- Email: [email protected]
A novel gene expressed during zebrafish gastrulation identified by differential RNA display
ELSEVIER
Mechanisms of Development52 (1995) 383-391
A novel gene expressed during zebrafish gastrulation identified by differential RNA display Greg Conway* Department of Molecular and Cellular Biology, Harvard University, Cambridge, MA 02138, USA
Received 4 April 1995; acceptedMay 22 1995
Abstract Vertebrate gastrulation is a dynamic period of development characterized by extensive cell migrations. This stage of development is likely to require the expression of a new genetic repertoire to initiate and direct these dramatic changes. The differential RNA display has been used to identify genes specifically expressed during the gastrula stage of a model vertebrate, the zebrafish. One of the genes isolated by the differential display technique has been sequenced and characterized for its spatial and temporal expression. This gene, called G12, is expressed during a narrow window of time during gastrulation and is restricted to a single cell type. At this time of development the zebrafish embryo consists of three cell types: the yolk cell, EVL cells and deep cells. Interestingly, both EVL and deep cells derive early in development from common progenitor cells but G12 expression is restricted only to the EVL lineage. Comparison of the amino acid sequence from this gene with the Genbank database indicates similarity to two previously reported mammalian genes. The similarity between these three genes suggests that they may serve a common function. The Cl2 gene is the first example of restricted gene expression in EVL cells of the zebrafish. The G12 gene should prove to be a useful model for the study of regulated gene expression during gastrulation. Keywords:
Zebrafish; Gastrulation; G12 Expression; Differential RNA display
1. Introduction Vertebrate gastrulation is characterized by extensive cell migrations which transform a roughly spherical ball of cells into an elongated structure possessing the major body axis. Expression of many new genes is probably required during this dynamic period of development. The differential RNA display technique has been used to isolate genes which are differentially expressed between blastula and gastrula stage embryos in a model vertebrate, the zebrafish. This report describes the characterization of one of these genes, called G12. This gene is expressed in one of the first three cell types of the zebrafish embryo and expression of G12 is coincident with many changes in the morphology of the cells in which it is expressed. Although the first three cell types of the zebrafish can be described at the morphological level, little is know about gene expression or the molecular changes which * Tel.:
+I 617 4950560;
Fax: +1 617 4964313;
Email: [email protected].
0925-4773/95/$09.50 SSDI
occur in them during gastrulation. The zebrafish embryo begins as a single cell with a large centrally located yolk. Upon fertilization, cytoplasmic streaming produces a bulge of cytoplasm, the blastodisk, at the site of sperm entry. The blastodisk undergoes cleavage, and by the fourth division produces a 4 by 4 array of cells with the outer cells of this array connected by cytoplasmic bridges. These cells will continue to undergo incomplete cleavages and eventually fuze to produce a large syncytium encompassing the yolk. The inner cells of the 4 by 4 array undergo complete cleavage and give rise to two types of cells. Superficial descendents of these cells will produce a single-cell-thick covering which will surround the embryo, called the enveloping layer (EVL), while the inner descendents, called deep cells, will give rise to ectoderm, endoderm and mesoderm, producing the embryo proper. As development proceeds, the deep cells sitting on top of the large yolk cell, produce a ball of cells covered by the EVL. This period of development is the blastula stage. At the onset of gastrulation, changes in the yolk cell, and migrations of EVL cells and deep cells, transform the
0 1995 Elsevier Science Ireland Ltd. All rights reserved
0925-4773(95)00416-X
384
G. Conway I Mechunisms of Development 52 (1995) 383-391
embryo into an elongated structure possessing the body axes of the adult. An understanding of this dynamic period of development requires a knowledge of the genes specifically activated during this time. The differential RNA display has been used to compare the mRNA population of blastula and early gastrula embryos. This technique involves the PCR amplification of first strand cDNA using random sequence oligonucleotides (Liang and Pardee, 1992; Welch et. al., 1992; Liang et al., 1993). Under these conditions a random but reproducible subset of cDNAs are amplified. Using the conditions described in this paper and employed by others, approximately 50-100 cDNAs are amplified per primer set. If this amplification procedure is employed on different RNA samples, the 50-100 amplified cDNAs can be compared in the two RNA populations by polyacrylamide gel electrophoresis. Using only 240 oligonucleotide primer sets it is theoretically possible to sample the entire mRNA population of a cell, roughly 15 000 different RNA species (Liang and Pardee, 1992). Using this technique, we have discovered several genes differentially expressed during blastula or gastrula stages. The temporal and spatial expression for one of these genes has been analyzed and it was found that it is expressed during gastrulation in EVL cells. G12 should prove to be a useful system for the analysis of temporal and spatial gene regulation during gastrulation. G12 is similar to genes found in other organisms suggesting an evolutionarily conserved function for these proteins. 2. Results 2.1. Differential RNA display of blastula and gastrula RNA
RNA from blastula (4 h of development) and gastrula (6 h of development) zebrafish embryos was compared using the differential RNA display. All developmental times in this paper indicate times postfertilization at 28°C. A region of the gel displaying the transcripts amplified by one set of oligonucleotide primers is shown in Fig. 1 (see Section 4 for a description of the olignucleotides). There are several examples of bands which are present in one RNA sample but missing or much less intense in the other sample. Bands which were differentially represented in either the blastula or gastrula samples were excised from the gel, reamplified and cloned. Using combinations of four oligonucleotides, a total of 15 bands were successfully reamplified and subcloned from differential display. Random primed probes were then made from these clones and used to probe northern blots containing blastula and gastrula RNA. Four clones showed stage specific expression which agreed with the results from the differential display. Two of these clones were blastula specific and two gastrula specific. Two clones which were gastrulation specific according to differential display detected an RNA species of equal in-
B
bases 404
G
-
-G12
309
242 238 217 201
-
Fig. 1. Differential RNA display of blastula and gastrula RNA. First strand cDNA was synthesized from blastula (B) and gastrula (G) RNA. The cDNA samples were then PCR amplified in the presence of 32PdATP (see Section 4 for a description of the PCR primers). The PCR samples were then size fractionated by denaturing electophoresis and autoradiographed. Several bands are present in one sample but less intense or missing in the other sample. One gene, called G12 (indicated by an arrow), which is present only in the gastrula sample was isolated from the gel, reamplified and cloned.
tensity in both blastula and gastrula samples and were thus falsely represented by the differential display. Unfortunately 9 of the clones did not give a signal on northern analysis even when 2,ug of poly-A+ RNA was probed. The differential display, utilizing PCR amplification, is likely to amplify RNA species of low abundance which are difficult to detect and analyze further. The four clones which showed stage specific expression were further analyzed by in situ hybridization of whole mount zebrafish embryos. One of these genes, called G12, gave an interesting spatial and temporal expression pattern and was studied further. 2.2. Sequence analysis of G12 A gastrula (6 h) cDNA library was probed with an oligonucleotide homologous to sequence found in the PCR
G. Conway J h4echanisms
1 61
121 181 241 301 361 421 481 541 601 661 721 781 841 901 961 1021 1081 1141 1201
of Devetopment
GAGAMCTAAGCACACTAAAGACACGCATTCAG
52 (I 995) 383-391
385
XUAAACTGCTCTCGAATCAAACCACC
ATATGTGAGAGCTATATTAATGAAGCAGAAACCCTGTGGAGGATGTCTCC
60 120 180 240 300 360 420 480 540 600 660 720 780 840 900 960 1020
TGCTAATGACCTG~G~TCTGTGAATCATCCTGTG;CTCTG
1080
~~AGATGTCCGAGCCCCTCAGCCAGAAGAACGCTCTCTACACGGCAATGAACCGC TTCCTCGGCGCCGTCAACAACATGGACCAGACGGTCATGGAT GTGCCGCTGGACCAGGAGAGAGCAGCAGAAGCTGACCACTACCTG CGGGAGGCTGAGGCCGACATGTACAGCTATTACAGTCAACC ATCGAATGGGGCGTCATCCGCTCGGAGGACCAGCGGCGCAGACACCTCTGCGTCG GAGCCTGTGCGCACAGAGGAGGAGAGCGACATGGATCTCGCTCCTGCAGTTCCAT CTGAAGGGCTTGCACGGGGTGCTGTCCCAACTCAC~GCC~GCC~C~CCTGACC~C CGATACAAGCAGGAGATTGGCATTTCTGGCTGGGGACAdTGATC CACATCTACTGCGATATGATGAACTGTTGTGTTGTTCTAACTGCTCTGCATTTG TGGAAGAGCCAA?iGGGGACTAATGGAGCAGAAGTTACT=CG
AAAAMGACAAACCGTGA
AGGGAAAGTGGATTTGTCGAT~TGGTTTTGGTGTTGAGGG CTGCTTCCTGTTTTGTGCCATTTGCACTGACAGTTTCTCTTTTT~CC~~~TCTC CGTTGAAGTGGCTGACTGCCATGCATTTACATTFLATATTT~GTCTATTTATGTATTA~ TAAAATTGAAAGCTTTTATTTTTCCATTTTTCTTTGTGATAGAGG~GT~TTGTCCTA TGATGTTGCCAAGTGTACCAGAGGACTGATGACTTTTTTTTAT~TT~CTGAGATTGCT
AAAATCACTGTTTGTGTAATGATGAGCCTCCATGAGATTCAGTGTGF~CTGTGCTCATA
1140
AATGCCAGTTGAATTGTTAAAGCATTGCTGTGTTTTTATGTTAATG~CTTTTTTTTTTT
1200
-ATAACAAAATAAACGTTTGTTACTTG
1226
Fig. 2. Nucleotide sequence of G12. Boxed letters indicate the start and stop codons. The cytoplasmic polyadenylation polyadenlyation sequence: AATAAA are underlined. The Genbank accession number for this sequence is U27 121.
