/. Embryol. exp. Morph. Vol. 44,pp. 181-189, 1978
Printed in Great Britain © Company of Biologists Limited 1978
Tissue specificity for incorporation of
[ H]thymidine by the 10- to 12-somite mouse
embryo: alteration by acute exposure
to hydroxyurea
3
By SUE ANN MILLER 1 AND MEREDITH N. RUNNER 2
From the Department of Environmental, Population and Organismic Biology
and Department of Molecular, Cellular and Developmental Biology,
University of Colorado
SUMMARY
Radioautograms from .10- to 12-somite mouse embryos labeled for 30min in vitro with
[3H]thymidine were examined for frequency and intensity of incorporation. Results from
ten tissues showed that values ranged from 82% of nuclei with a mean of 16-6 grains for
visceral yolk sac to .17% of nuclei labeled with a mean of 4-4 grains for epithelium of the
anterior gut tube. Labeling in the ten tissues indicated (1) a tissue-specific spectrum of
incorporation of [3H]thymidine, (2) close correlation between frequency and intensity of
labeling within a tissue and (3) asymmetrical quantities of incorporation between right and
left somatopleure.
Treatment with hydroxyurea in vitro reduced the frequency of labeled nuclei by 85 % to
12% of control values. Mean numbers of grains over treated nuclei, 3-3-4-6 grains, were well
above background but were clustered below the low end of the control range. Tissues exposed to hydroxyurea showed (1) labeling of significant numbers of nuclei, (2) inhibition of
labeling in selected tissues and (3) equalization of bilateral asymmetry in quantity (frequency
and intensity) of incorporation in somatopleure. The selective reduction of thymidine
incorporation and equalization of asymmetrical rates of proliferation may constitute
mechanisms by which hydroxyurea causes abnormal morphogenesis.
INTRODUCTION
Comparative patterns of incorporation of DNA precursor into specific
tissues of the intact embryo contribute to understanding teratogenic responses.
Wegener, Hollweg & Maurer (1964) compared labeling indices of various fetal
tissues, while other investigators have examined individual embryonic tissues
(e.g. Janners & Searls, 1970; Searls & Janners, 1971; Stark & Searls, 1973,
chick wing; Kauffman, 1966, mouse neural tube; Ritter, Scott & Wilson, 1973>
rat limb). Comparative data from early embryonic stages has been reported
1
Author's address: Department of Life Sciences, Hamilton and Kirkland Colleges,
Clinton, New York 13323, U.S.A.
2
Author's address: Department of Molecular, Cellular and Developmental Biology,
University of Colorado, Boulder, Colorado, 80302, U.S.A.
182
S. A. MILLER AND M. N. RUNNER
(e.g. Barlow, Owen & Graham, 1972, 5- to 256-cell mouse; Solter, Skreb &
Damjanov, 1971, mouse egg-cylinder). Quantitative comparisons of incorporation for all tissues in the intact, early mammalian embryo during organogenesis
seem not to have been reported.
Hydroxyurea is teratogenic in the rat (Chaube & Murphy, 1966, 1973;
Murphy & Chaube, 1964; Philips et al. 1967; Scott, Ritter & Wilson, 1971).
The drug retarded DNA synthesis in the 8-5-day, 10- to 12-somite mouse within
15 min of in vitro exposure (Runner, 1975). Administration of hydroxyurea to
mice at this age produced fetal gastroschisis. Precisely controlled labeling of the
10- to 12-somite mouse was practicable because such embryos will undergo
sustained growth in vitro (Clarkson, Doering & Runner, 1969). The present
study attempts to associate the hydroxyurea-related anomaly, gastroschisis,
with tissue-specific rates of thymidine incorporation and with selective inhibition
by hydroxyurea.
We report tissue-specific differentials for incorporation of thymidine and
differential alteration of that specificity by presence of hydroxyurea. Hydroxyurea tended to equalize asymmetrical labeling of somatopleure and selectively
inhibited labeling of other tissues.
