J. Embryol. exp. Morph. 74, 159-168 (1983)
Printed in Great Britain © The Company of Biologists Limited 1983
159
Rabbit a-globin messenger RNA translation by the
mouse ovum
By KARL M. EBERT 1 AND R. L. BRINSTER 1
From the Laboratory of Reproductive Physiology, University of Pennsylvania
SUMMARY
Fertilized mouse ova were injected with messenger RNA for rabbit globin. The ova were
labelled with [3H]leucine and the synthesized rabbit a-globin measured by immunoprecipitation. Injection of rabbit globin mRNA from 4-0 to 15-8pg/ovum resulted in the synthesis of
~1200 dpm/ovum of a-globin during an 18 h incubation. A concentration of mRNA that was
translated below the maximum level of globin synthesis was used to determine translation
efficiency. The efficiency of translation was 14-1 molecules of a-globin synthesized/cell/h for
each molecule of a-globin mRNA injected. The radioactivity in a-globin was 1 % of the total
incorporation in acid-precipitable material. Total incorporation of control and injected ova
was not significantly different. This study shows that the fertilized mouse ovum can translate
an injected foreign message with essentially the same efficiency as that reported for the
Xenopus oocyte but substantially lower than the reticulocyte and that the translational capacity of the mouse fertilized ovum for injected mRNA is limited with little if any spare translational ability.
INTRODUCTION
The fertilized mouse ovum can translate mRNA that is microinjected into the
cytoplasm (Brinster, Chen, Trumbauer & Avarbock, 1980; Brinster, Chen,
Trumbauer & Paynton, 1981). The ability of the ovum to translate exogenous
mRNA along with the development of techniques to identify and quantify
message products allow for the direct study of mRNA translation during early
development. Presently there is little information on how the mammalian ovum
uses mRNA during the early developmental stages. The absence of demonstrable transcriptional activity of the 1-cell mouse embryo suggests the lack of
continuous mRNA production (Knowland & Graham, 1972; Levey, Stull &
Brinster, 1978; Young, Sweeney & Bedford, 1978). The problem of how the
germ cell maintains protein synthesis without detectable transcriptional activity
has been frequently explained by theories on stored and inactivated maternal
mRNA (Van Blerkom & McGaughey, 1978) and/or the lack of significant turnover of mRNAs previously produced during the maturation of the oocyte (Johnson, Handyside & Braude, 1977). The abundance of polypeptides and mRN As
in the fertilized ovum makes it difficult to study translational activity of an
1
Authors' address: Laboratory of Reproductive Physiology, School of Veterinary
Medicine, University of Pennsylvania, Philadelphia, Pennsylvania 19104, U.S.A.
160
K. M. EBERT AND R. L. BRINSTER
endogenous message. This study was designed to quantitate the translational
capacity and efficiency of the fertilized mouse ovum following injection of rabbit
globin mRNA into the cytoplasm.
MATERIALS AND METHODS
Ova collection
Fertilized 1-cell ova were collected from C57 x SJL and C57 x DBA hybrid
mice following superovulation and mating to C57 x SJL males according to
Brinster (1972). The average time interval between pregnant mares serum
gonadotropin (Gestyl, Organon) and human chorionic gonadotropin (HCG,
Sigma CG-2) injections was 48-0h.
Fertilized ova were collected in BMOC-2 medium (Brinster, 1972) from
animals that had a vaginal plug 21 h following HCG injection. BMOC-2 medium
was modified to contain 5-2g/1 of NaCl and 5mg/ml bovine serum albumin.
Cumulus cells were removed by treatment with hyaluronidase (Sigma, Type IV)
(300 units/ml in PBS containing 1 % PVP 40T, Sigma) diluted to 60 units/ml
with culture medium. Ova were washed twice in 2 ml of culture medium to
remove debris.
Injection
Ova were injected as previously described (Brinster et al. 1980). Prior to and
during the injection procedure ova were exposed to cytochalasin B (Aldrich) at
a concentration of 5 jUg/ml diluted in culture medium with 25 mM HEPES salts
(pH7-4). An average volume of 7-9pl/ovum was injected as determined by
[3H]uridine injection. During a 3h interval, 46% of injected ova survived
resulting in 40-50 injected ova per experiment. Ova were labelled beginning 1 h
after injections were completed.