fragment of G12. The longest cDNA clone, called pG12 was sequenced and is presented in Fig. 2. Two reading frames contain stop codons throughout the sequence and cannot be the correct reading frame. One reading frame is devoid of stop codons for the first 525 nucleotides of sequence. The first methionine in this reading frame begins at position 64 and is a likely candidate for the start codon. However, upstream from this methionine and in the same reading frame there are no stop codons. This raised the possibility that the cDNA clone, pG12, was a partial cDNA. To determine if there are additional nucleotides missing from pG12, 5’ RACE was performed on gastrula cDNA. From the 5’ RACE analysis, two additional nucleotides were identified and are the first two residues shown in Fig. 2. With the addition of these nucleotides to the sequence derived from pG12, the methionine at position 64 in Fig. 2 is the first initiation codon, indicating that the first 63 nucleotides are 5’ untranslated sequence. The G12 gene encodes a long 3’ untranslated region of 706 nucleotides, over half the length of the transcript. Interestingly this 3’ sequence contains a cytoplasmic polyadenylation sequence: UUUUUAU. This motif is found in several genes and is necessary for developmentally regulated cytoplasmic lengthening of the poly-A tail (Fox et al., 1989; McGrew et al., 1989; Vassalli et al., 1989; Jackson, 1993). In several instances this lengthening of the poly-A tail has been correlated with increased translation of the mRNA (McGrew et al., 1989; Vassalli et al., 1989). The possibility that G12 undergoes cytoplasmic polyadenylation requires further investigation.
sequence:
TTTTTAT
and the
The amino acid sequence of G12 was compared to other sequences in the Genbank database and found to be similar to two previously reported sequences: rat spot 14 and human est00887. The amino acid sequence of G12 is compared to the sequence of spot 14 and human estO0887 in Fig. 3. Amino acids which are identical in G12 and spot 14 or est00887 are shown as bold letters. Both the G12 protein and spot 14 protein are nearly identical in size, roughly 17 kDa. Unfortunately the complete sequence of human est00887 is lacking. Human est00887 is an expressed sequence tag isolated from a human hippocampal cDNA library. Until this clone is fully sequenced it will not be possible to determine if the size of the est00887 protein is similar to that of G12 and spot 14. The greatest degree of similarity between G12, spot 14 and estOO887 lies in the carboxyl region of these proteins. For human est00887, 68% of the last 41 amino acids are identical to G12. Compared to G12, 37% of the last 38 residues of spot 14 are identical. The high degree of similarity in the carboxyl region of these proteins suggests that this region may fulfill a common function in these proteins. In the carboxy terminal sequence of G12 there is a possible leucine zipper characterized by the periodic spacing of four leucine resides every seven amino acids (shown by vertical lines in Fig. 3). 2.3. Developmental expression profile of G12 From northern blot analysis the G 12 gene produces a 1.4 Kb transcript (Fig. 4A). The size of this transcript is in agreement with the 1234 nucleotide long sequence for G12, adding roughly
G. Crmwuy I Mechanisnu ofDevelopment 52 (1995) 383-391
386
1
50
zebrafish G12 human est00887 rat spot 14
MQVLTKRYPKNCLLKVMDRYSAWRNMEQVVMIPSLLRDVEL........
zebrafish G12 human est00887 rat spot 14
51 100 LTNDPGSYLREAEADMYSYYSQLKSIRNNIEWGVIRSEDQRRKKDTSASE PPTGLGXXRAVPGSPKDILVDLG ..SGSGGSVQDGAPDLYTYFTMLKSICVEVDHGLLPREEWQAKVAGNEGS
zebrafish G12 human est00887 rat spot 14
PVRTEEES.......... .DMDLEQLLQFHLKGLHGVLSQLTSQANNLTN HLEGSDAG.......... .EEDLEQXFHYHLRGLHTVLSKLTRKANILTN EAENEAAETEEAEEDRLSEELDLEAQFHLHFSSLHHILTHLTQKAQEVTQ
zebrafish G12 human est00887 rat spot 14
151 163 RYKQEIGISGWGQ RYKQEIGFGNWG KY.QEMTGQV
MQMSEPLSQKNALYTPMNRFLGAVNNMDQTVMVPSLLRDVPLDQEKEQQK
101
150
I
I
I
I
Fig. 3. Amino acid sequence of G12 and comparison with rat spot 14 and human est00887. The amino acid sequence of Gl2, rat spot 14 and human est00887 have been aligned to maximized similarity. Amino acids which arc identical in G12 and spot 14 or human est 00887 are in bold. The position of leucine resides in Cl2 which could theoreticaly form a leucine zipper are indicated by vertical lines.
200 nucleotides of poly-A. The expression of G12 RNA throughout development was analyzed by RT-PCR and compared to the expression of a housekeeping gene which is required for translation, EFl alpha. As shown in Fig. 4B, there is low level expression of G12 throughout development, however, this gene is intensely expressed at 6 h of development and is again at low levels by 8 h. The level of EFl alpha on the other hand is constant from 6 h throughout development. The increased expression of
EFl alpha at 6 h of development dramatically shows the switch to zygotic gene expression. In the zebrafish, zygotic transcription does not begin until the midblastula stage (3 l/3 h of development). This is known as the midblastula transition or MBT (Kane and Kimmel, 1993). Prior to this stage, embryonic development is driven by maternal RNAs placed in the egg by the ovary. Cl2 is activated coincident with EFl alpha, at or shortly after MBT, however, in contrast to EFl alpha G12 expression
A G12
MW
Kb - 8.5 - 7.5
- 2.4
G12 EFl alpha
Fig. 4. Northern blot and RT-PCR analysis of Gl2 expression. (A) A northern blot containing 2yg poly-A+ RNA from gastrula stage (6 h) zebrafish embryos was probed for G12 transcripts. A transcript of 1.4 Kb is detected. (B) Total RNA, isolated from staged embryos or adults, served as template for RT-PCR using oligonucleotide primers complementary to G12 or EFl alpha. PCR products were then size fractionated, transfered to nylon membranes and probed with an oligonucleotide complementary to a region between the PCR primers. Hours of development indicate times postfertilization at 28°C.
G. Conway I Mechanisms
3hrs
6hrs
Fig. 5. labeled probes tion at
qf Development 52 (1995) 383-391
4hrs
387
Shrs
1Zhrs
8hrs
Developmental profile of Cl2 RNA revealed by in situ hybridization, Staged embryos were probed with either sense or antisense digoxigenin RNA probes to G12. The presence of G12 RNA is indicated by blue staining. The orange color of the yolk seen with both sense and antisensc is due to endogenous phosphatase activity and does not indicate the presence of G12 RNA. Hours of development indicate times postfertiliza28°C.
Sense
Antisense
, ,.*w, i b ‘-
,
Fig. 6. In situ localization of G12 RNA using thin sectioned embryos. Sections (5pm thick) were made from 5 h zebra&h embryos and probed with either sense or antisense digoxigenin labeled RNA probes to G12. The presence of Cl2 RNA is indicated by blue staining. The arrow points to EVL cells which show the presence of G12 RNA. The blue color of the yolk seen with both sense and antisense probes is due to endogenous phosphatnsc activity and does not indicate the presence of G12 RNA.
388
dramatically decreases, window for this gene.