METHODS AND MATERIALS
Randomly bred female mice were opened on the ninth day (8-5 days of
gestation) after a vaginal plug was detected. Embryos were dissected in
phosphate-buffered saline and those that had developed to the 10- to 12-somite
stage by 13.00 h of the ninth day of gestation were prepared with intact visceral
yolk sac (Miller & Runner, 1975). Incubation took place in 35 mm Petri dishes
(Falcon no. 3001), gently agitated on a rocking tray inside a tissue culture
incubator at 37 °C in air. Four embryos per 35 mm Petri dish were incubated
for 30 min with l m l 3 x 10~7 M methyl [3H]thymidine (5/iCi/ml; sp.act.
17 Ci/mmole; Schwarz-Mann). Experimental embryos were pretreated in vitro
for 15 min with 4 X 1 0 ~ 3 M hydroxyurea (Sigma) followed by 30 min of incubation with label in continued presence of hydroxyurea. Embryos were then
prepared by standard histological methods, serially sectioned at 5/*m and
processed for radioautography (Kopriwa & LeBlond, 1962).
Regions of a section to be counted were selected objectively at low power
(100 x). Areas of tissue from which data were taken were then defined by placing
an ocular grid over the area selected at low power. The grid enclosed 15625 /mi2
at 400 x magnification in the predetermined area. Nuclei of all tissues represented in this study were 10-12 /im in diameter. Every fifth section of an
embryo was sampled to avoid multiple representation of nuclei in the data.
Background counts were approximately 30 grains per 15625 jam* or 0-15 grains
per nucleus. This density was insufficient to account for random placement of
three grains over the area of a nucleus 10 fim in diameter. Twenty times back-
Thymidine labeling of mouse embryo
183
Table 1. Incorporation of thymidine by the 10- to 12-somite mouse embryo
Grains over labeled nuclei
Tissue type
Yolk sac
Amnion
Open midgut endoderm
Posterior gut portal
Left somatopleure
Neural epithelium
Right somatopleure
Anterior gut portal
Posterior gut tube
Anterior gut tube
Labeling
index
Mean
Median
0-82
0-73
0-68
0-64
0-52
0-48
0-39
0-38
0-33
017
16-6
110
8-4*
91
85
7-2*
7-9
7-8
5-7
44
160
100
80
90
70
50
80
80
60
40
Nuclei No. of No. of
Mode counted embryos mothers
100
90
80
6,80
40
40
60
60
40
40
500
400
950
350
1400
1175
1350
350
350
350
.10
8
7
7
9
10
9
7
7
7
5
5
4
5
5
6
5
5
5
5
Embryos encapsulated in visceral yolk sac were incubated in tritiated thymidine (5 /tCi/ml;
sp.act. 17 Ci/mmole) for 30 min. Tissues are arranged in order of decreasing labeling index.
The order established by mean numbers of grains differs by displacing the open midgut and
the neural epithelium downward by two positions. These displacements (*) seem unimportant,
because they are changes within the intermediate group of tissues having relatively similar
mean numbers of grains, 7-2-91, and labeling indices, 0-38-0-68.
ground, i.e. three grains per nucleus, was adopted as the criterion that a nucleus
be classified as 'labeled'.
Seven to ten embryos representing four to six litters and 350 to 1400 replicate
counts were used in compiling the data. Chi-square tests of differences between
labeling indices were used for comparing the data. Probability values of 0-01
or less were considered significant.
RESULTS
Labeling indices (frequencies) and mean numbers of grains (intensities) for
ten tissues in 10- to 12-somite, untreated mouse embryos are arranged in order
of decreasing values in Table 1. Frequencies with which nuclei were labeled
(labeling index x 100) ranged from 82 % to 17 %, with highest frequencies in
yolk sac and amnion and lowest frequencies in gut endodermal epithelium.
Mean numbers of grains ranged from 16-6 over yolk sac nuclei to 4-4 over
nuclei of the anterior gut tube. Except for minor displacements noted in
Table 1, the rank order of tissues established by the number of grains per
nucleus is the same as the rank order established by the frequency of labeled
nuclei.
Mitotic figures were observed in all tissues studied but were rare in the
visceral yolk sac. No label was detected over mitotic figures in any of the ten
tissues.
