Molecular Marine Biology and Biotechnology (1998) 7(4), 248–258
Effect of exogenous DNA microinjection on early
development response of the seabream (Sparus aurata)
Sonia Garcı́a-Pozo, Julia Béjar, Monica Shaw,
and M. Carmen Alvarez*
Departamento de Genética, Facultad de Ciencias,
Universidad de Málaga, 29071 Málaga, Spain
fulness of GFP for diagnosis of transgenesis is enhanced by the transparency of embryos and larvae
in S. aurata.
Introduction
Abstract
DNA transfer techniques allow genetic manipulation of commercial fish. However, marine species
have received little attention because of their difficult zootechnical requirements. The seabream
(Sparus aurata) has become one of the most important species in the aquaculture of Mediterranean
countries, and the development of suitable DNA
transfer procedures represents a main step in its
genetic improvement. To assess the response of the
seabream to exogenous DNA, naturally fertilized
eggs were injected with the plasmids pCMV-CAT,
pCMVTklacZ, and pEGFP-N1, in supercoiled and
linearized forms. Embryo and larval survival, DNA
fate, and reporter gene expression were analyzed
during early development. The survival results indicate that microinjection is an effective transfer
method in spite of the unfavorable conditions. Linearized plasmids were more efficiently polymerized than supercoiled ones; however, no significant
differences were detected either in their persistence or in their expression levels. Reporter gene
expression was initiated after mid-blastula transition. The duration of transient expression varied
between the promoter-gene combinations, and no
integration of transgenes into fish chromosomes
was detected. Results suggest that the main factor
affecting the persistence and expression of DNA
seems to be related to developmental processes.
Among the markers used, CAT proved to be the
most sensitive, but GFP had obvious methodologic
advantages over the spatial marker lacZ. The use-
* Correspondence should be sent to this author; tel 34-52131967; fax 34-5-2132000; e-mail [email protected].
娀 1998 Springer-Verlag New York Inc.
248
The transfer of exogenous DNA to fish and to other
animal embryos has two main objectives, the study
of the function of regulatory sequences during development and the genetic manipulation of commercial fish species to produce transgenic lines
with higher commercial value. For the former,
short-term or transient expression assays have been
preferred (Chong and Vielkind, 1989; Gong and
Hew, 1993; Reinhard et al., 1994) owing to the difficulties in obtaining suitable long-term expression
strains. So far, foreign genes have been introduced
mostly by microinjecting newly fertilized eggs with
DNA constructs, though other innovative procedures to improve efficiency have been attempted
with variable success (reviewed in Iyengar et al.,
1996).
Transgenic fish from model and commercial
species have been produced in the last decade, although most experiments have been restricted so
far to freshwater teleosts (reviewed in Maclean and
Rahman, 1994), partly because they are easier to
maintain in the laboratory and partly because of
their tradition in aquaculture. In contrast, marine
species have received much less attention despite
their increasing importance in aquaculture (Gordin, 1989).
The marine fish S. aurata has emerged in the last
few years as one of the most important species in
the aquaculture of the European Mediterranean
countries. It can be considered as a model in the
domestication process of various other sparid fish.
Consequently, there is considerable interest in the
isolation and characterization of genes involved in
the productive traits of this species for possible genetic improvement.
Because evolutionary diversity among fish is
greater than in all other vertebrate classes together,
the use of species-specific regulatory elements for
Behavior of exogenous DNA in Sparus aurata
Figure 1. DNA constructs used in the microinjection
process after linearization with the corresponding restriction enzyme (see text): (a) the pCMV-CAT; (b) the pCMVTklacZ; and (c) the pEGFP-N1 plasmids. Bacterial
plasmid regions are represented with striped boxes, open
boxes are the regulatory regions, closed boxes are the coding sequences, and shaded boxes are the polyadenylation
sequences. The relevant restriction sites are indicated.
transgene expression is recommended when possible (Devlin et al., 1994; Takagi et al., 1994). This
work is a step toward using S. aurata as an in vivo
system for studying the function of autologous
regulatory sequences, mostly related with genes involved in productive traits. Previous to these studies it was convenient to assess the response of S.
aurata to DNA transfer processes. For this purpose
the plasmids pCMV-CAT, pCMVTklacZ, and
pEGFP-N1 were injected. The embryo and larval
survival rates and the respective DNA fate and expression during the early development were evaluated. The relative utility of the three markers was
emphasized, especially the advantages of the GFP
over the lacZ marker.
249
vival values of embryos microinjected with the
plasmid pCMV-CAT and an uninjected control
group at different times (Figure 2). The survival
rates, expressed as the means of the percentages of
living embryos, showed a general decreasing tendency with respect to the increasing development
time. The survival values of the two groups were
compared by the statistical test for differences between two regression coefficients (Sokal and Rohlf,
1996) and found to be significantly different (F =
4.58, p < .01). The estimated average reduction in
viability in the transgenics relative to the controls
was around 23%.