G. Conwuy I Mechanisms
producing
2.4. Sites of Cl2 expression
a narrow
of’ Development 52 (1995) 383-391
expression
revealed by in situ
hybridization
To determine the spatial localization of Cl2 expression, in situ hybridization was performed with whole mount zebrafish embryos using anti-sense RNA probes. The in situ hybridizations are shown in Fig. 5; the presence of G12 transcripts is indicated by a blue color. Sense strand probes were used as a control and showed no staining. G12 RNA is first detected at 4 h of development in isolated cells. Staining is most intense at 5 h, begins to disappear at 6 h and random staining of isolated cells is seen at 8 h of development. Often cells which express G12 at 8 h of development are in clumps, although never observed with a pattern to this grouping. From 12 to 72 h of development Cl2 expression is not detected. The temporal pattern of staining revealed by in situ hybridization parallels that seen with RT-PCR. The staining of 5 h embryos shown in Fig. 5 suggests that just the outer surface layer of cells, the EVL, expresses G12. To determine if EVL cells are the only cells expressing G12, in situ hybridization was performed on sectioned embryos (Fig. 6). Embryos at 5 h of development were sectioned and hybridized with sense and antisense probes. Again sense probes did not show staining, however, the antisense probe stained only the EVL cells (shown by an arrow head). The blue staining of the yolk is due to endogenous phosphatase activity and is seen with both sense and antisense probes and does not indicate the presence of G12 RNA. 3. Discussion The beginning of gastrulation is likely to require the expression of new genes. To isolate genes activated during gastrulation we have used the technique of the differential RNA display to compare the mRNA population of blastula and gastrula zebrafish embryos. The differential RNA display is a rapid method for comparing RNA samples from different developmental stages. Using only a few primer combinations, four genes were isolated which are differentially expressed in either blastula or gastrula zebrafish embryos. One of these genes is described in detail in this report. There are some caveats to the display technique which should be noted. Some transcripts which are amplified by the display technique are of low abundance and can not be detected on Northern blot analysis. These genes may be interesting but are difficult to analyze further. Also, some genes which according to the display technique are differentially expressed were later found to be expressed at equal levels in both stages (2 of the 6 genes which gave signals on Northern analysis). The presence of these falsely represented transcripts makes it impossible to es-
timate the general percentage of stage verses nonstagespecific gene expression. Any gene detected to be differentially expressed between 2 RNA samples by differential display should be further checked on Northern blots, or by a similar means such as RT-PCR, to ensure that these genes are actually differentially expressed. The reason for the false distribution sometimes seen with the differential display is unknown. Also it should be noted that the intensity of bands on the differential display does not reflect the abundance of the mRNA in the RNA sample. Some of the most intense bands on the display failed to detect a transcript on northern analysis. The intensity of bands on differential display is dependent not only upon the abundance of that species in the RNA sample but also on the degree of homology between the RNA target and the PCR primers. The gene reported here, G12, has an interesting temporal and spatial expression pattern. G12 mRNA is detected in 4 to 8 h embryos, with peak expression at 5 h of development. This gene is activated at or shortly after MBT, but unlike the housekeeping gene EFl alpha, expression decreases shortly after its rapid induction. In addition to temporal regulation, G12 is also spatially restricted with expression only in the EVL cells, one of the first three cell types of the zebrafish embryo. Interestingly, both EVL and deep cells originate from common progenitor cells, yet expression is restricted to only one of these lineages. The restricted spatial expression of G12 indicates that transcriptional activation of this gene is not the result of general nonspecific activation of the transcriptional apparatus in all cells of the embryo at MBT. Expression of G12 is maximal at the onset of gastrulation and ends shortly before the completion of gastrulation. At the begining of gastrulation EVL cells adopt an epithelial morphology and are required to stretch and cover a greater surface area due to extensive cell movements which occur at this time. It would be interesting to determine if the G12 gene plays a causal role in any of these changes. The origin of EVL cells (an extraembryonic structure) and deep cells (the cells which will give rise to the embryo) is similar to the origin of extraembryonic and embryonic structures in mammals. During mammalian development the early blastula contains cells which can give rise to either embryonic or extraembryonic tissues. As mammalian development proceeds, the outer cells of the embryo (the trophectoderm) give rise to extraembryonic tissues while inner cells (the inner cell mass) produce the embryo proper. In many ways the EVL and trophectoderm are analogous structures; both are organized into an epithelium and originate from superficial cells. Deep cells and the inner cell mass are similar in that both are interiorly located, give rise to embryonic tissues and remain pluripotent for a considerable time in development. The expression of G12 in EVL cells raises the interesting possibility that a similar gene(s) may exist in mammals
G. CrmwuyI Mechanismsof Development52 (1995) 383-391 and be expressed in extraembryonic tissues early in development. A search of the Genbank database indicates that genes similar to G12 do exist in mammals. Two previously reported genes are similar to G12: rat spot 14 and human estO0887. Rat spot 14 was first discovered on 2 dimensional gels as a liver protein whose abundance rapidly increased in response to thyroxin (Liaw and Towle, 1984; Narayan et al., 1990). The gene was subsequently cloned and shown to be regulated by insulin and retinoic acid as well as thyroxin (Jump et al., 1990; MacDougald et al., 1992). Spot 14 has become a model for the study of regulated gene expression in the liver. The promoter region of this gene has been extensively analyzed and several cis-elements necessary for response to glucose and thyroxin determined (Jump et al., 1993; Sudo et al., 1993; Shih and Towle, 1994). It would be interesting to see if G 12 is also regulated by insulin, retinoic acid or thyroxin. Despite an impressive wealth of knowledge regarding the regulation of spot 14, little is know about its function. Antibodies have been produced to spot 14 and immunohistochemistry performed (Kinlaw et al., 1992, 1993). Interestingly, spot 14 is a nuclear protein. Unfortunately the antibodies produced to spot 14 are against peptides which are not conserved in G12 and are of no value in determining the localization of G12 in EVL cells. Currently, antibodies are being produced against zebrafish G12. All that is known about est00887 is that it is expressed in the human hippocampus, since it was derived from a hippocampal cDNA library. It would be interesting to examine the expression of spot 14 and est00887 during development; perhaps they are expressed at an earlier time. The amino acid sequence conservation seen in the carboxy terminal region of all three proteins suggests that this sequence may be a functional domain and fulfill an analogous role in all three. It would be valuable to search for other proteins with a sequence similar to this carboxyl region. G12, spot 14 and human estO0887 may be three members of a much larger gene family. 4. Materials and methods 4. I. RNA isolation All developmental times refer to development at 28°C. Total RNA was isolated from staged zebrafish embryos by the guanidinium isothiocyanate/phenol method using a Stratagene Total RNA Isolation Kit as described by the manufacturer. Poly-A+ RNA was isolated from total RNA using a Fasttrack mRNA isolation Kit from Invitrogen as described by the manufacturer. 4.2. Differential RNA displa) Differential RNA display was performed using a modified protocol of Liang and Pardee (1992) and Welch et al. (1992). First strand cDNA was synthesized from
389