The restricted distributions of frequency of numbers of grains over nuclei
from open midgut (Fig. 1A) has been illustrated because this tissue had an
184
S. A. MILLER AND M. N. RUNNER
s loo-i
Mean - 8-4 M e d j a n =
g
(A) Open midgut-control
Labelling index = 0-68
Total nuclei = 950
4
i i i i i i i i n
i r "i i i i r i I r i i i i i
6
8
10 12 14 16 18 20 22 24 26 28
Grains per labelled nucleus
•5 100- Mean = 3-5
I Median = 3
= 80-
(B) Open midgut-hydroxyurea
Labelling index = 016
Total nuclei = 900
73
1 60— —
•o
40-
I
20H
PI
30
l
I I I I I I I I I
10
I I I I I I 1 I I 1 I I 1
12 14 16 18 20 22
Grains per labelled nucleus
24
26
28
30
Fig. 1. Distributions of numbers of grains over nuclei in epithelium, of the open
midgut endoderm are representative of nine of the ten tissues in this study. Yolk
sac showed a wider distribution of numbers of grains in controls. Conditions for
labeling and treatment with hydroxyurea are described under Tables 1 and 2.
intermediate frequency of labeled nuclei. Nine of the ten tissues, i.e. all except
visceral yolk sac, showed distributions skewed toward the lower end of the
range of numbers of grains. The skew of distributions precluded comparison of
standard errors of numbers of grains and of labeling indices.
Both frequencies and intensities of labeled nuclei were also determined for
embryos that had been incubated in 4xl0~ 3 M hydroxyurea. The resulting
data are arranged in Table 2 in order of decreasing frequencies of labeled nuclei.
Labeling intensity, when hydroxyurea was present, ranged between 4-6 grains
per labeled nucleus in neural epithelium and 3-3 grains in the posterior gut tube.
Thymidine was incorporated into nuclei in the presence of inhibitor well above
background levels, but 9 of the 10 means fell below the low end of the control
range seen in Table 1.
The distribution of numbers of grains over labeled nuclei as modified by
hydroxyurea is illustrated for open midgut in Fig. 1B. The other nine tissues
showed similar reductions in the range of grains over nuclei. Hydroxyurea not
only reduced the numbers of grains, it also further skewed their distributions to
the left.
Percentage inhibition by hydroxyurea is included in Table 2. Somatopleure,
amnion and most areas of the gut endoderm, i.e. seven of the ten tissues, fell
Thymidine labeling of mouse embryo
185
Table 2. Effect of hydroxyurea on incorporation of thymidine
Tissue type
Yolk sac
Neural epithelium
Openmidgutendoderm
Anterior gut tube
Anterior gut portal
Posterior gut portal
Amnion
Right somatopleure
Left somatopleure
Posterior gut tube
Mean
Labeling Inhibition numbers Inhibition Nuclei No. of No. of
index
(%)
of grains
(%)
counted embryos mothers
0-46*
0-41
016*
015
012*
012*
011*
011*
008*
005*
44
15
76
12
68
81
85
72
85
85
3-8
46
3-5
3-7
3-8
3-8
3-6
41
3-8
3-3
77
36
58
16
51
58
67
48
55
42
500
.1075
900
400
400
350
400
650
650
350
8
8
6
8
8
7
8
7
7
7
3
3
3
3
3
3
3
3
3
3
Embryos were incubated 45 min in 4 x 10~3 M hydroxyurea and exposed to [ 3 H]thymidine
(as for Table 1) during the last 30 min. Mean numbers of grains were calculated for all
nuclei overlain by three or more grains. Hydroxyurea produced a significant difference in
labeling indices when compared to the respective labeling indices in Table 1. Chi-square tests
for eight of the ten tissues (*) gave probabilities of less than 0001.
between 85 % and 68 % inhibition of labeling by [3H]thymidine. Labeled
nuclei of neural epithelium and anterior gut tube showed 15% and 12%
inhibition respectively, and did not differ significantly from the control values.
Yolk sac, the most heavily labeled tissue in both control and treated embryos,
showed 44% inhibition (P< 0-001). Thus the three tissues, yolk sac, neural
epithelium and anterior gut tube, that were relatively resistant to inhibition by
hydroxyurea came from high, intermediate and low portions of the spectrum
of incorporation by control tissues, i.e. degree of inhibition by hydroxyurea
was not correlated with intensity of labeling in control tissues.
