/ . Embryol. exp. Morph. Vol. 57, pp. 61-70, 1980
Printed in Great Britain © Company of Biologists Limited 1980
Qualitative changes in DNA indicating differential
DNA replication during early embryogenesis of
the newt Triturus vulgaris
By KLAUS LOHMANN 1 AND LORE SCHUBERT
From the Institute of Zoology, University of Koln
SUMMARY
During gastrulation of the newt Triturus vulgaris considerable changes in the melting
behaviour and in the CsCl density gradient pattern of DNA occur. The melting curves of
DNAs from mid to late gastrulae (stages 11 b-12c) deviate from the regular shape. Whereas
the Tm values are identical in the stages 11 a (early gastrula) and 15 (early neurula), and
correspond to the standard DNA (stage 36 = tailbud), a significant rise of Tm (0-8-1 -2 °C)
has been recorded in the stages 12 a/b (yolk plug). The differences in melting behaviour become
visible by the deviation of the curves above Tm. These deviations from the normal sigmoidal
shape are caused by the fact that a portion of DNA melts at higher temperatures than usual.
Therefore the thermal denaturation of DNA is completed at approximately 3-4 °C later than
in standard DNA. Both the derivative curves and the plots on normal probability paper
demonstrate a heterogeneity of DNA in the stages 11 b-12c which indicates the presence of
an additional GC-rich satellite fraction. These findings are confirmed by CsCl density
gradient studies. Thus, in the stages 12a/b a slight shoulder on the heavy side of the gradients
occurs, being absent in the other stages. From these facts we have to conclude that there is a
stage-dependent multiplication and elimination of GC-rich sequences.
INTRODUCTION
In the last decade many exceptions from the rule of DNA constancy have been
found which can be classified into four major groups: gene amplification, DNA
magnification, DNA under-replication and DNA elimination (Brown &Dawid,
1968; Ritossa, 1968; Kaback & Halvorson, 1977; Rudkin, 1969; Beermann,
1977; see Tobler, 1975 and Nagl, 1976 for review). The best-studied case of gene
amplification involves the excessive multiplication of gene copies coding for
ribsomal RNAs (rDNA) in oocytes of many vertebrates and invertebrates (Gall,
1969; Lima de Faria, Birnstiel & Jaworska, 1969). Both the fact that gene amplification is not restricted to oogenesis but also occurs within somatic cells, for
example in the DNA puffs of sciarids (Pavan & Da Cunha, 1969), and the
phenomenon of DNA under-replication in the course of polytenization of some
insect tissues (see Laird, 1973 for review) led to the assumption that differential
1
Author's address: Zoologisches Institut der Universitat zu Koln, Lehrstuhl fur Experimentelle Morphologie, Weyertal 119, D 5000 Koln 41 B.R.D.
5
EMB
57
62
K. LOHMANN AND L. SCHUBERT
DNA replication may be a more common mechanism of differentiation in
eukaryotic systems than is currently believed.
Under this aspect we studied the early phases of determination and differentiation in the development of the newt Triturus vulgaris with respect to quantitative
and qualitative changes in the nuclear DNA. From cytophotometric measurements it became evident that the nuclear DNA content is not constant during
early development, but varies with stage and region specifically (Lohmann &
Vahs, 1969; Lohmann, 1972). To answer the question whether there is a stagespecific amplification of DNA sequences differing in the mean guanosinecytosine content from bulk DNA, in the present study the DNA isolated from
various developmental stages was characterized by thermal denaturation analysis
and CsCl density gradient centrifugation. A short report on this subject has been
presented elsewhere (Lohmann & Schubert, 1978).
MATERIALS AND METHODS
Animals. All studies were carried out with embryos of Triturus vulgaris from
the early gastrula to the early neurula stage (Harrison stages 10-15). Late tailbud
stages (stages 35, 36) have been used for reference. In order to obtain sufficient
numbers of eggs the freshly collected females were treated with human gonadotropin (Predalon, Organon GmbH, Miinchen). In 24 h intervals three doses of
3 i.u. each were injected. In order to prepare homogeneous egg material, the eggs
were collected within 3 h intervals and kept at 18 °C in Holtfretersolution, containing Cibazol (0-1 %, Ciba, Wehr/Baden) and penicillin (100 i.u. per ml), until
the desired stage was reached. In the gastrula stages 50-60 dejellied embryos
were pooled and preserved in liquid nitrogen until used.