1 ,ug blastula (4 h embryos) or gastrula (6 h embryos) total RNA. Conditions for first strand synthesis were 50 mM Tris (pH 8.3), 75 mM KC], 10 mM DTT, 5 mM MgCl2 and 20pM dNTPs in the presence of 1 pm anchor primer in a total volume of 20~1. Two different anchor primers were separately used to synthesize cDNA. The sequence of these primers were: Anchor1 S-CCGAATTCWCG-3’ and Anchor 2 5’-CCGAATTC-CA-3’. Samples were heated to 65°C for 5 min, followed by 10 min at 37°C. To the reaction 200 units MMLV reverse transcriptase (BRL Inc.) were added. After 50 min at 37”C, the reverse transcriptase was inactivated by incubation at 95°C for 5 min. PCR was then performed using 5 ~1 of the first strand reaction in a total reaction volume of 50~1. The PCR reactions contained 30 mM Tris (pH 8.3), 32.5 mM KCI, 3 mM MgCl*, 50pM dNTPs, 5 units Taq polymerase (Cetus Inc.), 20,& 32P-dATP (NEG-0122, New England Nuclear), 1 PM of an anchor primer and 1 PM of a 5’-primer. Two different 5’-primers were separately used in combination with the two anchor primers. The sequence of the 5’primers were: 5’-primer1 5’-GCACGGATCCGGTAGATAG-3’ and 5’-primer2 5’-GCACGGATCCTGATCCATG-3’. PCR reactions were subjected to one cycle of 94°C for 5 min, 40°C for 5 min, and 72°C for 5 min. Reactions were then subjected to 30 cycles of 1 min 94”C, 1 min 55”C, and 1 min at 72°C. Aliquots (5~1) of the PCR reactions from blastula and gastrula samples were run in adjacent lanes of a 6% denaturing polyacrylamide gel. Following electrophoresis, the gel was wrapped in Saran wrap, autoradiographed and bands present in either the blastula or gastrula lanes excised with a razor blade. The gel slices were soaked for 10 min in water to remove urea. This water was then replaced with 50~1 of water, heated for 2 h at 95°C and 10~1 used for reamplification. Individual eluted bands were amplified by PCR using the reaction conditions described above and were subjected to 30 cycles of 94°C for 1 min, 55°C for 1 min, and 72°C for 1 min. Aliquots (5~1) of the reamplified bands were checked on 6% non-denaturing polyacrylamide gels stained with ethidium bromide to insure that fragments the correct size were geneated during the second round of amplification. The remaining PCR reaction was then cloned into the pCRI1 vector (Invitrogen Inc.) as described by the manufacturer. 4.3. Sequencing of PCR fragments and Northern blot analysis Random primed probes were generated from PCR fragments cloned into the pCRI1 vector using a Random Prime Kit (BRL Inc.) as described by the manufacturer. These probes were used to screen Northern blots containing 10 pug of total RNA and 2,ug of poly-A+ RNA from blastula (4 h) and gastrula (6 h) stage embryos. Northern analysis was performed as described by Sambrook et al. (1989). Clones which gave signals that agreed with the
390
G. Conwuy
I Mechanisms
of Development
restricted and staged expression found from the original differential display were sequenced. Clones were sequenced using double stranded DNA templates and a Sequenase Kit (United States Biochemical) as described by the manufacturer. 4.4. Library screening, sequencing, and 5’ RACE A gastrula (6 h) cDNA library was constructed in the lambda Zap11 vector using a Zap cDNA Library Kit from Stratagene Inc. as described by the manufacturer. This library was probed with an oligonucleotide found in the PCR fragment of one of the genes, called G12, identified by the differential display and confirmed by northern analysis to be expressed during gastrulation but not during the blastula stage. This oligonucleotide had the sequence: S-GACAAATCCACTTTCCCTICACGGTITGTC-3’. Phage were plated and transfered to nylon disks (Biodyne, Pall Inc.) as described in Sambrook et al. (1989). Membranes were probed at 42°C and washed at 55°C using buffers described by Church and Gilbert (1984). Positive clones were plaque purified and converted to pBluescript plasmids (Stratagene, Zap cDNA Kit) as described by the manufacturer. The longest clone, called pG12, was sequenced using a Sequenase Kit as described by the manufacturer. To insure that 5’ sequence was not missing from pG12, 5’ RACE was performed using a 5’-AmpliFINDER RACE Kit (Clonetech Inc.) as described by the manufacturer. Only two 5’ nucleotides were found to be missing from this cDNA clone. 4.5. Analysis of RNA by RT-PCR First strand cDNA was synthesized from 1 pg total RNA as described above. For the developmental analysis of EFl alpha 1~1 of the 20~1 first strand reaction was diluted 1: 150 with water and 1~1 used for PCR. For the analysis of G12, cDNA was diluted 1:30 with water and 1~1 used for PCR. Reaction conditions for PCR were the same as described above, however, the final reaction volume was 50~1 and the reactions cycled 25 times: 95°C for 30 s, 55°C for 30 s and 72°C for 1 min. Oligonucleotide primers for EFl alpha had the sequence: 5’-TGGGCACTCTACTTAAGGAC-3’ and 5’-TGTGGCAACAGGTGCAGTTC-3’. Oligonucleotide primers for analysis of G 12 were: S’-TCATTAGCAGGAGACATCCT-3’ and 5’-AAGTGGCTGACTGCCATGCA-3’. Following PCR, for each developmental stage 2~1 of the 50~1 reactions for EFl alpha and G12 were pooled and subjected to 5% denaturing polyacrylamide gel electrophoresis. Gels were then electrobloted to nylon membranes (Hybond-N, Amersham Inc.). Blots were then baked, UV crosslinked and probed with oligonucleotides to EFl alpha (5’-TGACTGATAGTGCCTCTITC-3’) or Cl2 (5’-GTCATCAGTCCTCTGGTAC-3’) at 45°C and washed at 55°C using buffers and conditions described in Church and Gilbert (1984). Filters were then subjected to autoradiography and bands quantitated using a Fuji Bio-
52 (1995)
383-391
Image Analyzer (Fuji Photo Film Co.. Ltd.). To insure that the intensity of the bands in the RT-PCR analysis reflect the abundance of the analyzed message, RT-PCR analysis was performed in triplicate using 1~1, 3 ~1. and 9 ,ul of cDNA diluted as described above. The intensity of bands was found to reflect the amount of input cDNA and hence is quantitative and within the linear range of PCR amplification. 4.6. In situ hybridization In situ hybridization was performed on whole mount embryos exactly as described by Schulte-Merker et al. (1992). Digoxigenin labeled sense strand RNA probes were transcribed from KpnI linearized pG12 using T3 RNA polymerase. Anti-sense RNA probes were synthesized from EcoRI linearized pG12 using T7 RNA polymerase. In situ hybridization was also performed on 5 micron thick cryostat sections. Embryo treatment, sectioning and hybridizations were performed as described in Westerfield (1993). 4.7. Genbank accession number The accession number for the sequence reported in this paper is U27 12 1. Acknowledgments I thank Jim Wilhelm and other members of the Gilbert laboratory for technical assistance and helpful discussions. Thanks to Doug Melton and members of his laboratory for helpful advise and use of equipment. I am grateful to Barrett Klein and Michiel C. J. Bliemer for help with the figures. Special thanks to Richard Roberts for advice on the manuscript and helpful discussions. G. C. was supported by an American Cancer Society postdoctoral fellowship and was a Medical Foundation/Charles A. King Trust Research Fellow. This work was supported by a grant from the National Insitutes of Health to Walter Gilbert (EY 08397). References Church G. and Gilbert, W. (1984) Proc. Natl. Acad. Sci. USA 81, 1991-199s. Fox, CA., Sheets, M.D. and Wickens, M.P. (1989) Genes Dev. 3. 2151-2162. Jackson, R.J. (1993) Cell 74,9-14. Jump, D.B., Bell, A. and Santiago, V. (1990) J. Biol. Chem. 265, 34743478. Jump, D.B., Clarke, S.D., MacDougaJd, 0. and Thelen, A. (1993) Proc. Natl. Acad. Sci. USA 90, 8454-8458. Kane, D.A. and Kimmel, C.B. (1993) Development 119,447-456. Kinlaw, W.B., Tron, P. and Friedmann, AS. (1992) Endocrinology 131,3120-3122. Kinlaw, W.B., Tron, P. and Witters, L.A. (1993) Endocrinology 133, 645-650. Liang. P. and Pardee. A.B. ( 1992) Science 257.967-97 1. Liang, P., Averboukh, L. and Pnrdee, A.B. (1993) Nucleic Acids Res. 21.3269-3275.
G. Conway I Mechanisms
of’Development
Liaw, C.W. and Towle, H.C. (1984) J. Biol. Chem. 259, 72537260. MacDougald, O.A. and Jump, D.B. (1992) Biochem. Biophys. Res. Commun. 188,4701176. McGrew, L.L., Dworkin-Rastl, E., Dworkin, M.B. and Richter, J.D. (1989) Genes Dev. 3.803-815. Narayan, P., Liaw, C. W. and Towle, H. C. (1984) Proc. Natl. Acad. Sci. USA 8 1.4687-469 1. Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989). Molecular Cloning: A Laboratory Manual, 2nd Edition. Cold Spring Harbor Laboratory Press, New York. Schulte-Merker, S., Ho, R.K., Herrmann, B.G. and Nusslein-Volhard,
52 (1995) 383-391
391
C. (1992) Development 116, 1021-1032. Shih, H. and Towle, H.C. (1994) J. Biol. Chem. 269,9380-9387. Sudo, Y., Goto, Y. and Mariash, C.N. (1993) Endocrinology 133, 1221-1229. Vassalli, J.-D., Huarte, J., Belin, D., Gubler, P., Vassalli, A.. O’Connell, M.L., Parton, L.S., Rickles. R.J. and Strickland, S. (1989) Genes Dev. 3, 2163-2171. Welsh, J., Chada, K., Dalal, S.S., Cheng. R., Ralph. D. and McClelland, M. (1992) Nucleic Acids Res. 20,496.5-4970. Westerfield, M. (1993) The Zebrafish Book: A Guide for the Laboratory Use of Zebrafish (Brachydanio rerio). University of Oregon Press, Eugene, OR.