DISCUSSION
Differential labeling by thymidine -frequency of labeled nuclei
The 8-5-day mouse embryo appears to be a mosaic of cell populations
(Miller, 1974). The varied frequencies with which different tissues became
labeled, 17 % to 82 %, might be accounted for by (a) accessibility of cells to
label in the incubation medium, (b) efficiency of transport of thymidine into
cells, (c) withdrawal of significant numbers of cells from the proliferating
population, (d) sizes of intracellular precursor pools and/or (e) selective
preferences for salvage pathways during synthesis of DNA.
A companion study has shown that the visceral yolk sac of the 8-5-day
mouse embryo in vitro is permeable to thymidine and to hydroxyurea (Miller &
Runner, 1975). The current study suggests that accessibility to label cannot
account for the differences observed between the ten regions of the embryo,
186
S. A. MILLER AND M. N. RUNNER
since differences in frequencies of labeled nuclei did not correlate with proximity
to or contact with label in the incubation medium. One or all of the other four
possibilities could contribute to the observed regional differences in frequencies.
Tissue specificity in relative incorporation of thymidine - intensity
of nuclear label
The differences in labeled nuclei in the ten tissues can be compared in terms
of a second parameter, relative incorporation of thymidine, or labeling intensity.
Intensities with which nuclei became labeled during short-term exposure to
tritiated thymidine have been used to indicate that tissues differ in relative
rates of DNA synthesis (Flickinger, Freedman & Stambrook, 1967; Wegener
et al. 1964). Differential rates of DNA synthesis in eukaryotic cells have been
summarized by Callan (1972). He suggested that the more rapid synthesis of
DNA in embryonic cells than in adult cells is accomplished not by increasing
the actual rate of progress of the polymerase molecule but by activating larger
numbers of replicons. Such a model should be tested for tissue differences in
intensity of labeling during organogenesis.
Direct correlation between frequency and intensity of labeling
The close correlation between the frequency with which nuclei were labeled
in the different tissues and the intensity with which they were labeled must be
incorporated into any attempt to explain the significance of tissue differentials
during organogenesis. Existing evidence seems to eliminate the possibility that
the correlated tissue differentials are caused by technical artifacts. The tissue
differentials are unlikely to be due to populations being predominantly in early
or late S-phase (the rate of polymerization is said to be high during late Sphase), because there is no evidence that the populations studied were synchronized. The correlation is judged to reflect some inherent property of the various
tissues, such as differences in the relative length of the S-phase.
The amount of thymidine incorporated (i.e. frequency and/or intensity of
labeling) during short periods of exposure to label is customarily accepted as a
measure of rate of proliferation of cells and tissues (e.g. Antoniades, Stathako
& Scher, 1975; Gospodarowicz, 1975). Callan (1972) has summarized evidence
in support of mechanisms for variable rates of proliferation within an embryo.
A reasonable postulate would be that the correlated frequencies and intensities
of label with tritiated thymidine, seen in our data, are attributable to inherent
differences in length of DNA S-phase and its associated rate of proliferation
(see also Flickinger et al. 1967).
Asymmetry of labeling frequency in the somatopleure and its relation to rotation
The mouse embryo at 8-5 days shows remarkable morphogenetic asymmetry
as the axis of the body rotates and closure of the open midgut is effected.
Asymmetry of labeling in somatopleure was found in this study. Left somato-
Thymidine labeling of mouse embryo
187
pleure showed significantly higher frequency of labeled nuclei than the right
(P< 0-001). Midbody folding of somatopleure was first seen on the left side, so
that the left side appears to rotate around the more stationary right side.
The frequencies of labeling in somatopleure, along with the morphology,
suggest that rapid proliferation within the left somatopleure at the 10- to 12somite stage is an important asymmetry in rotation of the mouse embryo and
in closure of the open gut.
Incomplete inhibition by hydroxyurea
Comparison of Tables 1 and 2 shows two effects of hydroxyurea: (1) labeling
indices after exposure to hydroxyurea were lowered for all tissues studied and
(2) the rank order of labeling indices is altered by hydroxyurea.