DNA persistence
The fates of both linear and supercoiled forms of
the three plasmids, pCMV-CAT, pCMVTklacZ, and
pEGFP-N1 (Figure 1), were followed by Southern
blotting. The results from the respective supercoiled forms are exemplified by the pattern obtained from the pEGFP-N1 plasmid (Figure 3). The
signal intensities, representative of the amounts of
pEGFP-N1, sharply decreased after 12 hours of development, while pCMVTklacZ and pCMV-CAT
decreased after 24 hours (data not shown). In lanes
corresponding to samples with undigested DNA
(odd lanes from 5 to 15), bands of different mobilities are shown, indicating that the plasmid is present in various configurations, such as supercoiled
(z), linearized (y), open circular (x), and higher molecular weight fractions (w and v). The interpretation of the monomeric conformations was based on
a Southern blot using DNA under the same condi-
Results
The seabream response to the transfer of pCMVCAT, pCMVTklacZ, and pEGFP-N1 plasmids (Figure 1) was assessed by analyzing embryo and larval
survival, DNA persistence, and transgene expression, in early development.
Survival analysis
In order to minimize the negative effect of the injected DNA on embryo viability (Fletcher and Davies, 1991), different DNA concentrations were
evaluated, (results not shown). Results from these
experiments indicated that the dose of 108 plasmid
copies (100 ng/µl) per embryo produced the highest
levels of expression without adversely affecting
survival. The effect of the microinjection technique
on mortality was evaluated by comparing the sur-
Figure 2. Survival rates of seabream embryos and larvae
over the first 3 days of development. Open circles represent controls; closed circles, microinjected embryos. The
standard error is represented by the T bars.
250
S. Garcı́a-Pozo et al.
Figure 3. Southern blot patterns of DNA from seabream
embryos microinjected with the plasmid pEGFP-N1 in
circular form: lane 1 represents a positive control sample
containing 5 pg of BamHI-digested plasmid. The band y
corresponds to the linearized plasmid, and the band x to
the undigested plasmid. Lane 2, positive control sample
containing 5 pg of undigested plasmid. The two bands
indicate supercoiled (z) and open circular (x) DNA. Lane
3, negative control sample. Lanes 4 to 15 represent DNA
samples from embryos from times indicated on the top.
Even-numbered lanes represent BamHI-digested DNA,
and odd lanes represent undigested DNA. The amount of
DNA loaded in each lane corresponds to 10 individuals
of the indicated development time. The different bands
are interpreted in the text.
tions as those prior to injection (data not shown).
The bands from the odd lanes indicate that the
DNA was present in the transgenic embryos in either monomeric or multimeric forms. In the even
lanes (4–14), where the DNA was linearized with
BamHI (Figure 1c), a strong band of 4.7 kb was
present. The conversion of the high molecular
weight bands (w and v) into unit size bands (y) after
digestion (even lanes) is indicative of either headto-tail multimers or concatenates. The patterns obtained from the samples with the pCMVTklacZ and
pCMV-CAT plasmids (data not presented) were
similar to those described for pEGFP-N1 (Figure 3)
in terms of the presence of monomeric and multimeric configurations of the transgenic DNA.
The results from the linearized forms of the three
plasmids are exemplified by the Southern blot pattern obtained from the pCMVTklacZ plasmid (Figure 4). Upon the injection of its XbaI linearized
form, the Southern blotting showed in lanes with
undigested DNA (5, 7, 9, 11, and 13), bands of approximately 16 kb (u), as well as faint bands of
higher molecular weight (t). In lanes 4, 6, 8, 10, and
12, in which the transgene was digested with KpnI
(Figure 1b), the presence of two bands of approximately 4.2 kb (x) and 3.8 kb (y), indicates that
bands from odd lanes correspond to head-to-tail
multimeric molecules. Two other bands of about 2
kb (z) and 5.6 kb (w), in lanes 4 and 6, suggest the
presence of a smaller amount of polymeric plasmids in either head-to-head or tail-to-tail conformations.
The patterns obtained from the linearized forms
of the pCMV-CAT and pEGFP-N1 plasmids (Figures 1a and 1c) were very similar in terms of signal
intensity decreasing after 12 to 24 hours and the
presence of only multimeric configurations (data
not presented).