Mechanisms of Development52 (1995) 383-391
A novel gene expressed during zebrafish gastrulation identified by differential RNA display Greg Conway* Department of Molecular and Cellular Biology, Harvard University, Cambridge, MA 02138, USA
Received 4 April 1995; acceptedMay 22 1995
Abstract Vertebrate gastrulation is a dynamic period of development characterized by extensive cell migrations. This stage of development is likely to require the expression of a new genetic repertoire to initiate and direct these dramatic changes. The differential RNA display has been used to identify genes specifically expressed during the gastrula stage of a model vertebrate, the zebrafish. One of the genes isolated by the differential display technique has been sequenced and characterized for its spatial and temporal expression. This gene, called G12, is expressed during a narrow window of time during gastrulation and is restricted to a single cell type. At this time of development the zebrafish embryo consists of three cell types: the yolk cell, EVL cells and deep cells. Interestingly, both EVL and deep cells derive early in development from common progenitor cells but G12 expression is restricted only to the EVL lineage. Comparison of the amino acid sequence from this gene with the Genbank database indicates similarity to two previously reported mammalian genes. The similarity between these three genes suggests that they may serve a common function. The Cl2 gene is the first example of restricted gene expression in EVL cells of the zebrafish. The G12 gene should prove to be a useful model for the study of regulated gene expression during gastrulation. Keywords:
Zebrafish; Gastrulation; G12 Expression; Differential RNA display
1. Introduction Vertebrate gastrulation is characterized by extensive cell migrations which transform a roughly spherical ball of cells into an elongated structure possessing the major body axis. Expression of many new genes is probably required during this dynamic period of development. The differential RNA display technique has been used to isolate genes which are differentially expressed between blastula and gastrula stage embryos in a model vertebrate, the zebrafish. This report describes the characterization of one of these genes, called G12. This gene is expressed in one of the first three cell types of the zebrafish embryo and expression of G12 is coincident with many changes in the morphology of the cells in which it is expressed. Although the first three cell types of the zebrafish can be described at the morphological level, little is know about gene expression or the molecular changes which * Tel.:
+I 617 4950560;
Fax: +1 617 4964313;
Email: [email protected].
0925-4773/95/$09.50 SSDI
occur in them during gastrulation. The zebrafish embryo begins as a single cell with a large centrally located yolk. Upon fertilization, cytoplasmic streaming produces a bulge of cytoplasm, the blastodisk, at the site of sperm entry. The blastodisk undergoes cleavage, and by the fourth division produces a 4 by 4 array of cells with the outer cells of this array connected by cytoplasmic bridges. These cells will continue to undergo incomplete cleavages and eventually fuze to produce a large syncytium encompassing the yolk. The inner cells of the 4 by 4 array undergo complete cleavage and give rise to two types of cells. Superficial descendents of these cells will produce a single-cell-thick covering which will surround the embryo, called the enveloping layer (EVL), while the inner descendents, called deep cells, will give rise to ectoderm, endoderm and mesoderm, producing the embryo proper. As development proceeds, the deep cells sitting on top of the large yolk cell, produce a ball of cells covered by the EVL. This period of development is the blastula stage. At the onset of gastrulation, changes in the yolk cell, and migrations of EVL cells and deep cells, transform the
0 1995 Elsevier Science Ireland Ltd. All rights reserved
0925-4773(95)00416-X
384
G. Conway I Mechunisms of Development 52 (1995) 383-391
embryo into an elongated structure possessing the body axes of the adult. An understanding of this dynamic period of development requires a knowledge of the genes specifically activated during this time. The differential RNA display has been used to compare the mRNA population of blastula and early gastrula embryos. This technique involves the PCR amplification of first strand cDNA using random sequence oligonucleotides (Liang and Pardee, 1992; Welch et. al., 1992; Liang et al., 1993). Under these conditions a random but reproducible subset of cDNAs are amplified. Using the conditions described in this paper and employed by others, approximately 50-100 cDNAs are amplified per primer set. If this amplification procedure is employed on different RNA samples, the 50-100 amplified cDNAs can be compared in the two RNA populations by polyacrylamide gel electrophoresis. Using only 240 oligonucleotide primer sets it is theoretically possible to sample the entire mRNA population of a cell, roughly 15 000 different RNA species (Liang and Pardee, 1992). Using this technique, we have discovered several genes differentially expressed during blastula or gastrula stages. The temporal and spatial expression for one of these genes has been analyzed and it was found that it is expressed during gastrulation in EVL cells. G12 should prove to be a useful system for the analysis of temporal and spatial gene regulation during gastrulation. G12 is similar to genes found in other organisms suggesting an evolutionarily conserved function for these proteins. 2. Results 2.1. Differential RNA display of blastula and gastrula RNA
RNA from blastula (4 h of development) and gastrula (6 h of development) zebrafish embryos was compared using the differential RNA display. All developmental times in this paper indicate times postfertilization at 28°C. A region of the gel displaying the transcripts amplified by one set of oligonucleotide primers is shown in Fig. 1 (see Section 4 for a description of the olignucleotides). There are several examples of bands which are present in one RNA sample but missing or much less intense in the other sample. Bands which were differentially represented in either the blastula or gastrula samples were excised from the gel, reamplified and cloned. Using combinations of four oligonucleotides, a total of 15 bands were successfully reamplified and subcloned from differential display. Random primed probes were then made from these clones and used to probe northern blots containing blastula and gastrula RNA. Four clones showed stage specific expression which agreed with the results from the differential display. Two of these clones were blastula specific and two gastrula specific. Two clones which were gastrulation specific according to differential display detected an RNA species of equal in-
B
bases 404
G
-
-G12
309
242 238 217 201
-
Fig. 1. Differential RNA display of blastula and gastrula RNA. First strand cDNA was synthesized from blastula (B) and gastrula (G) RNA. The cDNA samples were then PCR amplified in the presence of 32PdATP (see Section 4 for a description of the PCR primers). The PCR samples were then size fractionated by denaturing electophoresis and autoradiographed. Several bands are present in one sample but less intense or missing in the other sample. One gene, called G12 (indicated by an arrow), which is present only in the gastrula sample was isolated from the gel, reamplified and cloned.
tensity in both blastula and gastrula samples and were thus falsely represented by the differential display. Unfortunately 9 of the clones did not give a signal on northern analysis even when 2,ug of poly-A+ RNA was probed. The differential display, utilizing PCR amplification, is likely to amplify RNA species of low abundance which are difficult to detect and analyze further. The four clones which showed stage specific expression were further analyzed by in situ hybridization of whole mount zebrafish embryos. One of these genes, called G12, gave an interesting spatial and temporal expression pattern and was studied further. 2.2. Sequence analysis of G12 A gastrula (6 h) cDNA library was probed with an oligonucleotide homologous to sequence found in the PCR
G. Conway J h4echanisms
1 61
121 181 241 301 361 421 481 541 601 661 721 781 841 901 961 1021 1081 1141 1201
of Devetopment
GAGAMCTAAGCACACTAAAGACACGCATTCAG
52 (I 995) 383-391
385
XUAAACTGCTCTCGAATCAAACCACC
ATATGTGAGAGCTATATTAATGAAGCAGAAACCCTGTGGAGGATGTCTCC
60 120 180 240 300 360 420 480 540 600 660 720 780 840 900 960 1020
TGCTAATGACCTG~G~TCTGTGAATCATCCTGTG;CTCTG
1080
~~AGATGTCCGAGCCCCTCAGCCAGAAGAACGCTCTCTACACGGCAATGAACCGC TTCCTCGGCGCCGTCAACAACATGGACCAGACGGTCATGGAT GTGCCGCTGGACCAGGAGAGAGCAGCAGAAGCTGACCACTACCTG CGGGAGGCTGAGGCCGACATGTACAGCTATTACAGTCAACC ATCGAATGGGGCGTCATCCGCTCGGAGGACCAGCGGCGCAGACACCTCTGCGTCG GAGCCTGTGCGCACAGAGGAGGAGAGCGACATGGATCTCGCTCCTGCAGTTCCAT CTGAAGGGCTTGCACGGGGTGCTGTCCCAACTCAC~GCC~GCC~C~CCTGACC~C CGATACAAGCAGGAGATTGGCATTTCTGGCTGGGGACAdTGATC CACATCTACTGCGATATGATGAACTGTTGTGTTGTTCTAACTGCTCTGCATTTG TGGAAGAGCCAA?iGGGGACTAATGGAGCAGAAGTTACT=CG
AAAAMGACAAACCGTGA
AGGGAAAGTGGATTTGTCGAT~TGGTTTTGGTGTTGAGGG CTGCTTCCTGTTTTGTGCCATTTGCACTGACAGTTTCTCTTTTT~CC~~~TCTC CGTTGAAGTGGCTGACTGCCATGCATTTACATTFLATATTT~GTCTATTTATGTATTA~ TAAAATTGAAAGCTTTTATTTTTCCATTTTTCTTTGTGATAGAGG~GT~TTGTCCTA TGATGTTGCCAAGTGTACCAGAGGACTGATGACTTTTTTTTAT~TT~CTGAGATTGCT
AAAATCACTGTTTGTGTAATGATGAGCCTCCATGAGATTCAGTGTGF~CTGTGCTCATA
1140
AATGCCAGTTGAATTGTTAAAGCATTGCTGTGTTTTTATGTTAATG~CTTTTTTTTTTT
1200
-ATAACAAAATAAACGTTTGTTACTTG
1226
Fig. 2. Nucleotide sequence of G12. Boxed letters indicate the start and stop codons. The cytoplasmic polyadenylation polyadenlyation sequence: AATAAA are underlined. The Genbank accession number for this sequence is U27 121.