It has been assumed, a priori, that regions of rapid proliferation within the
embryo would be the ones most susceptible to inhibitors of synthesis of DNA.
Such associations have led to interpretation of causal teratogenesis, e.g. Ritter
et al. (1973). Sinclair (1965, 1967) has shown that under the influence of
hydroxyurea, cultured mammalian cells are blocked during DNA synthesis and
eventually die. Nuclei in other phases of the cell cycle progress to but do not
enter the synthesis phase. It has been assumed to follow a posteriori that in an
embryo with divergent cell types, those tissues with rapidly proliferating cells,
i.e. those with a large percentage of cells in DNA synthesis, would show
massive cell death and be highly susceptible to inhibitory effects of hydroxyurea.
However, even after pretreatment with inhibitor, an appreciable number of
nuclei were capable of incorporating thymidine and presumably were synthesizing DNA (for example, 41 % and 46 % of nuclei in neural epithelium and
yolk sac respectively were labeled in the presence of hydroxyurea), and the
percentage inhibition by hydroxyurea failed to correlate with the frequency or
intensity with which nuclei were labeled in control tissues. Most notable is the
observation that yolk sac, the most frequently and most heavily labeled tissue,
and the tissue most exposed to the medium, showed only moderate sensitivity to
hydroxyurea.
When hydroxyurea gains entrance to cells, inhibition of labeling by [3H]thymidine is relatively uniform with respect to tissue groupings. Tissue differences
with respect to inhibition by hydroxyurea are most readily seen in frequency
of uptake of label. This contrast in effect of hydroxyurea is interpreted to represent
tissue differences in permeability to the drug which operate independently of
location in the embryonic mass. Once hydroxyurea has gained entrance to the
cell, inhibitory effects are similar, regardless of intensity of labeling in the absence
of hydroxyurea.
Gastroschisis
Gastroschisis has been induced in the rat embryo by hydroxyurea (Murphy
& Chaube, 1964; Chaube & Murphy, 1966, 1973; Scott et al. 1971). Exposure
188
S. A. MILLER AND M. N. RUNNER
of mouse embryos to hydroxyurea in utero at 8-5 days' gestation has resulted in
gastroschisis at 18 days' gestation (20 %, Runner, 1975). The explanted 8-5-day
mouse embryo, subjected to hydroxyurea, was tested for possible relation
between version and failure of closure of the body wall, i.e. gastroschisis.
Hydroxyurea resulted in 85 % inhibition of labeling in left and 72 % inhibition
in right somatopleure. The asymmetrical labeling differences of 52 % in left
and 39 % in right somatopleure were equalized under treatment by an agent
known to increase the frequency of gastroschisis. This association supports the
hypothesis that differential rates of proliferation are causally related to axial
rotation.
Selectively altered rates of proliferation, even in the absence of subsequent
cell death, may account for developmental anomalies. As an example, loss of
proliferative asymmetry is proposed as a mechanism by which hydroxyurea
can produce a high frequency of gastroschisis.
The authors wish to acknowledge support of this study by grants from the National
Institutes of Health, National Institute of Dental Research (DE-198) and from the National
Science Foundation (GB 14662) to M. N. Runner.
REFERENCES
H. N., STATHAKO, D. & SCHER, C. D. (1975). Isolation of a cationic polypeptide from human serum that stimulates proliferation of 3T3 cells. Proc. natn. Acacl.
Sci. U.S.A. 72, 2635-2639.
ANTONIADES,
BARLOW, P., OWEN, D. A. M. & GRAHAM, C. (1972). DNA synthesis in the preimplantation
mouse embryo. /. Embryol. exp. Morph. 27, 431-445.
H. G. (1972). Replication of DNA in the chromosomes of eukaryotes. Proc. R. Soc.
Lond. B181, 19-41.
CHAUBE, S. & MURPHY, M. L. (1966). The effects of hydroxyurea and related compounds
on the rat fetus. Cancer Res. 26, 1448-1457.
CHAUBE, S. & MURPHY, M. L. (1973). Protective effect of deoxycytidylic acid (CdMP) on
hydroxyurea induced malformations in rats. Teratology 7, 79-88.