Reporter gene expression
The expression efficiencies of the three constructs
were assessed through the proportions of positive
individuals from various developmental stages, after being assayed for the corresponding reporter
gene. The CAT activity shown by transgenic embryos is documented in Figure 5, in which all
groups screened in the plate were positive. The expression frequencies are summarized in the histogram of Figure 6a. Exogenous CAT expression initiated at 7 hours after fertilization with linearized
and supercoiled plasmids and was maintained
through 4 days of development. To evaluate statis-
Figure 4. Southern blot analysis of DNA from seabream
embryos of various developmental stages microinjected
with the pCMVTklacZ plasmid in linearized form. Lane 1
represents positive control sample containing 5 pg of the
KpnI-digested plasmid, producing two bands: x of 4.2 kb
and y of 3.8 kb. Lane 2 is a positive control sample containing 5 pg of the XbaI linearized plasmid, producing
one band of about 8 kb (v). Lane 3, negative control
sample. Lanes 4 to 13 represent DNA samples from embryos from times indicated on the top. There are two
lanes for each sample, lanes with even numbers correspond with digested (KpnI) DNA, and lanes with odd
numbers represent undigested DNA. The amount of DNA
loaded in each lane corresponds to 10 individuals of the
indicated development time. The different bands are interpreted in the text.
Behavior of exogenous DNA in Sparus aurata
Figure 5. CAT assay plate. Lane 1 is the positive control, lane 2 is a negative control, and lanes 6 and 10 are
uninjected controls. Lanes 3–5, 7–9, and 11–12 correspond to samples derived from pCMV-CAT-injected embryos of 8 hours development. All samples are positive
and show variable CAT intensities. Only monoacetyl derivatives have been detected.
tically the influence of plasmid conformation and
time of development on the expression levels, a
two-way analysis of variance (ANOVA) was performed, revealing a significant effect as a function
of the plasmid forms (F = 3.40, p < .005) as well as
the time of development (F = 8.02, p < .005).
Evidence of ␤-galactosidase activity by immunohistochemical assays could be detected in microinjected embryos but never in controls, indicating
that the activities were exclusively of exogenous
origin. The results of the temporal expression of the
pCMVTklacZ plasmid are shown in Figure 5b. The
blue patches started to appear at 7 hours, and the
proportion of positive individuals gradually decreased to the lowest values at 72 hours. No expression was detected beyond this time. According to
the ANOVA test, the expression activity was influenced by time of development (F = 3.58, p < .005),
but not by the plasmid conformation (F = 9.18, p >
.005). The spatial expression patterns (Figures 7a–
7d) were mosaic at all stages with a large variation
in the extent and distribution of the stained areas.
In general the extension of the stained areas never
surpassed 50% of the whole embryo, was most extensive at about 12 hours (Figure 7a), and progressively decreased thereafter. No tissue-specific patterns of expression were seen (Figure 7b), as would
be expected owing to the ubiquity of the promoter
used, so that the expression areas involved both
extraembryonic and embryonic tissues (Figure 7b).
However, expression in ventral tissues surrounding
the resorbing yolk sac was relatively common (Figure 7c).
251
The temporal expression patterns from the supercoiled pEGFP-N1 plasmid (Figure 6c) revealed
fluorescent cells in about 50% of the embryos at
around 7 hours after fertilization, shortly after the
mid-blastula transition. At 24 hours the level of
positive reaction sharply decreased, and no fluorescence was detected after 72 hours. The patterns
were very similar to those found with the two forms
of the lacZ plasmid. However, with the linearized
GFP plasmid, expression was not observed earlier
than 12 hours, and the GFP was still visible in 4day-old embryos and even persisted in later stages
(unpublished observations). The ANOVA test revealed a significant effect of developmental time (F
= 4.02, p < .005), but not a significant influence of
plasmid conformation (F = 9.28, p <.005), on the
expression rates. The spatial expression patterns
were similar to those described for lacZ in terms of
mosaicism, ubiquity, and lack of tissue specificity
(Figures 8a and 8d).
Discussion
Effect of microinjection on embryo viability
Because of the reported technical difficulties in microinjecting seabream embryos (Knibb et al., 1994),
the first step of this study consisted in evaluating
the damage produced by the microinjection technique itself, without using any method to prevent
the chorion from hardening. This was accomplished by comparing the survival rates of the microinjected groups with their respective uninjected
controls (Figure 2). The results showed a parallelism between the microinjected groups and the respective controls and a clear tendency of decreasing survival with development time. Previous microinjection attempts in this species had reported
survival rates of around 10% at 48 hours (Cavari et
al., 1993), or around 34% at 24 hours (Knibb et al.,
1994). In the latter study different chemical treatments had to be used to inhibit the chorion from
hardening to ease the microinjection process. Their
reported survival values are clearly lower than
ours, which reach around 50% (Figure 2). Survival
data from other fish are available only from freshwater species, with less hard chorions (Fletcher
and Davies, 1991). In general these data vary considerably between species and experiments, but the
highest values are in all cases very close to ours.
Both kinds of comparisons, within the same species and among other species, indicate that in spite
of the unfavorable natural conditions of seabream
eggs for microinjection, the procedure adopted in
252
S. Garcı́a-Pozo et al.
Figure 6. Temporal expression pattern of the seabream embryos and larvae from different stages, microinjected with the pCMV-CAT plasmid (a), pCMVTklacZ
plasmid (b), and pEGFP-N1 plasmid (c). The number of
individuals per sample ranged from 52 to 71, from 33 to
36, and from 110 to 128, respectively. Development
times are represented along the abscissa and the percentage of positive samples on the ordinate.
this work without using any chorion treatment
proved relatively successful in terms of viability.