fragment of G12. The longest cDNA clone, called pG12 was sequenced and is presented in Fig. 2. Two reading frames contain stop codons throughout the sequence and cannot be the correct reading frame. One reading frame is devoid of stop codons for the first 525 nucleotides of sequence. The first methionine in this reading frame begins at position 64 and is a likely candidate for the start codon. However, upstream from this methionine and in the same reading frame there are no stop codons. This raised the possibility that the cDNA clone, pG12, was a partial cDNA. To determine if there are additional nucleotides missing from pG12, 5’ RACE was performed on gastrula cDNA. From the 5’ RACE analysis, two additional nucleotides were identified and are the first two residues shown in Fig. 2. With the addition of these nucleotides to the sequence derived from pG12, the methionine at position 64 in Fig. 2 is the first initiation codon, indicating that the first 63 nucleotides are 5’ untranslated sequence. The G12 gene encodes a long 3’ untranslated region of 706 nucleotides, over half the length of the transcript. Interestingly this 3’ sequence contains a cytoplasmic polyadenylation sequence: UUUUUAU. This motif is found in several genes and is necessary for developmentally regulated cytoplasmic lengthening of the poly-A tail (Fox et al., 1989; McGrew et al., 1989; Vassalli et al., 1989; Jackson, 1993). In several instances this lengthening of the poly-A tail has been correlated with increased translation of the mRNA (McGrew et al., 1989; Vassalli et al., 1989). The possibility that G12 undergoes cytoplasmic polyadenylation requires further investigation.
sequence:
TTTTTAT
and the
The amino acid sequence of G12 was compared to other sequences in the Genbank database and found to be similar to two previously reported sequences: rat spot 14 and human est00887. The amino acid sequence of G12 is compared to the sequence of spot 14 and human estO0887 in Fig. 3. Amino acids which are identical in G12 and spot 14 or est00887 are shown as bold letters. Both the G12 protein and spot 14 protein are nearly identical in size, roughly 17 kDa. Unfortunately the complete sequence of human est00887 is lacking. Human est00887 is an expressed sequence tag isolated from a human hippocampal cDNA library. Until this clone is fully sequenced it will not be possible to determine if the size of the est00887 protein is similar to that of G12 and spot 14. The greatest degree of similarity between G12, spot 14 and estOO887 lies in the carboxyl region of these proteins. For human est00887, 68% of the last 41 amino acids are identical to G12. Compared to G12, 37% of the last 38 residues of spot 14 are identical. The high degree of similarity in the carboxyl region of these proteins suggests that this region may fulfill a common function in these proteins. In the carboxy terminal sequence of G12 there is a possible leucine zipper characterized by the periodic spacing of four leucine resides every seven amino acids (shown by vertical lines in Fig. 3). 2.3. Developmental expression profile of G12 From northern blot analysis the G 12 gene produces a 1.4 Kb transcript (Fig. 4A). The size of this transcript is in agreement with the 1234 nucleotide long sequence for G12, adding roughly
G. Crmwuy I Mechanisnu ofDevelopment 52 (1995) 383-391
386
1
50
zebrafish G12 human est00887 rat spot 14
MQVLTKRYPKNCLLKVMDRYSAWRNMEQVVMIPSLLRDVEL........
zebrafish G12 human est00887 rat spot 14
51 100 LTNDPGSYLREAEADMYSYYSQLKSIRNNIEWGVIRSEDQRRKKDTSASE PPTGLGXXRAVPGSPKDILVDLG ..SGSGGSVQDGAPDLYTYFTMLKSICVEVDHGLLPREEWQAKVAGNEGS
zebrafish G12 human est00887 rat spot 14
PVRTEEES.......... .DMDLEQLLQFHLKGLHGVLSQLTSQANNLTN HLEGSDAG.......... .EEDLEQXFHYHLRGLHTVLSKLTRKANILTN EAENEAAETEEAEEDRLSEELDLEAQFHLHFSSLHHILTHLTQKAQEVTQ
zebrafish G12 human est00887 rat spot 14
151 163 RYKQEIGISGWGQ RYKQEIGFGNWG KY.QEMTGQV
MQMSEPLSQKNALYTPMNRFLGAVNNMDQTVMVPSLLRDVPLDQEKEQQK
101
150
I
I
I
I
Fig. 3. Amino acid sequence of G12 and comparison with rat spot 14 and human est00887. The amino acid sequence of Gl2, rat spot 14 and human est00887 have been aligned to maximized similarity. Amino acids which arc identical in G12 and spot 14 or human est 00887 are in bold. The position of leucine resides in Cl2 which could theoreticaly form a leucine zipper are indicated by vertical lines.
200 nucleotides of poly-A. The expression of G12 RNA throughout development was analyzed by RT-PCR and compared to the expression of a housekeeping gene which is required for translation, EFl alpha. As shown in Fig. 4B, there is low level expression of G12 throughout development, however, this gene is intensely expressed at 6 h of development and is again at low levels by 8 h. The level of EFl alpha on the other hand is constant from 6 h throughout development. The increased expression of
EFl alpha at 6 h of development dramatically shows the switch to zygotic gene expression. In the zebrafish, zygotic transcription does not begin until the midblastula stage (3 l/3 h of development). This is known as the midblastula transition or MBT (Kane and Kimmel, 1993). Prior to this stage, embryonic development is driven by maternal RNAs placed in the egg by the ovary. Cl2 is activated coincident with EFl alpha, at or shortly after MBT, however, in contrast to EFl alpha G12 expression
A G12
MW
Kb - 8.5 - 7.5
- 2.4
G12 EFl alpha
Fig. 4. Northern blot and RT-PCR analysis of Gl2 expression. (A) A northern blot containing 2yg poly-A+ RNA from gastrula stage (6 h) zebrafish embryos was probed for G12 transcripts. A transcript of 1.4 Kb is detected. (B) Total RNA, isolated from staged embryos or adults, served as template for RT-PCR using oligonucleotide primers complementary to G12 or EFl alpha. PCR products were then size fractionated, transfered to nylon membranes and probed with an oligonucleotide complementary to a region between the PCR primers. Hours of development indicate times postfertilization at 28°C.
G. Conway I Mechanisms
3hrs
6hrs
Fig. 5. labeled probes tion at
qf Development 52 (1995) 383-391
4hrs
387
Shrs
1Zhrs
8hrs
Developmental profile of Cl2 RNA revealed by in situ hybridization, Staged embryos were probed with either sense or antisense digoxigenin RNA probes to G12. The presence of G12 RNA is indicated by blue staining. The orange color of the yolk seen with both sense and antisensc is due to endogenous phosphatase activity and does not indicate the presence of G12 RNA. Hours of development indicate times postfertiliza28°C.
Sense
Antisense
, ,.*w, i b ‘-
,
Fig. 6. In situ localization of G12 RNA using thin sectioned embryos. Sections (5pm thick) were made from 5 h zebra&h embryos and probed with either sense or antisense digoxigenin labeled RNA probes to G12. The presence of Cl2 RNA is indicated by blue staining. The arrow points to EVL cells which show the presence of G12 RNA. The blue color of the yolk seen with both sense and antisense probes is due to endogenous phosphatnsc activity and does not indicate the presence of G12 RNA.
388
dramatically decreases, window for this gene.