Globin mRNA
The purified rabbit globin mRNA was a gift from Dr J. Lingrel. It was
prepared by the method of Gorski, Morrison, Merkel & Lingrel (1974) except
that polysomes were dissociated with SDS and applied directly to oligo (dT) cellulose as described by Krystosek, Cawthon & Kabat (1975). The content of
a- and /3-globin message was equal (Lingrel, personal communication). The
mRNA was dialysed and then lyophilized in 2 0 ^ aliquots. Each aliquot was
reconstituted in distilled water (previously boiled for 20min) as needed and
stored in glass tubes at - 7 0 °C.
Labelling
Ova were labelled for 18 h in a 25 ^1 droplet of culture medium containing
fA of [3H]leucine (TRK-510; Amersham) with a specific activity of
Translational efficiency of ova
161
130 Ci/mmol. Under these labelling conditions 28-2 % and 20-4 % of the control
and injected ova cleaved to the 2-cell stage compared to 93-7 % cleavage of 1-cell
ova cultured in the absence of the radioactivity. After labelling, the ova were
washed twice in 2 ml of culture medium, transferred to a 1-5 ml Eppendorf tube
and frozen at - 7 0 °C.
Immunoprecipitation
Because of the variation in labelling between experimental groups of ova each
mRNA-injected group had a corresponding control group (non-injected) that
was treated identically in all experimental phases except injection. Samples were
solubilized in 100^1 of 20 % normal rabbit serum in 0-25 % Triton X-100 and
frozen and thawed five times. They were ultracentrifuged in a Sorvall OTD-2
ultracentrifuge (SW-50 rotor) at 25000rpm (75 000 g) for 30min at 4°C. The
supernatant was transferred to a 1-5 ml Eppendorf tube and a 5jul aliquot
removed for determination of total acid-insoluble (TCA) protein incorporation.
The procedure for immunoprecipitation of a-globin was essentially the
method outlined by Boyer, Smith, Noyes & Mullen (1974). Antibody and
labelled globin used in the experiments were a gift from Dr S. Boyer. Anti-rabbit
globin was purified, characterized and shown to bind at least 88% of a 20ng
(1-3 pmole) quantity of a-globin with 1 /ig of antibody (6-5 pmoles). Cross reactivity with [3H]/3 globin was only 4 % in the presence of an equivalent immunoprecipitate (Boyer, personal communication). Five microlitres (5jUg) of
anti-rabbit a-globin was added to each supernatant, mixed and refrigerated
(4°C) for 1 h. After the initial reaction, lOjul (10 jig) of goat IgG carrier (Cappel)
plus 40 (A (160 jug) of rabbit anti-goat IgG (Cappel: F(abx) 2 Fragment, H*L)
were added and the secondary immunoprecipitation reaction performed at 37 °C
for 1 h. The reaction mixture was layered on a 1 ml sucrose pad containing 1 Msucrose, lOOmM-leucine, 1 % TX-100 and 1 % sodium deoxycholate that had
been previously centrifuged at 15 000 g for 30min (Rhoads, McKnight &
Schimke, 1973). The sample was centrifuged at 15 000 g for 15 min at 4°C. The
supernatant and sucrose pad were removed by aspiration. The pellet was washed
five times with 0-005 M-borate-saline buffer (pH8-4) containing 0-5 % TX-100
and 100 mM leucine. The final pellet was solubilized with 1-0 ml of 90% NCS
(New England Nuclear) and 10 ml of LSC complete (Yorktown Research). Samples were counted in a liquid scintillation counter (Intertechnique SL-30) and
dpm determined. Counts used as a measure of incorporation into a-globin were
based on the difference between the injected group and its corresponding control
immunoprecipitate.
Electrophoresis
Immunoprecipitates were solubilized in 20 \A of SDS-sample buffer and
electrophoresed on a 9-15 % linear gradient SDS-polyacrylamide gel according
toLaemmli(1970).
