to endoreduplicate during the culture period, we incubated blastocysts or TVs in [3H]thymidine for 24 h periods between the 6th and 11th EGD. We found that from the earliest times investigated, i.e. shortly after the onset of outgrowth of trophoblast cells, phenotypically mutant blastocysts or TVs incorporate the isotope, but fewer of the trophoblast nuclei are labeled when compared to those in normal outgrowths in the experimental or control cross (Table 5). As the culture period proceeds, the proportion of labeled trophoblast nuclei in both mutant and normal outgrowth falls, but the decline is more precipitous in the mutant populations. Even so, between the 10th and 11th EGD in the latter case, a few trophoblast nuclei can be found to incorporate [3H]thymidine. On the other hand, the percentage of labeled trophoblast nuclei in normal blastocysts levels off or even rises at this stage in culture as relatively small presumptive secondary trophoblast cells (Sherman, 1975 b) can be seen to migrate out along the culture dish from under some of the ICMs. These cells, which incorporate label into their nuclei, are not seen in mutant blastocysts since the latter do not have ICMs. The coincidental rise in labeled nuclei in phenotypically normal TVs on EGD 10·5 is not due to the presence of secondary trophoblast cells, but is probably an artifact related to the small number of nuclei measured at that stage.
Incorporation of [3H] thymidine by normal and t6-mutant blastocyst and trophectodermal vesicle outgrowths
![Incorporation of [3H] thymidine by normal and t6-mutant blastocyst and trophectodermal vesicle outgrowths](https://cob.silverchair-cdn.com/cob/content_public/journal/dev/48/1/10.1242_dev.48.1.127/4/m_develop_48_1_127tb5.png?Expires=1739971001&Signature=xfChBRzIr6aiOR3USuDOShpQvWfnW-lBaaneuaOsHcvrhtmiNg~1Cw9-xm4qtSB4dVCl~YV-C5Cu64~WYOIQBxzo673CIGSl-3d4QJezmGL~MXTSbkPI-ylJeO7o5nwzEljC5FEd-vyUCP6M8JIM0JXI8f5-bz0jFUzm5gpEpzI7P2wcCFYDrSh4RcvmiZkE6wRpD~WDt7~ACRtUZcL45TRRHJtW9wDASleDU6LXjQSY2ArwjeBqHVi0BYrwIXWAN~e0NnU-nI6RN6vnS4~BT4Q2i9PXi8i6RR1q5C6DpQ6Hm47tb9zIbU54gOXKCkr6iE16MjNHVtZXvKzb6ERuUg__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
Since trophoblast nuclei have disparate sizes, shapes and distributions of chromatin, we felt that grain counts would not provide a reliable indication of labeling intensity. However, by visual inspection, it appeared as though the grain density was consistently lower over mutant than over normal nuclei, most obviously for the EGD 9·5 samples.
A further series of experiments was carried out to determine whether t6-mutant trophoblast cells continued to incorporate [3H]uridine into RNA throughout the culture period. Even when the labeled nucleoside is administered to blastocysts for as little as 4 h, virtually all ( > 90%) trophoblast cells, whether phenotypically normal or mutant, are labeled when studied by autoradiography. Although grain densities appear by visual inspection to be lower over mutant than over phenotypically normal trophoblast cells on EGD 8, gross differences are not apparent either at earlier stages or beyond EGD 8, at which time the grain density seems to diminish over both mutant and normal outgrowths.
t6-mutant: normal chimeric embryos
We generated chimeric blastocysts by fusing pairs of 8-cell experimental cross embryos in order to investigate whether (1) t6/t6 ICMs die in culture because they have subnormal cell numbers (Sherman (1975a) has shown this to be the case for miniblastocysts developing from disaggregrated blastomeres) and (2) mutant trophoblast cells might polyploidize normally if in direct cell-cell contact with normally replicating trophoblast cells.
When we inspected outgrowths developing from chimeric blastocysts on the 10th EGD, we found three types that could be distinguished by morphology (Fig. 9): the first type, from presumptive normal:normal chimeras, looks like a large, phenotypically wild-type embryo with a substantial ICM and many giant trophoblast cells; the second type of outgrowth, presumably from mutant: mutant embryos, contains only undersized, although healthy-looking, tropho-blast cells, and little or no evidence of surviving ICM cells; finally, the putative normal:mutant chimeras give rise to structures containing some ICM cells and two discrete clusters of trophoblast cells, one of normal size, the other distinctly undersized. The results indicate that t6/t6 ICMs are not rescued in vitro by being combined in pairs, and that t6/t6 trophoblast cells do not acquire a normal phenotype by coming into contact with normal trophoblast cells.
Outgrowths of experimental cross chimeric blastocysts. Structures were photographed on the 10th EGD. (A), Presumptive normal: normal chimera; (B) presumptive t6-mutant: t6-mutant chimera; (C) presumptive normal: t6-mutant chimera. Scale marker (in C) = 50 μm.