DNA isolation. The isolation of nuclei was performed according to the methods
of Imoh & Minamidani (1973) and Faulhaber (1972). From the nuclei fractions
the DNA was prepared by the method of Marmur (1961) with some modifications. From the supernatant obtained after the removal of RNA by RNase
(Ribonuclease A, 100 /*g/ml) and deproteinization with an equal volume of
chloroform-isoamyl alcohol (24:1, v/v) the DNA was precipitated by ethyl
alcohol containing 0-2 M sodium acetate and dissolved in 0-1 SSC (0015 M-NaCl,
0-0015 M-Na3 citrate, pH 7-0). The remaining amounts of polysaccharides and
proteins were removed by digestion with a-amylase (100 /*g/ml) and pronase
(500 /tg/ml), followed by two treatments with chloroform-isoamyl alcohol. After
dialysing the supernatant against 0-1 SSC the DNA was precipitated with ethanol
and washed several times with 80 % ethanol.
Thermal denaturation. The melting profiles of native DNA were obtained as
described by Mandel & Marmur (1968) with a Zeiss PM Q2 spectrophotometer
equipped for thermal denaturation studies. In all cases the DNA was dissolved
in 0-1 SSC. The samples showed a 260/280 nm absorption ratio of 1-76-1-82,
and a 230/260 nm ratio of 0-42-0-47. The hyperchromicity values of different
Differential DNA replication in Triturus embryogenesis
63
H (%)
100
80 60
80 60
80 60
Temperature (°C)
80 60
80 60
80
Fig. 1. Upper part: thermal denaturation profiles of DNA from different developmental stages (early gastrula to early neurula, and tailbuds). Ordinate: relative
absorbance (hyperchromicity) (percentages). Abscissa: temperature (°C). Lower
part: the appropriate derivative melting curves.
samples varied between 33 and 36 %. The mean Tm value of 28 melting curves of
' standard'-DNA (stage 36) is 69-43, with a standard error a = 0-15. The plots
on normal probability paper were carried out according to Knittel, Black,
Sandine & Fraser (1968).
CsCl ultracentrifugation. Native unsheared DNA was centrifuged in a Beckman preparative ultracentrifuge in a type-65 fixed-angle rotor. Centrifugation
was performed in neutral CsCl at 33000 rev./min for 65 h at 25 °C. In all cases
10 ml gradients, overlaid with paraffin oil, were loaded with equal amounts of
DNA (0-6 OD unit, approx. 30 ju,g) from different developmental stages. The
initial density of the gradients was 1-695 g/ml at 25 °C. After centrifugation
the content of each tube was fractionated by the bottom harvesting technique.
The u.v. gradient profiles were continuously recorded by using a flow-through
micro-cuvette in a ZeissPM Q2 spectrophotometer. Forty-six fractions of 185/^1
each were collected and the density of each fraction was determined by a Zeiss
refractometer.
RESULTS
1. The melting behaviour of DNA
Figure 1 illustrates the melting behaviour of DNA isolated from different
developmental stages. Parallel to each embryonic DNA the standard DNA of
late tailbuds (stage 36) has been measured. Standard Triturus DNA has a regular
melting profile with a mean Tm value of 69-4 °C, whereby the extreme values
are at 69-2 and 69-6 °C, respectively. Corresponding melting curves have been
recorded in the early gastrula stages (10/1 la) and early neurula (stage 15). In
the stages 116-12 c, however, the melting profiles deviate from the regular
5-2
K. LOHMANN AND L. SCHUBERT
64-
64-
•76
64-
-78
64-
•80
65
65-
-79
•77
-77
Temperature (°C)
Fig. 2. Thermal denaturation curves of DNA (stages as in Fig. 1), plotted on
normal probability paper.
sigmoidal shape at temperatures above Tm being accompanied by an increase of
the Tm value. Maximal values have been found in the stages \2afb that range
in different samples between 0-8 to 1-2 °C above the mean Tm of standard DNA.