DNA persistence
The strategy to inject linearized or supercoiled
plasmids was influenced by the discrepant results
in other fish species in terms of persistence and
expression. The comparison of the Southern blot
patterns obtained from both forms of the plasmids
(Figures 3 and 4) allowed us to evaluate the fates of
each form of DNA. A common observation from all
of the plasmids in the supercoiled and linear con-
figurations was a decrease in the signal intensity in
samples after 24 to 48 hours, which can be interpreted as an overall reduction in the total amount
of the foreign DNA remaining in the embryos. Similar patterns of degradation have been reported in
other fish species (Winkler et al., 1991; Volckaert et
al., 1994), presumably owing to random nuclease
action, occurring at a higher rate than the replication process. This is similar to what happens in
Xenopus (Marini et al., 1988). Although the signals
are very faint or nonexistent beyond 3 days of development, DNA persistence after that time has
Behavior of exogenous DNA in Sparus aurata
been implied from expression activity detected at
13 days (data not shown). To delimit the actual
duration of the extrachromosomal plasmid persistence, further developmental stages must be
screened.
Both monomeric and multimeric forms of the
three plasmids persisted in all stages screened (Figure 4 and data not shown). Digestion of the multimeric forms suggests that they are concatenates in
head-to-tail orientations, as seen in other fish species (e.g., Chong and Vielkind, 1989). However,
when linear plasmids were used (Figure 4b), no
monomeric forms were observed. Rather, all of the
exogenous DNA appeared in high molecular weight
conformations in all possible combinations (headto-tail, head-to-head, and tail-to-tail), presumably
as the result of random ligation (reviewed in Iyengar et al., 1996). These results suggest that linear
plasmids were rearranged into concatemers at a
much higher rate than circular plasmids, which
were more often maintained as monomers.
When linearized and supercoiled forms are correlated with the exogenous DNA persistence, our
results reveal no significant differences between
them, as both forms of the plasmid decrease around
the same developmental time (Figures 3 and 4).
This suggests that degradation processes equally affect both forms. Our findings are supported by
other studies in fish, such as Chong and Vielkind
(1989) in medaka and Volckaert et al. (1994) in the
African catfish. However, an enhanced persistence
of the linear versus the supercoiled form has been
found in medaka (Winkler et al., 1991) and rainbow
trout (Iyengar and Maclean, 1995). In the latter, the
authors proposed that this difference is due to a
more efficient concatemerization, though no experimental evidence was presented. These observations are supported by the evidence in Xenopus
that DNA conformations from linear plasmids are
more efficiently replicated and persist longer than
those derived from circular molecules (e.g., Endean
and Smithies, 1989; Fu et al., 1989). In contrast, we
found a higher efficiency for concatemer formation
with linear plasmids, but contrary to the experience of others, the persistence was not appreciably
different. This lack of agreement could be due to
other variables that influence DNA persistence,
such as the sequences used in the constructs
(Asano and Shiokawa, 1993), or the number and
size of the plasmid molecules (Winkler et al., 1991).
In our experiments persistence was not only similar when comparing linear versus circular forms,
but also for the three plasmids, which have differ-
253
ent sequences and sizes (Figure 1). Accordingly, we
suggest that developmental events are the main factor influencing plasmid persistence.
Although the main objective of this work was
not to measure the integration of the transgenic
constructs, the Southern blot patterns in Figure 3
suggested that all high molecular weight bands
were multimers rather than integrants. No integration could be inferred from the blotting data, which
is not improbable if one takes into account the low
integration rate expected in fish (Rahman et al.,
1997) and the limitations of this technique for reliable detection.
Reporter gene expression
The temporal patterns of both forms of the three
plasmids (Figures 6a–6c) reveal that expression is
initiated at about 7 hours after fertilization, which
corresponds approximately with the late-blastula
stage in seabream (Alessio and Gandolfi, 1975).
This event has been associated in zebrafish (Kane
and Kimmel, 1993) and in medaka (Tsai et al.,
1995) with the activation of endogenous genes,
which almost invariably occurs after the midblastula transition (MBT), irrespective of the promoter used. The 5-hour delay in initiation of expression from the linearized GFP plasmid relative
to the circularized form may be due to differences
in time needed to acquire sufficient product for detection of fluorescence (Davis et al., 1995).