Differentiation of t6-mutant trophoblast cells
Table 6 indicates that plasminogen activator activity can be detected in blastocysts by the fibrin-agar overlay assay between the 6th and 7th EGD and the enzyme continues to be secreted by all embryos, mutant and normal, up to the 10th EGD. Since beyond that point plasminogen activator production is more marked in parietal endoderm than trophoblast cells (Strickland et al. 1976) and only normal blastocysts would possess the former cell types, we carried out similar studies on TVs, wherein both normal and mutant outgrowths would contain only trophoblast. We found that similar numbers of outgrowths from both the mutant and normal crosses produce plasminogen activator. As reported earlier (Sherman et al. 1976), the amount of enzyme produced and secreted by TVs drops beyond EGD 10, so that the size of the fibrinolysis zones, and, eventually, the number of outgrowths secreting detectable amounts of enzyme, decreases. There is a greater number of negatives in the EGD 12·5 sample in the experimental cross than in the control cross because, as we have mentioned previously (Wudl et al. 1977), cell death is apparent by this time in t6/t6 TVs, and the remaining viable trophoblast cells in these structures pre-sumably cannot produce sufficient amounts of plasminogen activator to be detected by our assay.
Plasminogen activator production by normal and t6-mutant cultured blastocysts and trophectodermal vesicles

We assayed blastocysts from control and experimental crosses for 3β-HSD activity by culturing them individually in the presence of pregnenolone (on the 9th EGD). The culture medium was collected and assayed for progesterone production 24 h later. All of the outgrowths produce and secrete progesterone (Fig. 10). However, the distribution of the amounts of progesterone detected is very different in the two crosses, with the average progesterone content being substantially higher in the experimental than in the control cross. This is mainly because the media from phenotypically mutant blastocysts contain on average much more progesterone than the media from normals (compare open and filled circles in Fig. 10 A). This is explained by the fact that the mutant blasto-cysts have lost their ICMs by the time of the analysis: we have observed pre-viously (Sherman & Atienza, 1975, 1977; Sherman, Atienza, Salomon & Wudl, 1977) that if ICMs are present, a substantial proportion of the progesterone formed may be metabolized and so will not be detected by the selective radio-immunoassay procedure that is used.
Progesterone production by individual blastocysts from the control and experimental crosses. 3β-HSD activities were estimated by the ability of 9th EGD blastocysts to convert pregnenolone in the culture medium to progesterone (see Methods). Prior to the collection of the culture medium for assay, blastocyst out-growths in the experimental cross (A) were phenotyped: ○, presumptive +/ t6 or + /+; ●, presumptive t6/t6. (B) blastocyst outgrowths from the control cross. Each circle represents the amount of progesterone produced by a single blastocyst outgrowth over a 24 h period.
Progesterone production by individual blastocysts from the control and experimental crosses. 3β-HSD activities were estimated by the ability of 9th EGD blastocysts to convert pregnenolone in the culture medium to progesterone (see Methods). Prior to the collection of the culture medium for assay, blastocyst out-growths in the experimental cross (A) were phenotyped: ○, presumptive +/ t6 or + /+; ●, presumptive t6/t6. (B) blastocyst outgrowths from the control cross. Each circle represents the amount of progesterone produced by a single blastocyst outgrowth over a 24 h period.
DISCUSSION
Whereas the ICMs of putative t6/t6 blastocysts die at an early stage of culture, trophoblast cells remain viable for several days thereafter, although they are obviously affected by the mutation in that they fail to endoreduplicate normally. It has been proposed that t mutations interfere with cell-cell organi-zation or interactions (see Bennett, 1975). However, t6-mutant trophoblast cells grow out in culture to give monolayers with a regular appearance; when com-pared to normals, mutant cells show no increased tendency either to migrate away from, or pile up on, each other (Sherman & Wudl, 1977; Wudl et al. 1977). In their gross morphology, presumptive t6/t6 trophoblast cells appear different from normals only in that their nuclei do not become very large (see, for example, Fig. 9). Although statistical analyses indicate that the nuclei of mutant tropho-blast cells contain slightly less DNA per unit area than those of normals (Fig. 6), this difference is not reflected by a visible alteration in nuclear morphology and its significance is unclear.
Taken together, the data we have presented indicate that the difference in nuclear size between presumptive t6/t6 and normal trophoblast cells is related primarily to differing DNA contents in the nuclei. Although other explanations of the data in Tables 4 and 5 and Fig. 5 might not be formally ruled out, we believe the most likely interpretation of these results is that polyploidization takes place continuously in t6-mutant trophoblast nuclei but at a consistently slower rate than in +/+ or +/ t6 trophoblast nuclei. Some presumptive t6/t6 trophoblast cells cease endoreduplication 1 or 2 days prior to normals, but others continue to synthesize DNA for the same length of time as normals. Autoradio-graphic experiments designed to detect the incorporation of uridine into RNA indicate that those trophoblast cells which have ceased replication, both mutant and normal, nevertheless continue to synthesize RNA.