The deviations from the normal shape become visible by the flattening of the
curves at lower hyperchromicity values (85 % H) than usual, and moreover by
the fact that the thermal denaturation of DNA is completed later (about 3-4°
in the stages 12 a/6) than in the other stages.
If we look at the derivative curves (lower part of Fig. 1) and the plots on
normal probability paper (Fig. 2), these changes in the melting profiles of DNA
can be most clearly seen. The melting behaviour of DNAs from the stages
10/1 \a, 15 and 36 corresponds to the Gaussian distribution with respect to the
mean base composition. The derivative curves of the stages 11 b-12c, by contrast,
show a shoulder at higher temperatures which indicates the presence of a second
DNA fraction. The latter is also demonstrated by the fact that on normal probability paper the hyperchromicity values deviate from the straight line, yielding
a second line at H-values above Tm. From the comparison of the temperature
scales in Fig. 2 it can be easily seen that the endpoint of thermal denaturation is
shifted to higher temperatures in the mid gastrula stages.
2. CsCl density gradient patterns
The characterization of DNA in neutral CsCl gradients yields corresponding
curves for the early gastrula (stage 10/11 a), early neurula (stage 15), and tailbuds
(stage 36). Besides the main band a slight shoulder (arrow) is present on the
lighter side of the gradient (Fig. 3d). In contrast, in the stages 12a/b (Fig. Ad)
a little deviation from the reference curve becomes visible on the heavy side of
the gradient too (arrow). In order to make this deviation more evident the
Differential DNA replication in Triturus embryogenesis
40
0-2
T
- 1-76 0-2 - r l O
20
0
Fraction number
(c)
r 1-76
r 1-76 0-2 -r 1 0
- 1-70
--
(d)
•70
L 1-64
- 1-64
-0-5
40
20
0
Fraction number
65
01 --
40
20
Fig. 3. CsCl density gradient patterns of D N A from the stages lla, 15 and 36
(Fig. 3«). The buoyant density profiles are shown in the upper part. In Fig. 3b-d
the heavy fractions 10-18 (d), the middle fractions 19-27 (c), and the light fractions
28-42 {b) of three identical gradients were pooled and recentrifuged.
fractions 10-18, 19-27 and 28-42 of three identical gradients were pooled and
recentrifuged in neutral CsCl. As expected, a symmetric profile has been monitored for the middle fractions (19-27, Fig. 4c); in the light fractions (28-42) the
shoulder on the light side appears more prominently (Fig. 4 b), and the mentioned
deviation on the heavy side of Fig. 4a can now be clearly seen in the fractions
10-18 as a significant shoulder (Fig. 4d). In the same manner the DNAs from
the stages 10/1 la, 15 and 36 have been fractionated, whereby no deviations
could be detected on the heavy side of the gradient (Fig. 3d). For the other
fractions, curves have been recorded (Fig. 3 b and c) corresponding with those of
Fig. 4.
On the one hand the density difference of DNA which causes the heavy shoulder
is obviously too small to separate itself from the main band as a satellite peak.
On the other hand the portion of this fraction is so small that the formation of a
separate peak is suppressed solely by the amount of main band DNA. To reduce
the portion of main band DNA significantly the heavier fractions (10-18) of one
66
K. LOHMANN AND L. SCHUBERT
r 1-76 >•
0-2 -r 10
0 1 --
40
0-2
- 1-76 0-2
T
T
10
20
0
Fraction number
r 176 0-2 -r 1 0
rl-76
- 1-70
- 1-64
0 1 -- 0-5
01 --
40
20
40
20
0
Fraction number
Fig. 4. CsCl density gradient profiles of DNA from the mid-gastrula stages 12a/b
(a). In (b)-(d) the same fractionation and recentrifugation was carried out as in
Fig. 3.
gradient with DNA from the stages 12a/b (Fig. 5, upper part) were pooled and
recentrifuged. The result is presented in the lower part of Fig. 5. Now, a separate,
heavy satellite peak (arrow) occurs. In order to determine the buoyant density of
Triturus DNA, gradients loaded with equal amounts of Triturus and Micrococcus
luteus marker DNA (30 fig each) were analysed. From these tracings the buoyant
density of Triturus main band DNA was estimated with 1-702 g/ml. With respect
to this value the buoyant density of the shoulder on the lighter side is 1-686 g/ml
and that of the stage-specific one on the heavy side of the gradient is 1-719 g/ml.