162
K. M. EBERT AND R. L. BRINSTER
RESULTS
Immunoprecipitates that were solubilized and electrophoresed on SDSpolyacrylamide gels showed rabbit a-globin in fertilized mouse ova that were
injected with rabbit globin mRNA (lane C) but not in corresponding controls
Mr
B
92,500
6aooo
46,000
mmm
30,000
mmm
12,000
Fig. 1. A 9-15 linear gradient SDS-polyacrylamide gel of immunoprecipitated
protein. (A) [3H]rabbit a-globin. (B) Control fertilized ova. (C) Rabbit globin
mRNA injected ova. (D) 1 4 C- standards. Immunoprecipitates were solubilized in
20 [A of SDS-sample buffer and placed in boiling water for 5 min. The a-globin in the
injected ova co-migrated with 3H-labelled rabbit a-globin. Solubilization of the
synthesized a-globin in the immunoprecipitates from injected ova resulted in the
formation of a smaller component (x) that was not present in control immunoprecipitates. Mr = relative molecular mass.
Translational efficiency of ova
163
3
(lane B). The a-globin comigrated with H-labelled rabbit a-globin (lane A).
Solubilizing the a-globin in the immunoprecipitate by SDS sample buffer resulted in the formation of a smaller component that migrated ahead of the authentic
a-globin and was not present in controls.
Recovery of extremely small amounts of rabbit globin present in the eggs
(1 x 108 molecules/ovum) was assumed to be 100 % under the conditions of
immunoprecipitation. Depending upon the amount of rabbit a-globin present in
the fertilized ova, approximately 25 % (range from 18 to 40 %) of the radioactivity in the immunoprecipitate of injected ova was labelled a-globin. However, by the use of simultaneously treated ova as controls the difference in
immunoprecipitate background for injected and control eggs was minimized.
Three to five experiments were performed at each of four message concentrations; 2-0, 1-0, 0-5, and 0-lpg/pl, which corresponded to 15-8, 7-9, 4-0,
and 0-8pg of message injected per ovum. Over an 18h labelling period the
average incorporation of pH]leucine into a-globin in the above four groups
was 78, 63, 58, and 34dpm/ovum/h respectively (Table 1). The amount of
a-globin synthesized at each concentration of mRNA was not significantly
different from the 2-0 pg/pl amount with the exception of the 0-1 pg/pl
concentration (p < 0-05; t-test). As shown in Fig. 2 the 0-1 pg/pl concentration
of mRNA did not fall on the linear regression line formed by the higher
concentrations of mRNA and represents a quantity of mRNA that was below
the apparent maximum translation of exogenous rabbit globin mRNA in the
fertilized mouse ovum.
The radioactivity in a-globin was approximately 1 % of the total incorporation
140 H
120
100
a.
3 80
£ 60
.S 40
O 20
0-1
0-5
10
Globin mRNA (pg/pl)
2-0
Fig. 2. Effect of globin messenger RNA concentration on a-globin synthesis by
mouse ova. A linear regression analysis between 0-5 and 2-0 pg/pl has a regression
equation of Y = 50-30-13-78 X and a correlation coefficient of 0-996. The 0-1 pg/pl
samples fell below the regression line.
164
K. M. EBERT AND R. L. BRINSTER
Table 1. a-Globin synthesis in fertilized one-cell mouse ova following microinjection of globin mRNA*
mRNA
pg/pl
2-0
1-0
0-5
0-1
a-Globin synthesis
dpm/ovum
Total pg
(mean ±
15-8
7-9
4-0
0-8
dpm/ovum/h
S.E.M.)
1403 ± 241 (5)t
1146 ± 457 (3)
1038 ± 244 (3)
616 ± 153 (4)
78 ±13
63 ±25
58 ±13
34 ±9t
* Ova incubated for 18 h in the presence of [3H]leucine
t ( ) = No. of experiments.
$ Significantly different from the 2-0 pg/pl group (P<0-05); student t-test.
Table 2. Efficiency of a-globin translation in the fertilized mouse ovum*
Experiment
1
2
3
4
Incorporation
(dpm/ova/h)
Molecules of a-globin/
cell/ht
6
29
53
13
41
Efficiency =
3-34 xlO
6-10 xlO 6
1-50 xlO 6
4-72 xlO 6
Gt
[%^K"*
Efficiency^
12-0
21-9
5-4
17-Q
14-1 ±3-5
t
Gt = Globin in immunoprecipitate
Mo = Message injected
Kg = Globin decay constant
Km = Message decay constant
* Calculated using the incorporation data in Table 1 for ova injected with 0-1 pg/pl (0-8 pg
injected per ovum) of globin mRNA, assuming that a-globin mRNA was 1/2 the message
injected and the molecular weight of 9S globin mRNA was 200000. An average of 1-2 x 106
molecules of a-globin mRNA was injected into each ovum. Efficiency equation was derived
by M. Pring at the University of Pennsylvania.
t Specific activity of [3H]leucine was 132CI/mmol. The a-globin molecule contains 18
leucine residues.