CLARKSON, S. G., DOERING, J. V. & RUNNER, M. N. (1969). Growth of postimplantation
mouse embryos cultured in a serum-supplemented, chemically defined medium. Teratology
2, .18.1-186.
FLICKINGER, R. A., FREEDMAN, M. L. & STAMBROOK, P. J. (1967). Generation times and
DNA replication patterns of cells of developing embryos. Devi Biol. 16, 457-473.
GOSPODAROWICZ, D. (1975). Purification of a fibroblast growth factor from bovine pituitary.
/. biol. Chem. 250, 2515-2520.
JANNERS, M. Y. & SEARLS, R. L. (1970). Changes in rate of cellular proliferation during the
differentiation of cartilage and muscle in the mesenchyme of embryonic chick wing. Devi
Biol. 23, 136-165.
KAUFFMAN, S. L. (1966). An autoradiographic study of the generation cycle in the ten-day
mouse embryo neural tube. Expl Cell Res. 42, 67-73.
KOPRIWA, B. M. & LEBLOND, C. P. (1962). Improvements in the coating technique of radioautography. / . Histochem. Cytochem. 10, 269-284.
3
MILLER, S. A. (1974). Pattern of short-term, in vitro [ H]thymidine labeling in the tissues of
the 8i-day mouse embryo. Anat. Rec. 178, 418-419.
MILLER, S. A. & RUNNER, M. N. (1975). Differential permeability of murine visceral yolk
sac to thymidine and to hydroxyurea. Devi Biol. 45, 74-80.
CALLAN,
Thymidine labeling of mouse embryo
189
MURPHY, M. L. & CHAUBE, S. (1964). Preliminary survey of hydroxyurea (NSC-32065) as a
teratogen. Cancer Chemother. Rept. 40, 1-7.
PHILIPS, F. S., STERNBERG, S. S., SCHWARTZ, H. S., CRONIN, A. P., SONDERGREN, J. E. &
VIDAL, P. M. (1967). Hydroxyurea. I. Acute cell death in proliferating tissues in rats.
Cancer Res. 27, 61-74.
RITTER, E. J., SCOTT, W. J. & WILSON, J. G. (1973). Relationship of temporal patterns of
cell death and development to malformations in the rat limb. Possible mechanisms of
teratogenesis with inhibitors of DNA synthesis. Teratology 7, 219-226.
RUNNER, M. N. (1975). Biosynthesis in the mouse embryo as affected by hydroxyurea. In
New Approaches to the Evaluation of Embryonic Development (ed. D. Neubert & H. J.
Merker), pp. 591-613. Stuttgart: G. Thieme Pub.
SCOTT, W. j . , RITTER, E. J. & WILSON, J. G. (1971). DNA synthesis inhibition and cell
death associated with hydroxyurea teratogenesis in rat embryos. Devi Biol. 26, 306-315.
SEARLS, R. L. & JANNERS, M. Y. (1971). The inhibition of limb bud outgrowth in the
embryonic chick. Devi Biol. 1A, 198-213.
SINCLAIR, W. K. (1965). Hydroxyurea: differential lethal effects on cultured mammalian
cells during the cell cycle. Science, N.Y. 150, 1721-1731.
SINCLAIR, W. K. (1967). Hydroxyurea: effects on Chinese hamster cells grown in culture.
Cancer Res. 27, 297-308.
SOLTER, D., SKREB, N. & DAMJANOV, I. (1971). Cell cycle analysis in the mouse egg cylinder.
Expl Cell Res. 64, 331-334.
STARK, R. J. & SEARLS, R. L. (1973). A description of chick wing bud development and a
model of limb morphogenesis. Devi Biol. 33, 138-153.
WEGENER, K., HOLLWEG, S. & MAURER, W. (1964). Autoradiographische Bestimmung der
DNA-Verdopplungszeit und anderer Teil-phasen des Zell-syklus bei fetalen Zellarten der
Ratte. Z. Zellforsch. mikrosk. Anat. 63, 309-326.
{Received 27 August 1975, revised 24 October 1977)
13
EMB 44