According to the formula of Schildkraut, Marmur & Doty (1962) the calculated
GC contents of the main band, the light and the heavy shoulder are 43 %,
27 % and 60 % respectively.
DISCUSSION
As shown by the thermal denaturation curves, significant changes in the
melting behaviour of DNA occur during gastrulation of Triturus vulgaris. The
deviation from the normal shape and the increase of the Tm value is caused by
Differential DNA replication in Triturus embryogenesis
67
0-2 -r
01 --
002 -40
20
0
Fraction number
Fig. 5. (a) Density gradient profile of gastrula DNA (stage 12a/b). (b) The gradient
pattern of the DNA from the fractions 10-18 of the upper gradient (Fig. 5 a).
the fact that a portion of DNA melts at higher temperatures than usual. From
the stage-specificity of the curves we have to conclude that the amount of the late
melting fraction increases from the early gastrula up to the yolk-plug stage and
becomes reduced in the following stages down to the normal level. There are
essentially two possibilities to explain the occurrence of a late melting fraction.
First, by selective methylation of DNA. For instance Dawid, Brown & Reeder
(1970) have shown that chromosomal rDNA has a higher thermal stability than
amplified rDNA which contains no 5-methyl cytosine. Our results pose the old
question of whether DNA modifications are generally involved in cell differentiation, as discussed by Scarano (1971) and Holliday & Pugh (1975).
Secondly, by amplification of DNA sequences, possessing a higher GC-content
than bulk DNA.
In comparison with the melting curves, by ultracentrifugation only slight
differences in the gradient patterns have been found between the stages studied.
Since it is well known that thermal satellites do not always appear as satellite
fractions in CsCl density gradients, and because we cannot decide whether and to
what extent methylation of DNA participates in the melting behaviour, this
point will not be discussed further. The small number of measurements at lower
temperatures, when AT-rich sequences denature preferentially, accounts for the
fact that the permanently present AT-rich satellite fraction, shown in the density
gradient patterns, has not been visualized in the melting profiles.
The occurrence of the shoulder on the heavy side of the gradient, however,
gives evidence for the existence of a stage-specific DNA fraction with a higher
density than main band DNA. Then, the appearance of a heavy satellite fraction
cannot be caused by methylation of DNA, because this phenomenon reduces
the buoyant density of DNA. On the other hand this argument includes the
68
K. LOHMANN AND L. SCHUBERT
possibility that this fraction is relatively under-methylated as a result of DNA
synthesis in the absence of methylation. But this effect could not account for the
differences in the melting curves. Therefore we think that the presented changes
in the melting behaviour together with the differences in the density gradient
patterns can be plausibly explained by differential DNA replication, precisely by
a stage-dependent amplification of GC-rich DNA sequences.
These results confirm our earlier studies in which we postulated that the
increase of nuclear DNA content in the late gastrula is at least partly caused by
amplification of GC-rich sequences, possibly rDNA (Lohmann, 1972; Lohmann
& Schubert, 1977). Filter hybridization experiments with radioactive ribosomal
RNA from HeLa cells (done very recently by Schubert (in preparation)), show
that the stage-specific satellite fraction contains amplified rDNA to only a small
extent. The function of the major component of the amplified DNA is not
known, but there is much indication that the stage-dependent increase of DNA
apparently serves as an enhancement of the transcription capacity in definite
regions of the embryo. Then, with a phase-shift of about 2 h the DNA amplification during gastrulation and neurulation is followed by a remarkable short-term
increase of the nuclear RNA and non-histone protein content (Lohmann &
Jansen, 1976; Lohmann, 1979).