:(: Molecules of a-globin/cell/h per molecule of a-globin mRNA injected.
in acid precipitable material. Total incorporation of controls (3979 ± 738 dpm/
ova/h) and injected ova (4012 ± 773 dpm/ova/h) was not significantly different.
A four-fold increase in exogenous mRNA (4-0 to 15-8pg) did not significantly
increase the amount of a-globin synthesized (Table 1).
Efficiency of a-globin translation can be calculated employing the equation in
Table 2. The efficiency of rabbit a-globin translation was 14-1 ± 3-5 molecules
of a-globin/cell/h per molecule of a-globin mRNA injected. The half lives of
a-globin mRNA and a-globin following injection of mRNA into fertilized mouse
Translational efficiency of ova
165
ova are 8-8 and 8-2 h respectively (Chen & Brinster, 1982) assuming that all
mRNA injected was translatable upon injection and that the concentration of
mRNA was not saturated with respect to translational capacity.
DISCUSSION
Increasing the amount of rabbit a-globin mRNA injected into the fertilized
mouse ovum fourfold (4-0 to 15-8pg) did not augment a-globin synthesis nor
alter endogenous mRNA translation. This constant level of a-globin synthesis
indicates that the mouse ovum has little spare translational capacity for injected
message and suggests that during early development the quantity of protein
synthesized may not be determined solely by the availability of translatable
message but may be, in large part, a function of the cells' ability to translate.
Injection of foreign mRNA into Xenopus oocytes likewise has been shown not
to alter total protein synthesis, but the foreign mRNA did lower the amount of
protein synthesized from endogenous mRNA (Laskey et al. 1977). While aglobin synthesis in the mouse ovum accounts for only 1 % of total protein
synthesis, in Xenopus as much as 40 % of synthesized protein is globin (Laskey,
Mills, Gurdon & Partinton, 1977). Thus, the Xenopus oocyte, like the mouse
ovum lacks spare translational capacity; however, other differences in translational activity between the amphibian oocyte and the mammalian fertilized
ovum do exist (Brinster et al. 1980).
The fertilized mouse ovum has a translational efficiency of 14-1 molecules of
a-globin/cell/h produced for each molecule of a-globin mRNA injected.
Xenopus oocytes have a translational efficiency for globin mRNA of 13-8 (Asselbergs, Van Verooij & Bloemendal, 1978) and 20-0 (Gurdon, 1973) polypeptides/mRNA molecule/h. A direct comparison of efficiency does not take into
consideration the temperature at which translation occurs {Xenopus, 19°C;
mouse, 37°C). The rate of globin translation in the oocyte and ovum may be
related to the incubation temperature as was shown to occur in reticulocytes
(Conconni, Bank & Marks, 1966). Therefore, the calculated translational
efficiency of rabbit a-globin mRNA by the mouse ovum may be three times
lower than the Xenopus oocyte if the Qio of 3-2 is applied as suggested by Brandis
& Raff (1978).
Translational efficiency of rabbit a-globin mRNA in the fertilized mouse
ovum is approximately 40 times lower than in the intact reticulocyte (Hunt, 1974)
and 8 times lower than in microinjected Hela cell (Stacey & Allfrey, 1976). The
large difference in globin mRNA translation efficiency between the mouse ovum
and the reticulocyte may reflect, in part, the absence of haemin in the fertilized
mouse ovum. Haemin has been shown to be effective in increasing translation
of a-globin mRNA in frog oocytes by three-fold (Giglioni et al. 1973).
Since there does not appear to be any spare translational activity in the fertilized mouse ovum, the translation of microinjected (foreign) mRNA should be
166
K. M. EBERT AND R. L. BRINSTER
related to the ability of the foreign mRNA to compete with the endogenous
mRNA for in vivo translation. Total mRNA content of the fertilized mouse
ovum is ~24pg/ovum (Bachvarova & De Leon, 1980; Piko & Clegg, 1982).