Once the exogenous DNA expression was initiated, in the case of the CAT plasmid (Figure 6a) it
continued over all the stages surveyed, and the expression levels were higher with the linearized
form. However, the two forms of the lacZ (Figure
6b) and the GFP (Figure 6c) plasmids showed progressive declines in their apparently similar expression levels after 10 to 12 hours. This decrease
paralleled their respective Southern blot signal intensities (Figures 3 and 4 and data not shown), as
well as with the extent of the stained regions during
development (Figure 6). The apparent lack of correlation between expression levels (Figure 6a) and
persistence in the CAT plasmid (data not shown),
as well as the apparently significant differences in
expression between the two forms, have to be interpreted cautiously, taking into account that each
sample corresponds to five individuals and that
statistical tests are not very reliable under these
conditions. Accordingly, we conclude that our results do not strongly support a marked influence of
the plasmid configuration either in plasmid persistence or in transient expression level, in agreement
Behavior of exogenous DNA in Sparus aurata
with the observations of Chong and Vielkind (1989)
in medaka, Volckaert et al. (1994) in the African
catfish, and Collas et al. (1996) in zebrafish, but
opposite to the findings of Iyengar and Maclean
(1995) in rainbow trout.
The three markers used in this study can be used
in S. aurata if put behind the highly efficient CMV
and Tk viral promoters, as has been done in other
fish species, both in vitro (Hernandez-Bétancourt et
al., 1993) and in vivo (Tewari et al., 1992). Nevertheless, the high percentage of samples showing
CAT activity during development times when DNA
signals were very faint suggests a higher degree of
sensitivity of CAT compared with lacZ and GFP,
even though CAT and GFP transcripts were transcribed from the same promoter. The two spatial
markers (lacZ and GFP) allowed us to locate the
transgene-expressing cells, demonstrated the inherent mosaic patterns of expression, the lack of
cell type specificity of the regulating sequences, as
well as the high transgene activity in extraembryonic tissues (Figure 7), including the yolk syncitial
layer (YSL). This last observation supports the findings of Williams et al. (1996) of higher levels of
expression activity in these tissues. According to
these authors, this aspect should be taken into consideration in quantifying expression levels in embryos with yolk sac that contain extraembryonic
tissues.
While CAT and lacZ have been widely used as
markers in fish, the recently described GFP (Chalfie
et al., 1994) has been tested only in zebrafish as a
convenient live staining marker (Amsterdam et al.,
1995; Moss et al., 1996). We found that when the
GFP gene was injected into S. aurata embryos,
green fluorescence was easily detectable and appeared to have no adverse effects on embryonic development despite widespread expression (data not
shown). The utility of the GFP in this species to
analyze the regulation of tissue-specific expression
255
is enhanced by the fact that seabream embryos are
nearly transparent during the first weeks of development and by day 3 the hatched larvae have almost completed their morphogenesis and rudimentary organs are easily observable. This advantage
can also be useful in production of transgenic lines
as an indicator of transgene expression in early embryos, thus allowing positive selection with higher
efficiency and lower costs. The primary advantage
of GFP over lacZ is that gene expression can be
continuously followed in living embryos, while
fixed embryos are lost to further analysis. Furthermore, the visualization of ␤-galactosidase activity
by the addition of chemical components in living
cells can be disturbing. However, the relative degrees of sensitivity of detection of these markers
cannot be determined from this study because their
promoters were different. However, a superior sensitivity of GFP over lacZ when driven by the same
promoters has been demonstrated in mice (Chiocchetti et al., 1997).
In this work we demonstrate the utility of this
marine species as a suitable system to be manipulated by DNA transfer and show parameters that
affect the persistence and transient expression of
transgenes in vivo. This information will be useful
for assessing the function of regulatory sequences
that have already been cloned from S. aurata.
Experimental Procedures
Maintenance of broodstock and egg manipulation
The broodstock from a private company, located in
the South coast of Spain, has been bred for several
generations and is able to spawn through almost
the complete year under artificial photoperiods.
Naturally fertilized eggs collected from the reproducer tanks were aligned into depressions shaped
with 0.9-mm capillaries in a 2% agarose disk
stained with methylene blue. Embryos at the one-
<
Figure 7. Expression of the exogenous lacZ gene in S. aurata embryos and larvae at different development stages: (a) blastula,
10 hours postfertilization; (b) the embryonic body formation, 32 hours; (c) the posthatching larvae, 48 hours, showing spots of
␤-galactosidase activity only in extraembryonic body areas (surrounding and yolk extension tissues, arrows); (d) middle region
of a 3-day larva presenting a large blue area in the floor plate and a small one in the roof plate of the notochord. The anal duct
is clearly visible. Magnification 80×.
Figure 8. Expression of the GFP gene in S. aurata embryos and larvae. (a and b) Anterior tip of an embryo at the
embryonic body formation stage, 32 hours postfertilization. Under fluorescence conditions (a), the GFP is detected both
inside and outside the embryo. The corresponding bright-field photograph is shown in b. The middle region of a
2-day-old larva is presented under fluorescence (c) and under bright field (d). In c, conspicuous green spots are seen in
the dorsal fin primordium, while very faint ones are seen in the notochord (arrowheads), and a very small one in the
resorbing yolk sac. A close correspondence between yellow areas (a and c) and natural pigmentation (b and d) is clearly
visible. Magnification 90×.