With respect to the relative amount of the additional DNA fraction, there is a
quantitative discrepancy between the cytophotometric measurements and the
presented findings. Since a correct estimation of the proportion of GC-rich
sequences is not possible from the gradient profiles shown in this paper, we
reproduced some of our experiments using a Gilford spectrophotometer (Type
250) equipped with a density gradient scanner. These tracings, which have been
recorded without changing the absorption range, show, besides the permanently
present AT-rich fraction (about 10 % of total DNA), a distinct GC-rich shoulder
of about 3-4 % of total DNA. This amount is obviously low compared to the
above-mentioned changes in the Feulgen-DNA content (ranging between 15
and 35 % in various regions of the embryo). If we consider that the Feulgen
values reflect changes in DNA content (this point has been discussed in detail by
Lohmann, 1975), then the cited discrepancy gives indirect evidence that not only
GC-rich sequences are amplified during early Triturus development but also,
and to a greater extent, DNA sequences with a mean GC-content similar to the
main band DNA.
While it is generally accepted that gene amplification is a common mechanism
during oogenesis, the occurrence of selective gene amplification in proliferating
somatic cells is still a very disputed question (e.g. Tobler, 1975; Wintersberger,
1978). Since the amplified DNA is accumulated for a period of a few cell cycles
(in neurula cells the S-phase lasts about 4 h (Callan, 1973)) there has to be a
mechanism that prevents the dilution of (extrachromosomally ?) amplified DNA
in the following mitoses, possibly by a temporal limited integration into the
genome. Evidence for the existence of such a mechanism comes also from recent
Differential DNA replication in Triturus embryogenesis
69
studies on ribosomal gene amplification in mitotically dividing spermatogonia
and oogonia of Xenopus (Bird, 1978) and on the ribosomal magnification in
Drosophila melanogaster (Graziani, Gaizzi & Gargano, 1977). In both cases
extra-chromosomal rDNA rings can be integrated into the chromosome.
Furthermore, Alt, Kellems, Bertino & Schimke (1978) could show that in
methotrexate-resistant lines of Sarcoma 180 murine lymphoma cells the increased
synthesis of dihydrofolate reductase is associated with a proportional increased
number of copies (200-250-fold) of the dihydrofolate reductase structural genes.
These findings indicate that in proliferating cells the phenomenon of gene amplification is not restricted to the ribosomal RNA genes, but that the machinery
required to amplify structural genes exists. In summary, we would like to interpret our results as consistent with the concept that the genetic information in
the eukaryotic genome is not statically fixed, but undergoes dynamic changes in
the course of embryonic differentiation.
This research was supported in part by the Ministerium fur Wissenschaft und Forschung
des Landes Nordrhein-Westfalen.
REFERENCES
W., KELLEMS, R. E., BERTINO, J. R. & SCHIMKE, R. T. (1978). Selective multiplication
of dihydrofolate reductase genes in methotrexate-resistant variants of cultured murine cells.
/. biol. Client. 253, 1357-1370.
BEERMANN, S. (1977). The diminution of heterochromatic chromosomal segments in Cyclops.
Chromosoma (Berl.) 60, 297-344.
BIRD, A. P. (1978). A study of early events in ribosomal gene amplification. Cold Spring Harb.
Symp. Quant. Biol. 42, 1179-1183.
BROWN, D. D. & DAWID, I. B. (1968). Specific gene amplification in oocytes. Science 160,
272-280.
CALLAN, H. G. (1973). Replication of DNA in eukaryotic chromosomes. Br. med. Bull. 29,
192-195.
DAWID, I. B., BROWN, D. D. & REEDER, R. H. (1970). Composition and structure of chromosomal and amplified ribosomal DNAs of Xenopus laevis. J. mol. Biol. 51, 341-360.
FAULHABER, I. (1972). Die Induktionsleistung subzellularer Fraktionen aus der Gastrula von
Xenopus laevis. Wilhelm Roux Arch. EntwMech. Org. Ill, 87-108.
GALL, J. G. (1969). The genes for ribosomal RNA during oogenesis. Genetics Suppl. 61,
121-132.
GRAZIANI, F., CAIZZI, R. & GARGANO, S. (1977). Circular ribosomal DNA during ribosomal
magnification in Drosophila melanogaster. J. mol. Biol. 112, 49-63.
HOLLIDAY, R. & PUGH, J. E. (1975). DNA modification mechanism and gene activity during
development. Science 187, 226-232,
IMOH, H. & MINAMIDANI, T. (1973). Changes in protein during development of Triturus
embryos. /. Embryol. exp. Morph. 30, 647-659.