Since trophoblast cells in genotypically normal TVs polyploidize at almost the same rate as those in blastocysts, it is unlikely that t6-mutant trophoblast cells fail to polyploidize normally due to lack of interaction with ICM cells. The likelihood is further reduced by our observation that putative mutant tropho-blast cells are retarded in their rate of endoreduplication even as part of a chimera containing phenotypically normal trophoblast and ICM cells.
Despite its adverse effects upon polyploidization, the t6 mutation does not prevent trophoblast cells from producing 3β-HSD and plasminogen activator, two specialized enzymes that are characteristic of trophoblast differentiation (Chew & Sherman, 1976; Strickland et al. 1976). From previous observations, we can state that these enzymes are first detectable in normal embryos at about the time at which we can discern clear differences between normal and mutant embryos (loss of ICM cells and subnormal levels of β-glucuronidase activity). Because we have not determined the time of transcription of the 3β-HSD and plasminogen activator genes, we do not know whether they were activated in t6-mutant cells prior to, during, or subsequent to interference with normal development. We can, however, argue that synthesis of these differentiated gene products is continued at substantial levels after the cells are otherwise affected by the mutation. It is unlikely that t6-mutant trophoblast cells are merely secreting plasminogen activator formed while they were still phenotypically normal because Strickland et al. (1976) found that the amount of enzyme present in blastocyst cells at any time is only a small fraction of the amount being secreted. Furthermore, the levels of 3β-HSD activity in cultured pre-sumptive t6/t6 embryos as reflected by progesterone production (1–2 pmoles/day) compares favorably with those observed for wild-type embryos in which the ICM had been eliminated by antimetabolite treatment (Sherman & Atienza, 1975). We conclude, therefore, that the t6 mutation does not interfere directly with known differentiative processes in trophoblast cells.
Since our observations render it unlikely that the primary action of the t6 mutation on trophoblast cells is interference with cell interactions or differen-tiation, we favor the view that the initial lesion is a metabolic one, that is, that the +t6 gene product(s) is directly involved in basic cellular metabolism (see Hillman, 1975; Sherman & Wudl, 1977). Such a lesion could explain why ICM cells, which might require intense metabolic activity for their rapid rates of cell division, would succumb to the effects of the t6 mutation abruptly whereas non-dividing trophoblast cells might be able to persist for a longer period of time under suboptimal conditions. In fact, this is what happens when blastocysts are cultured in the presence of appropriate doses of various antimetabolites (Rowinski, Solter & Koprowski, 1975; Sherman & Atienza, 1975; Pedersen & Spindle, 1976). Because replication is an energy-requiring process, it is not difficult to visualize a drop in the rate of polyploidization of t6-mutant trophoblast cells under conditions of metabolic insufficiency. We have found recently that certain suboptimal culture conditions prevent expression of some trophoblast differen-tiation markers but not of others (M. H. Sellens, L. R. Wudl & M. I. Sherman, unpublished observations); this could be analogous to the pronounced degree of interference by the t6 mutation with trophoblast polyploidization relative to 3β-HSD and plasminogen activator production.
Nadijcka & Hillman (1975) have reported that presumptive t6/t6 embryos can be distinguished from normal embryos by ultrastructural analyses at sub-stantially earlier stages than those in which we have been able to discriminate between the two on the basis of β-glucuronidase activities and death of ICM cells. This would imply that t6-mutant gene expression occurs at early stages of development with relatively minor consequences, yet its effect becomes lethal, and abruptly so in so far as the ICM is concerned, at a later period in embryo-genesis. Such a phenomenon might be understood by taking into account the observations that cleavage-stage embryos can develop to the expanded blasto-cyst stage in a medium containing only salts, glucose, pyruvate and albumin (e.g. Biggers, Whitten & Whittingham, 1971), yet, for further development, a complex mixture including amino acids, vitamins and serum factors is necessary (e. g. Gwatkin, 1966). Thus, it appears as though the embryo is progressing through a period of increased ‘metabolic sophistication’ during the pheno-critical period for the t6 mutation, and a gene product which played only a limited role just prior to this period could become crucial for normal cellular metabolism within a very short time.
We have found the comparative study of normal and t6-mutant TVs to be relatively simple because only a single cell type is involved, and advantageous because it is possible to unambiguously phenotype homozygous t6-mutant cells several days before their death. This advantage does not apply to later-acting t mutations. For example, normal and mutant TV outgrowths from a ( + /tw5 × + /tw5) cross are morphologically indistinguishable from each other as late as the 11th EGD (unpublished observations). Therefore, the t6 TV system might prove to be particularly useful in our attempts to understand the mode of action of this mutation. This in turn might help to clarify some of the mystery surrounding T-complex mutations in general.
ACKNOWLEDGEMENTS
We are grateful to Ms Jill Lunn for statistical analyses and to Dr A. Weissbach for comments on the manuscript.