256
S. Garcı́a-Pozo et al.
to two-cell stage were microinjected in the cytoplasm with 8 nl of TE buffer (10 mM Tris-HCl, 1
mM EDTA, pH 8) colored with 0.2% red phenol,
containing approximately 108 plasmid copies (100
ng/µl). No specific treatment to weaken or to remove the chorion was applied, and eggs were kept
in marine water throughout the process. Microinjection was performed with borosilicate needles
and a mechanical micromanipulator. Because the
one-cell stage is clearly visible only 15 minutes
prior to the first cell division, most microinjections
were performed at the two-cell stage, which is visible over a longer duration. To slow down the
cleavage rate, the incubating water temperature
prior to injection was reduced to 14°C. Subsequent
to injection, the treated and the control embryos
were incubated in slow-flowing water tanks at 18°C
until they were collected for further analysis.
The gene constructs
The plasmids used were 5-kb pCMV-CAT (Figure
1a) containing the enhancer and promoter of the
immediate early gene of human cytomegalovirus
(CMV), fused to the chloramphenicol acetyl transferase (CAT) gene, along with the early SV40 splice
and polyadenylation signal (Foecking and Hofstetter, 1986); the 8.05-kb pCMVTklacZ (Figure 1b)
containing the CMV enhancer and herpes simplex
thymidine kinase (Tk) promoter, fused to the E. coli
lacZ gene and SV40 polyadenylation signal (Winkler et al., 1994); the 4.7-kb pEGFP-N1 plasmid
from Clontech (Figure 1c) containing the CMV promoter and SV40 polyadenylation signal. To obtain
the corresponding linearized forms, the pCMVCAT and pCMVTklacZ plasmids were digested
with XbaI, while the pEGFP-N1 plasmid was digested with ClaI.
Extraction and analysis of DNA
For DNA screening, embryos and larvae from the
same state were homogenized in 55 µl of 250 mM
Tris-HCl, pH 7.5, and subjected to three cycles of
freezing-thawing (by freezing in liquid nitrogen
and thawing at 37°C for 5 minutes each) to lyse the
cells. After centrifugation for 20 minutes at 4°C, the
pellets were kept at −80°C. In samples microinjected with the pCMV-CAT, the supernatant was
used for CAT assays.
For Southern analysis the DNA from 20 individuals was pooled together. For DNA extraction,
the reserved homogenate was incubated with 150
µl of the extraction buffer (10 mM Tris, pH 8, 100
mM EDTA, 0.5% sodium dodecyl sulfate [SDS],
and 200 µg/ml of proteinase K). Upon phenol and
chloroform extraction, the DNA was precipitated
by addition of 5 M NaCl and 2× volume of ethanol,
washed in 70% ethanol, and resuspended in TE
buffer. Subsequently half of the DNA was restricted
with the appropriate enzyme, and the other half
remained undigested.
DNA fragments were separated by electrophoresis on 0.8% agarose gels, stained with ethidium
bromide, denatured, and neutralized after being
treated for 10 minutes with 0.25 M HCl and then
transferred to a nylon membrane (Dupont, Boeringher Mannheim). The complete plasmid was used as
probe and 32P-labeled with Klenow polymerase using random primers (Amersham, Megaprime Kit).
Filters were prehybridized in 25 ml of hybridization buffer (0.5 M phosphate buffer, pH 7.2, 7%
SDS, and 1 mM EDTA) at 65°C for 5 hours. Hybridization occurred in the same solution after adding
the probe at 65°C overnight (18 hours). Filters were
washed two times in 2% SDS/2× SSC at room temperature and two more times in 0.1% SDS/0.5×
SSC at 65°C. After that they were exposed to radiographic film (X-Omat Kodak) in a cassette at −80°C
for 1 day to several days.
CAT assay
Groups of five embryos or hatchlings were pooled
in each sample for CAT activity screening. The supernatant obtained from the centrifugation was
heated at 55°C for 10 minutes. The protein extract
was incubated for 2 hours at 37°C with 8 µl 14Clabeled chloramphenicol (Amersham; 25 µCi/ml)
and 20 µl of 4 mM aqueous solution of acetyl coenzyme A (Sigma). The 150 µl of mix was completed with 250 mM Tris-HCl, pH 7.5. After extraction with ethyl acetate, the reaction products were
dried down in a Speed-Vac, redissolved in ethyl
acetate, spotted on a silica gel thin layer plate, and
run with chloroform-methanol mixture (95:5) for
50 minutes. The plates were air-dried, and the autoradiogram was developed after 4 to 5 days (Gorman et al., 1982).