Injection of 16pg of rabbit globin mRNA or a 67 % increase in total mRNA
resulted in only 1 % of the total incorporation in rabbit a-globin. In addition,
increasing the amount of exogenous globin mRNA (4-0 to 15-8 pg) did not result
in a significant increase in a-globin synthesis. Similarly, only 2-0 % of the total
incorporation (18 h labelling with [3H]leucine) in fertilized mouse ova was
labelled ovalbumin following the cytoplasmic injection of 2-7 pg of ovalbumin
mRNA (unpublished observation). This low level of competition is compatible
with the idea of a selective process for translation and degradation of maternal
mRNA by the fertilized ovum (see discussion, Kidder & Pedersen, 1982).
Exogenous mRNA competition experiments with Xenopus oocytes (Asselbergs,
Van Verooij & Bloemendal, 1979) show that globin mRNA translation can be
competitively inhibited by simultaneously increasing the concentration of a
second injected mRNA. This suggests the presence of a required translational
component that is limited in vivo. The lack of effective competition of the rabbit
globin mRNA with endogenous mRNA for either a similar translational component or access to the endogenous preformed translational machinery may
account for the absence of an increase in a-globin synthesis when the concentration of injected rabbit globin mRNA was increased.
Changes in the efficiency of mRNA translation during development may
contribute to the control of protein synthesis at the cytoplasmic level. This
control may be completely separate from the control of message translation by
the 'masking' of some maternal messages (Braude, Pelham, Flach & Lobatto,
1979). While the translational efficiency of cells varies for different injected
mRNAs (Brinster etal. 1980; Asselbergs etal. 1978; Gurdon, 1973; Chan, Kohler & O'Malley, 1976), a direct comparison of efficiencies of translation of
identical messages at different stages (i.e., growing oocytes to fertilized embryos), not yet reported, could contribute to our understanding of the changes
occurring in translation during development. The microinjection of mRNA
could be a valuable means of determining changes in translational activity that
may accompany the development of the mammalian ovum.
We thank Drs Lingrel and Boyer for providing us with rabbit globin mRNA and anti-rabbit
globin respectively. We also are grateful for the helpful suggestions of Dr Paynton and Dr
Hammer. The derivation of the efficiency formula was by Dr Pring.
Financial support was provided by NIH grant HD 15477 and NSF grants PCM 78-22931 and
PCM 81-07172. K.M.E. was a trainee on HD-06210.
REFERENCES
ASSELBERGS, A. M., VAN VEROOIJ, W. J. & BLOEMENDAL, H. (1978). Synthesis of lens crystallins in Xenopus oocytes as determined by quantitative immunoprecipitation. Eur. J.
Biochem. 87, 517-524.
Translational efficiency of ova
167
A. M., VAN VEROOIJ, W. J. & BLOEMENDAL, H. (1979). Messenger RNA competition in living Xenopus oocytes. Eur. J. Biochem. 94, 249-254.
BACHVAROVA, R. & D E LEON, V. (1980). Polyadenylated RNA of mouse ova and loss of
maternal RNA in early development. Devi Biol. 74, 1-8.
BOYER, S. H., SMITH, K. D., NOYES, A. N. & MULLEN, M. A. (1974). Immunological characterization of rabbit hemaglobin a and /? chain-synthesizing polysomes. /. biol. Chem. 249,
7210-7219.
BRANDIS, J. W. &RAFF, R. A. (1978). Translation of oogenetic mRNA in sea urchin eggs and
early embryos. Devi Biol. 67, 99-113.
BRAUDE, P., PELHAM, H., FLACH, G. & LOBATTO, R. (1979). Post-transcriptional control in the
early mouse embryo. Nature 282, 102-105.
BRINSTER, R. L. (1972). Cultivation of the mammalian embryo. In Growth, Nutrition and
Metabolism of Cells in Culture, Vol. II. (ed. G. Rothblat & V. Cristofalo), pp. 251-286.
New York and London: Academic Press, Inc.
BRINSTER, R. L., CHEN, H. Y., TRUMBAUER, M. E. & AVARBOCK, M. R. (1980). Translation
of globin messenger RNA by the mouse ovum. Nature 283, 499-501.
BRINSTER, R. L., CHEN, H. Y., TRUMBAUER, M. E. & PAYNTON, B. V. (1981). Secretion of
proteins by the fertilized mouse ovum. Expl Cell Res. 134, 291-296.