KABACK, D. B. & HALVORSON, H. O. (1977). Magnification of genes coding for ribosomal
RNA in Saccharomyces cerevisiae. Proc. natn. Acad. Sci., U.S.A. 74, 1177-1180.
KNITTEL M. D., BLACK, C. H., SANDINE, W. E. & FRASER, O. K. (1968). Use of normal
probability paper in determining thermal melting values of deoxyribonucleic acid. Canad.
J. Microhiol. 14, 239-245.
LAIRD, C. D. (1973). DNA of Drosophila chromosomes. Ann. Rev. Genet. 7, 177-204.
LIMA DEFARIA, A., BIRNSTIEL, M. &JAWORSKA, H. (1969). Amplification of ribosomal cistrons
in the heterochromatin of Acheta. Genetics Suppl. 61, 145-159.
ALT, F.
70
K. LOHMANN AND L. SCHUBERT
K. (1972). Untersuchungen zur Frage der DNS-Konstanz in der Embryonalentwicklung (Urodela, Triturus vulgaris). Wilhelm Roux Arch. EntwMech. Org. 169, 1^40.
LOHMANN, K. (1975). Das Hydrolyseverhalten des Chromatins von Zellkernen aus friihembryonalen 7W/Mn/.s-Geweben. Zur Frage der Proportionality von Feulgen-Werten und
DNS-Mengen. Wilhelm Roux' Arch, devl Biol. 177, 285-299.
LOHMANN, K. (1979). Tissue-specific changes in nuclear RNA content during early development of Triturus vulgaris. Histo chemistry 63, 47-56.
LOHMANN, K. & JANSEN, U. (1976). Cytochemical evidence for stage-specific changes of
nuclear RNA and nonhistone protein content during early development of Triturus vulgaris
Experientia {Basel) 32, 380-381.
LOHMANN, K. & SCHUBERT, L. (1977). Changes in the thermal denaturation profiles of DNA
from different developmental stages of the newt Triturus vulgaris. Experientia {Basel) 33,
1518-1519.
LOHMANN, K. & SCHUBERT, L. (1978). Differentielle DNS-Replikation in der Embryogenese
des Teichmolchs Triturus vulgaris. Verh. dtsch. Zool. Ges. 1978, 234.
LOHMANN, K. & VAHS, W. (1969). Cytochemischer Nachweis phasen- und regionsspezifischer
Yeranderungen des DNS-Gehaltes wahrend der Embryogenese von Triturus vulgaris.
Experientia {Basel) 25, 1315-1316.
MANDEL, M. & MARMUR, J. (1968). Use of ultraviolet absorbance-temperature profile for
determining the guanine plus cytosine content of DNA. In Methods in Enzymology, xnb
(ed. L. Grossman & K. Moldave), pp. 195-206.) New York-London: Academic Press.
MARMUR, J. (196.1). A procedure for the isolation of deoxyribonucleic acid from microorganisms. /. mol. Biol. 3, 208-217.
NAGL, W. (1976). Zellkern undZellzyklen. Stuttgart: Eugen Ulmer Verlag.
PAVAN, C. & DA CUNHA, A. B. (1969). Gene amplification in ontogeny and phylogeny of
animals. Genetics Suppl. 61, 289-304.
RITOSSA, F. M. (1968). Unstable redundancy of genes for ribosomal RNA. Proc. natn. Acad.
ScL, U.S.A. 60, 509-516.
RUDKIN, G. T. (1969). Non replicating DNA in Drosophila. Genetics Suppl. 61, 227-238.
SCARANO, E. (1971). The control of gene function in cell differentiation and in embryogenesis.
Adv. Cytopharmacol. 1, 13-23.
SCHILDKRAUT, C. L., MARMUR, J. & DOTY, P. (1962). Determination of the base composition of deoxyribonucleic acid from its buoyant density in CsCl. / . mol. Biol. 4, 430443.
TOBLER, H. (1975). Occurrence and developmental significance of gene amplification. In
The Biochemistry of Animal Development, III (ed. R. Weber), pp. 91-173. New YorkLondon, Academic Press.
WINTERSBERGER, E. (1978). DNA replication in eukaryotes. Rev. Physio/. Biochem. Pharmacol.
84,93 143.
LOHMANN,
(Received 27 June 1979, revised 20 December 1979)