␤-Galactosidase assay
For LacZ expression, undechorionated embryos
were fixed for 30 minutes at 4°C in 2% paraformaldehyde, 0.2% glutaraldehyde, and 0.02% nonidet P-40, in phosphate-buffered saline (PBS). The
embryos were stained for at least 16 hours at 30°C,
in a solution containing a mix of 974 ml of the
Behavior of exogenous DNA in Sparus aurata
solution A and 26 ml of solution B. Solution A: 0.2
M Na-phosphate buffer, pH 7.3, 5 M NaCl, 1 M
MgCl2, 0.1 M K4(Fe3(CN)6), and 0.1 M K3(Fe2(CN)6).
Solution B: 8% X-gal (5-bromo-4-chloro-3-indolyl␤-D galactosidase) in DMF (dimetil formamide).
Stained embryos were kept in 1× PBS.
GFP observation
The EGFP protein is a red-shifted mutated variant
that is more fluorescent than the wild form. To be
visualized, live embryos and larvae were placed on
a glass dish with seawater and observed under an
Olympus 1 × 50 SBF inverted microscope equipped
with an IX-FLA attachment. One inconvenience encountered in observing the GFP was autofluorescence produced by melanophores. Though these
fish are initially transparent over the first 40 days,
they start developing abundant melanophores soon
after the embryoid body is formed, which extends
along the long axis and interferes with observation
of the GFP. To distinguish the green signal of GFP
from the yellow autofluorescence (Figures 8a and
8c), we had to optimize the use of the fluorescein
filter set (BP 460-490 and BA 515-550). Owing to
the larval mobility, the bright-field and fluorescent
photographs were taken on high-speed film.
Acknowledgments
The authors acknowledge the Co. Cupimar, S.A.
(San Fernando, Cádiz), for providing the eggs and
the facilities. Special thanks to E.R. Bejarano for
technical advice.
The plasmids pCMV-CAT and pCMVTklacZ
were kindly supplied by D. Chourrout and M.
Schartl, respectively. This work has been supported by grant BIO2-CT93-0430 from the European Union.
References
Alessio, G., and Gandolfi, G. (1975). Riproduzione artificiale
di orata, Sparus aurata (L.) (Osteichthyes, Sparidae).
Memorie dell’Istituto Lombardo–Accademia de Scienze e
Lettere 26:Memoria 3.
Amsterdam, A., Lin, S., and Hopkins, N. The Aquorea victoria green fluorescent protein can be use as a reporter in
live zebrafish embryos. Dev Biol 171:123–129.
Asano, M., and Shiokawa, K. (1993). Behavior of exogenously-introduced DNAs in early embryos of Xenopus
laevis. Zool Sci 10:197–222.
Cavari, B., Funkestein, B., Chen, T.T., González-Villaseñor,
L.I., and Schartl, M. (1993). Effect of growth hormone on
the growth rate of the gilthead seabream (Sparus aurata),
257
and use of different constructs for the production of transgenic fish. Aquaculture 111:189–197.
Chalfie, M., Tu, Y., Euskirchen, G., Ward, W., and Prasher, D.
(1994). Green fluorescent protein as a marker for gene
expression. Science 263:802–805.
Chiocchetti, A., Tolosano, E., Hirsch, E., Silengo, L., and Altruda, F. (1997). Green fluorescent protein as a reporter of
gene expression in transgenic mice. Biochim Biophys
Acta 1352:193–202.
Chong, S.S.C., and Vielkind, J.R. (1989). Expression and fate
of CAT reporter gene microinjected into fertilized medaka
(Oryzias latipes) eggs in the form of plasmid DNA, recombinant phage particles and its DNA. Theor Appl Genet
78:369–380.
Collas, P., Husebye, H., and Alestrom, P. (1996). The nuclear
localization sequence of the SV40 T antigen promotes
transgene uptake and expression in zebrafish embryo nuclei. Transgen Res 5:451–458.
Davis, Y., Girdham, Ch.H., and O’Farrel, P.H. (1995). A
nuclear GFP that marks nuclei in living Drosophila embryos; maternal supply overcomes a delay in the appearance of zygotic fluorescence. Dev Biol 170:726–729.
Devlin, R.H., Yesaki, T.Y., Biagi, C.A., Donalson, E.M., Swanson, P., Chan, W. (1994). Extraordinary salmon growth.
Nature 371:209–210.
Endean, D., and Smithies, O. (1989). Replication of plasmid
DNA in fertilized Xenopus eggs is sensitive to both the
topology and size of the injected template. Chromosoma
97:307–314.
Fletcher, G.L., and Davies, P.L. (1991). Transgenic fish for
aquaculture. In: Setlow, J.K. (ed.). Genetic Engineering.
New York: Plenum Press, 331–370.
Foecking, M.K., and Hofstetter, H. (1986). Powerful and versatile enhancer-promoter unit for mammalian expression
vectors. Gene 45:101–105.
Fu, T., Hosokawa, K., and Shiokawa, K. (1989). Expression of
circular and linearized bacterial chloramphenicol acetyltransferase genes with or without viral promoters after
injection into fertilized eggs, unfertilized eggs and oocytes of Xenopus laevis. Roux’s Arch Dev Biol 198:148–
156.