G. Conwuy I Mechanisms
producing
2.4. Sites of Cl2 expression
a narrow
of’ Development 52 (1995) 383-391
expression
revealed by in situ
hybridization
To determine the spatial localization of Cl2 expression, in situ hybridization was performed with whole mount zebrafish embryos using anti-sense RNA probes. The in situ hybridizations are shown in Fig. 5; the presence of G12 transcripts is indicated by a blue color. Sense strand probes were used as a control and showed no staining. G12 RNA is first detected at 4 h of development in isolated cells. Staining is most intense at 5 h, begins to disappear at 6 h and random staining of isolated cells is seen at 8 h of development. Often cells which express G12 at 8 h of development are in clumps, although never observed with a pattern to this grouping. From 12 to 72 h of development Cl2 expression is not detected. The temporal pattern of staining revealed by in situ hybridization parallels that seen with RT-PCR. The staining of 5 h embryos shown in Fig. 5 suggests that just the outer surface layer of cells, the EVL, expresses G12. To determine if EVL cells are the only cells expressing G12, in situ hybridization was performed on sectioned embryos (Fig. 6). Embryos at 5 h of development were sectioned and hybridized with sense and antisense probes. Again sense probes did not show staining, however, the antisense probe stained only the EVL cells (shown by an arrow head). The blue staining of the yolk is due to endogenous phosphatase activity and is seen with both sense and antisense probes and does not indicate the presence of G12 RNA. 3. Discussion The beginning of gastrulation is likely to require the expression of new genes. To isolate genes activated during gastrulation we have used the technique of the differential RNA display to compare the mRNA population of blastula and gastrula zebrafish embryos. The differential RNA display is a rapid method for comparing RNA samples from different developmental stages. Using only a few primer combinations, four genes were isolated which are differentially expressed in either blastula or gastrula zebrafish embryos. One of these genes is described in detail in this report. There are some caveats to the display technique which should be noted. Some transcripts which are amplified by the display technique are of low abundance and can not be detected on Northern blot analysis. These genes may be interesting but are difficult to analyze further. Also, some genes which according to the display technique are differentially expressed were later found to be expressed at equal levels in both stages (2 of the 6 genes which gave signals on Northern analysis). The presence of these falsely represented transcripts makes it impossible to es-
timate the general percentage of stage verses nonstagespecific gene expression. Any gene detected to be differentially expressed between 2 RNA samples by differential display should be further checked on Northern blots, or by a similar means such as RT-PCR, to ensure that these genes are actually differentially expressed. The reason for the false distribution sometimes seen with the differential display is unknown. Also it should be noted that the intensity of bands on the differential display does not reflect the abundance of the mRNA in the RNA sample. Some of the most intense bands on the display failed to detect a transcript on northern analysis. The intensity of bands on differential display is dependent not only upon the abundance of that species in the RNA sample but also on the degree of homology between the RNA target and the PCR primers. The gene reported here, G12, has an interesting temporal and spatial expression pattern. G12 mRNA is detected in 4 to 8 h embryos, with peak expression at 5 h of development. This gene is activated at or shortly after MBT, but unlike the housekeeping gene EFl alpha, expression decreases shortly after its rapid induction. In addition to temporal regulation, G12 is also spatially restricted with expression only in the EVL cells, one of the first three cell types of the zebrafish embryo. Interestingly, both EVL and deep cells originate from common progenitor cells, yet expression is restricted to only one of these lineages. The restricted spatial expression of G12 indicates that transcriptional activation of this gene is not the result of general nonspecific activation of the transcriptional apparatus in all cells of the embryo at MBT. Expression of G12 is maximal at the onset of gastrulation and ends shortly before the completion of gastrulation. At the begining of gastrulation EVL cells adopt an epithelial morphology and are required to stretch and cover a greater surface area due to extensive cell movements which occur at this time. It would be interesting to determine if the G12 gene plays a causal role in any of these changes. The origin of EVL cells (an extraembryonic structure) and deep cells (the cells which will give rise to the embryo) is similar to the origin of extraembryonic and embryonic structures in mammals. During mammalian development the early blastula contains cells which can give rise to either embryonic or extraembryonic tissues. As mammalian development proceeds, the outer cells of the embryo (the trophectoderm) give rise to extraembryonic tissues while inner cells (the inner cell mass) produce the embryo proper. In many ways the EVL and trophectoderm are analogous structures; both are organized into an epithelium and originate from superficial cells. Deep cells and the inner cell mass are similar in that both are interiorly located, give rise to embryonic tissues and remain pluripotent for a considerable time in development. The expression of G12 in EVL cells raises the interesting possibility that a similar gene(s) may exist in mammals
G. CrmwuyI Mechanismsof Development52 (1995) 383-391 and be expressed in extraembryonic tissues early in development. A search of the Genbank database indicates that genes similar to G12 do exist in mammals. Two previously reported genes are similar to G12: rat spot 14 and human estO0887. Rat spot 14 was first discovered on 2 dimensional gels as a liver protein whose abundance rapidly increased in response to thyroxin (Liaw and Towle, 1984; Narayan et al., 1990). The gene was subsequently cloned and shown to be regulated by insulin and retinoic acid as well as thyroxin (Jump et al., 1990; MacDougald et al., 1992). Spot 14 has become a model for the study of regulated gene expression in the liver. The promoter region of this gene has been extensively analyzed and several cis-elements necessary for response to glucose and thyroxin determined (Jump et al., 1993; Sudo et al., 1993; Shih and Towle, 1994). It would be interesting to see if G 12 is also regulated by insulin, retinoic acid or thyroxin. Despite an impressive wealth of knowledge regarding the regulation of spot 14, little is know about its function. Antibodies have been produced to spot 14 and immunohistochemistry performed (Kinlaw et al., 1992, 1993). Interestingly, spot 14 is a nuclear protein. Unfortunately the antibodies produced to spot 14 are against peptides which are not conserved in G12 and are of no value in determining the localization of G12 in EVL cells. Currently, antibodies are being produced against zebrafish G12. All that is known about est00887 is that it is expressed in the human hippocampus, since it was derived from a hippocampal cDNA library. It would be interesting to examine the expression of spot 14 and est00887 during development; perhaps they are expressed at an earlier time. The amino acid sequence conservation seen in the carboxy terminal region of all three proteins suggests that this sequence may be a functional domain and fulfill an analogous role in all three. It would be valuable to search for other proteins with a sequence similar to this carboxyl region. G12, spot 14 and human estO0887 may be three members of a much larger gene family. 4. Materials and methods 4. I. RNA isolation All developmental times refer to development at 28°C. Total RNA was isolated from staged zebrafish embryos by the guanidinium isothiocyanate/phenol method using a Stratagene Total RNA Isolation Kit as described by the manufacturer. Poly-A+ RNA was isolated from total RNA using a Fasttrack mRNA isolation Kit from Invitrogen as described by the manufacturer. 4.2. Differential RNA displa) Differential RNA display was performed using a modified protocol of Liang and Pardee (1992) and Welch et al. (1992). First strand cDNA was synthesized from
389