CHAN, L., KOHLER, P. O. & O'MALLEY, B. W. (1976). Translation of ovalbumin mRNA in
Xenopus laevis oocytes: Characterization of the system and effects of estrogen on injected
mRNA populations. /. Clin. Invest. 57, 576-585.
CHEN, H. Y. & BRINSTER, R. L. (1982). Stability of rabbit globin and its messenger RNA in
the mouse ovum. Expl Cell Res. In press.
CONCONNI, F. M., BANK, A. & MARKS, P. A. (1966). Polyribosomes and control of protein
synthesis: Effects of sodium fluoride and temperature in reticulocytes. /. molec. Biol. 19,
525-540.
GIGLIONI, B., GIANNI, A. M., COMI, P., OTTOLENGHI, S. &RUNGGER, D. (1973). Translational
control of globin synthesis by haemin in Xenopus oocytes. Nature New Biology 246,99-102.
GORSKI, J., MORRISON, M. R., MERKEL, C. G. & LINGREL, J. B. (1974). Size heterogeneity of
polyadenylate sequences in mouse globin messenger RNA. /. molec. Biol. 86, 363-371.
GURDON, J. B. (1973). The translation of messenger RNA injected in living oocytes of
Xenopus laevis. In Karolinska Symposia on Research Methods in Reproductive Endocrinology. 6th Symposium. Protein Synthesis in Reproductive Tissue (ed. E. Diczfalusy), pp.
225-237. Denmark: Bogtrykkeriet Forum Copenhagen.
HUNT, J. A. (1974). Rate of synthesis and half-life of globin messenger ribonucleic acid: Rate
of synthesis of globin messenger ribonucleic acid calculated from data of cell haemoglobin
content. Biochem. J. 138, 499-510.
JOHNSON, M. H., HANDYSIDE, A. H. & BRAUDE, P. R. (1977). Control mechanisms in early
mammalian development. In Development in Mammals II (ed. M. H. Johnson), pp. 67-97.
Amsterdam, New York, Oxford: North-Holland.
KIDDER, G. M. & PEDERSEN, R. A. (1982). Turnover of embryonic messenger RNA in
preimplantation mouse embryos. J. Embryol. exp. Morph. 67, 37-49.
KNOWLAND, J. & GRAHAM, C. (1972). RNA synthesis at the two-cell stage of mouse development. /. Embryol. exp. Morph. 27, 167-176.
KRYSTOSEK, A., CAWTHON, M. L. & KABAT, D. (1975). Improved methods for purification and
assay of eukaryotic messenger ribonucleic acids and ribosomes. J. biol. Chem. 250,
6077-6084.
LAEMMLI, U. K. (1970). Cleavage of structural proteins during the assembly of the head of
bacteriophage T 4 . Nature 227, 680-685.
ASSELBERGS,
LASKEY, R. A., MILLS, A. D., GURDON, J. B. & PARTINTON, G. A. (1977). Protein synthesis
in oocytes of Xenopus laevis is not regulated by the supply of messenger RNA. Cell 11,
345-351.
LEVEY, I. L., STULL, G. B. & BRINSTER, R. L. (1978). Poly (A) and synthesis of polyadenylated
RNA in the preimplantation mouse embryo. Devi Biol. 64, 140-148.
PIK6, L. & CLEGG, K. B. (1982). Quantitative changes in total RNA, total poly (A), and
ribosomes in early mouse embryos. Devi Biol. 89, 362-378.
168
K. M. EBERT AND R. L. BRINSTER
R. E., MCKNIGHT, G. S. & SCHIMKE, R. T. (1973). Quantitative measurement of
ovalbumin messenger ribonucleic acid activity. /. biol. Chem. 248, 2031-2039.
STACEY, D. W. & ALLFREY, V. G. (1976). Microinjection studies of duck globin messenger
RNA translation in human and avian cells. Cell 9, 725-732.
VANBLERKOM, J. &MCGAUGHEY, R. W. (1978). Molecular differentiation of the rabbit ovum.
I. During oocyte maturation in vivo and in vitro. Devi Biol. 63, 139-150.
YOUNG, R. J., SWEENEY, K. & BEDFORD, J. M. (1978). Uridine and guanosine incorporation
by the mouse one-cell embryo. /. Embryol. exp. Morph. 44, 133-148.
RHOADS,
{Accepted 6 November 1982)