Gong, Z., and Hew, C.L. (1993). Promoter analysis of fish
antifreeze protein genes. In: Hoochachka, P.W., and
Mommsen, T.M. (eds.). Biochemistry and Molecular Biology of Fishes, vol. 2. New York: Elsevier, 307–324.
Gordin, H. (1989). Coastal desert mariculture: an approach to
development. In: Abel, R., and Haldey, S. (eds.). Proceedings of the International Symposium on Coastal Ocean
Space Utilization. New York: Elsevier Science.
Gorman, G.M., Moffat, L., and Howard, B. (1982). Recombinant genomes which express chloramphenicol acetyl
transferase in mammalian cells. Mol Cell Biol 2:1044–
1051.
Hernandez Bétancourt, O., Attal, J., Théron, M.C., Puissant,
C., and Houdebine, L.M. (1993). Efficiency of introns from
various origins in fish cells. Mol Mar Biol Biotechnol 2:
181–188.
Iyengar, A., and Maclean, N. (1995). Transgene concatemerization and expression in rainbow trout (Oncorhyncus
mykis). Mol Mar Biol Biotechnol 4:248–254.
Iyengar, A., Müller, F., and Maclean, N. (1996). Regulation
258
S. Garcı́a-Pozo et al.
and expression of transgenes in fish—a review. Transgen
Res 5:147–166.
Kane, D.A., and Kimmel, C.B. (1993). The zebrafish midblastula transition. Development 119:447–456.
Knibb, W.R., Elizur, A., Moaz, B., and Robins, A. (1994).
Inhibition of egg hardening in the marine teleost, Sparus
aurata. Mol Mar Biol Biotechnol 3:23–29.
Maclean, N., and Rahman, A. (1994). Transgenic fish. In:
Maclean, N. (ed.) Animals with Novel Genes. Cambridge,
U.K.: Cambridge University Press, 63–105.
Marini, N.J., Etkin, L.D., and Benbow, R.M. (1988). Persistence and replication of plasmid DNA microinjected into
early embryos of Xenopus laevis. Dev Biol 127:421–434.
Moss, J.B., Price, A.L., Raz, E., Driever, W., and Rosenthal, N.
(1996). Green fluorescent protein marks skeletal muscle
in murine cell lines and zebrafish. Gene 173:89–98.
Rahman, M.A., Iyengar, A., Maclean, N. (1997). Coinjection
strategy improves integration efficiency of a growth hormone gene construct, resulting in lines of transgenic tilapia (Oreochromis niloticus) expressing an exogenous
growth hormone gene. Transgen Res 6:369–378.
Reinhard, E., Nedivi, E., Wegner, J., Skene, J.H.P., and Westerfield, M. (1994). Neural selective activation and temporal regulation of a mammalian GAP-43 promoter in zebrafish. Development 120:1767–1775.
Sokal, R.R., and Rohlf, F.J. (1986). Introducción a la bioestadı́stica. Barcelona: Ed. Reverté.
Takagi, S., Sasado, T., Tamiya, G., Ozato, K., Wakamatsu, Y.,
Takeshita, A., and Kimura, M. (1994). An efficient expres-
sion vector for transgenic medaka construction. Mol Mar
Biol Biotechnol 3:192–199.
Tewari, R., Michard-Vanhee, C., Perrot, E., and Chourrout, D.
(1992). Mendelian transmission, structure and expression
of transgenes following their injection into the cytoplasm
of trout eggs. Transgen Res 1:250–260.
Tsai, H.J., Wang, S.H., Inoue, K., Takagi, S., Kimura, M.,
Wakamatsu, Y., and Ozato, K. (1995). Initiation of the of
the transgenic lac-Z gene expression in medaka (Oryzias
latipes) embryos. Mol Mar Biol Biotechnol 4:1–9.
Volckaert, F.A., Hellemans, B.A., Galbusera, P., Sekkali, B.,
Belayew, A., and Ollevier, F. (1994). Replication, expression, and fate of foreign DNA during embryonic and larval
development of the African catfish (Clarias gariepinus).
Mol Mar Biol Biotechnol 3:57–69.
Williams, D.W., Muller, F., Lavender, F.L., Orbán, L., and
Maclean, N. (1996). High transgene activity in the yolk
syncytial layer affects quantitative transient expression
assays in zebrafish (Danio rerio) embryos. Transgen Res
5:433–442.
Winkler, C., Vielkind, J.R., and Schartl, M. (1991). Transient
expression of foreign DNA during embryonic and larval
development of the medaka fish (Oryzias latipes). Mol
Gen Genet 226:129–140.
Winkler, C., Wittbrodt, J., Lammers, R., Ullrich, A., and
Schartl, M. (1994). Ligand-dependent tumor induction in
medakafish embryos by a Xmrk receptor tyrosine kinase
transgen. Oncogene 9:1517–1525.