1 ,ug blastula (4 h embryos) or gastrula (6 h embryos) total RNA. Conditions for first strand synthesis were 50 mM Tris (pH 8.3), 75 mM KC], 10 mM DTT, 5 mM MgCl2 and 20pM dNTPs in the presence of 1 pm anchor primer in a total volume of 20~1. Two different anchor primers were separately used to synthesize cDNA. The sequence of these primers were: Anchor1 S-CCGAATTCWCG-3’ and Anchor 2 5’-CCGAATTC-CA-3’. Samples were heated to 65°C for 5 min, followed by 10 min at 37°C. To the reaction 200 units MMLV reverse transcriptase (BRL Inc.) were added. After 50 min at 37”C, the reverse transcriptase was inactivated by incubation at 95°C for 5 min. PCR was then performed using 5 ~1 of the first strand reaction in a total reaction volume of 50~1. The PCR reactions contained 30 mM Tris (pH 8.3), 32.5 mM KCI, 3 mM MgCl*, 50pM dNTPs, 5 units Taq polymerase (Cetus Inc.), 20,& 32P-dATP (NEG-0122, New England Nuclear), 1 PM of an anchor primer and 1 PM of a 5’-primer. Two different 5’-primers were separately used in combination with the two anchor primers. The sequence of the 5’primers were: 5’-primer1 5’-GCACGGATCCGGTAGATAG-3’ and 5’-primer2 5’-GCACGGATCCTGATCCATG-3’. PCR reactions were subjected to one cycle of 94°C for 5 min, 40°C for 5 min, and 72°C for 5 min. Reactions were then subjected to 30 cycles of 1 min 94”C, 1 min 55”C, and 1 min at 72°C. Aliquots (5~1) of the PCR reactions from blastula and gastrula samples were run in adjacent lanes of a 6% denaturing polyacrylamide gel. Following electrophoresis, the gel was wrapped in Saran wrap, autoradiographed and bands present in either the blastula or gastrula lanes excised with a razor blade. The gel slices were soaked for 10 min in water to remove urea. This water was then replaced with 50~1 of water, heated for 2 h at 95°C and 10~1 used for reamplification. Individual eluted bands were amplified by PCR using the reaction conditions described above and were subjected to 30 cycles of 94°C for 1 min, 55°C for 1 min, and 72°C for 1 min. Aliquots (5~1) of the reamplified bands were checked on 6% non-denaturing polyacrylamide gels stained with ethidium bromide to insure that fragments the correct size were geneated during the second round of amplification. The remaining PCR reaction was then cloned into the pCRI1 vector (Invitrogen Inc.) as described by the manufacturer. 4.3. Sequencing of PCR fragments and Northern blot analysis Random primed probes were generated from PCR fragments cloned into the pCRI1 vector using a Random Prime Kit (BRL Inc.) as described by the manufacturer. These probes were used to screen Northern blots containing 10 pug of total RNA and 2,ug of poly-A+ RNA from blastula (4 h) and gastrula (6 h) stage embryos. Northern analysis was performed as described by Sambrook et al. (1989). Clones which gave signals that agreed with the
390
G. Conwuy
I Mechanisms
of Development
restricted and staged expression found from the original differential display were sequenced. Clones were sequenced using double stranded DNA templates and a Sequenase Kit (United States Biochemical) as described by the manufacturer. 4.4. Library screening, sequencing, and 5’ RACE A gastrula (6 h) cDNA library was constructed in the lambda Zap11 vector using a Zap cDNA Library Kit from Stratagene Inc. as described by the manufacturer. This library was probed with an oligonucleotide found in the PCR fragment of one of the genes, called G12, identified by the differential display and confirmed by northern analysis to be expressed during gastrulation but not during the blastula stage. This oligonucleotide had the sequence: S-GACAAATCCACTTTCCCTICACGGTITGTC-3’. Phage were plated and transfered to nylon disks (Biodyne, Pall Inc.) as described in Sambrook et al. (1989). Membranes were probed at 42°C and washed at 55°C using buffers described by Church and Gilbert (1984). Positive clones were plaque purified and converted to pBluescript plasmids (Stratagene, Zap cDNA Kit) as described by the manufacturer. The longest clone, called pG12, was sequenced using a Sequenase Kit as described by the manufacturer. To insure that 5’ sequence was not missing from pG12, 5’ RACE was performed using a 5’-AmpliFINDER RACE Kit (Clonetech Inc.) as described by the manufacturer. Only two 5’ nucleotides were found to be missing from this cDNA clone. 4.5. Analysis of RNA by RT-PCR First strand cDNA was synthesized from 1 pg total RNA as described above. For the developmental analysis of EFl alpha 1~1 of the 20~1 first strand reaction was diluted 1: 150 with water and 1~1 used for PCR. For the analysis of G12, cDNA was diluted 1:30 with water and 1~1 used for PCR. Reaction conditions for PCR were the same as described above, however, the final reaction volume was 50~1 and the reactions cycled 25 times: 95°C for 30 s, 55°C for 30 s and 72°C for 1 min. Oligonucleotide primers for EFl alpha had the sequence: 5’-TGGGCACTCTACTTAAGGAC-3’ and 5’-TGTGGCAACAGGTGCAGTTC-3’. Oligonucleotide primers for analysis of G 12 were: S’-TCATTAGCAGGAGACATCCT-3’ and 5’-AAGTGGCTGACTGCCATGCA-3’. Following PCR, for each developmental stage 2~1 of the 50~1 reactions for EFl alpha and G12 were pooled and subjected to 5% denaturing polyacrylamide gel electrophoresis. Gels were then electrobloted to nylon membranes (Hybond-N, Amersham Inc.). Blots were then baked, UV crosslinked and probed with oligonucleotides to EFl alpha (5’-TGACTGATAGTGCCTCTITC-3’) or Cl2 (5’-GTCATCAGTCCTCTGGTAC-3’) at 45°C and washed at 55°C using buffers and conditions described in Church and Gilbert (1984). Filters were then subjected to autoradiography and bands quantitated using a Fuji Bio-
52 (1995)
383-391
Image Analyzer (Fuji Photo Film Co.. Ltd.). To insure that the intensity of the bands in the RT-PCR analysis reflect the abundance of the analyzed message, RT-PCR analysis was performed in triplicate using 1~1, 3 ~1. and 9 ,ul of cDNA diluted as described above. The intensity of bands was found to reflect the amount of input cDNA and hence is quantitative and within the linear range of PCR amplification. 4.6. In situ hybridization In situ hybridization was performed on whole mount embryos exactly as described by Schulte-Merker et al. (1992). Digoxigenin labeled sense strand RNA probes were transcribed from KpnI linearized pG12 using T3 RNA polymerase. Anti-sense RNA probes were synthesized from EcoRI linearized pG12 using T7 RNA polymerase. In situ hybridization was also performed on 5 micron thick cryostat sections. Embryo treatment, sectioning and hybridizations were performed as described in Westerfield (1993). 4.7. Genbank accession number The accession number for the sequence reported in this paper is U27 12 1. Acknowledgments I thank Jim Wilhelm and other members of the Gilbert laboratory for technical assistance and helpful discussions. Thanks to Doug Melton and members of his laboratory for helpful advise and use of equipment. I am grateful to Barrett Klein and Michiel C. J. Bliemer for help with the figures. Special thanks to Richard Roberts for advice on the manuscript and helpful discussions. G. C. was supported by an American Cancer Society postdoctoral fellowship and was a Medical Foundation/Charles A. King Trust Research Fellow. This work was supported by a grant from the National Insitutes of Health to Walter Gilbert (EY 08397). References Church G. and Gilbert, W. (1984) Proc. Natl. Acad. Sci. USA 81, 1991-199s. Fox, CA., Sheets, M.D. and Wickens, M.P. (1989) Genes Dev. 3. 2151-2162. Jackson, R.J. (1993) Cell 74,9-14. Jump, D.B., Bell, A. and Santiago, V. (1990) J. Biol. Chem. 265, 34743478. Jump, D.B., Clarke, S.D., MacDougaJd, 0. and Thelen, A. (1993) Proc. Natl. Acad. Sci. USA 90, 8454-8458. Kane, D.A. and Kimmel, C.B. (1993) Development 119,447-456. Kinlaw, W.B., Tron, P. and Friedmann, AS. (1992) Endocrinology 131,3120-3122. Kinlaw, W.B., Tron, P. and Witters, L.A. (1993) Endocrinology 133, 645-650. Liang. P. and Pardee. A.B. ( 1992) Science 257.967-97 1. Liang, P., Averboukh, L. and Pnrdee, A.B. (1993) Nucleic Acids Res. 21.3269-3275.
G. Conway I Mechanisms
of’Development
Liaw, C.W. and Towle, H.C. (1984) J. Biol. Chem. 259, 72537260. MacDougald, O.A. and Jump, D.B. (1992) Biochem. Biophys. Res. Commun. 188,4701176. McGrew, L.L., Dworkin-Rastl, E., Dworkin, M.B. and Richter, J.D. (1989) Genes Dev. 3.803-815. Narayan, P., Liaw, C. W. and Towle, H. C. (1984) Proc. Natl. Acad. Sci. USA 8 1.4687-469 1. Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989). Molecular Cloning: A Laboratory Manual, 2nd Edition. Cold Spring Harbor Laboratory Press, New York. Schulte-Merker, S., Ho, R.K., Herrmann, B.G. and Nusslein-Volhard,
52 (1995) 383-391
391
C. (1992) Development 116, 1021-1032. Shih, H. and Towle, H.C. (1994) J. Biol. Chem. 269,9380-9387. Sudo, Y., Goto, Y. and Mariash, C.N. (1993) Endocrinology 133, 1221-1229. Vassalli, J.-D., Huarte, J., Belin, D., Gubler, P., Vassalli, A.. O’Connell, M.L., Parton, L.S., Rickles. R.J. and Strickland, S. (1989) Genes Dev. 3, 2163-2171. Welsh, J., Chada, K., Dalal, S.S., Cheng. R., Ralph. D. and McClelland, M. (1992) Nucleic Acids Res. 20,496.5-4970. Westerfield, M. (1993) The Zebrafish Book: A Guide for the Laboratory Use of Zebrafish (Brachydanio rerio). University of Oregon Press, Eugene, OR.