GENOMICS
46, 337–344 (1997) GE975117
ARTICLE NO.
cDNA Libraries from Single Human Preimplantation Embryos James Adjaye,*,1 Rob Daniels,* Virginia Bolton,† and Marilyn Monk* *Molecular Embryology Unit, Institute of Child Health, 30 Guilford Street, London WC1N 1EH, United Kingdom; and †Assisted Conception Unit, Department of Obstetrics and Gynaecology, King’s College School of Medicine and Dentistry, Denmark Hill, London SE5 8RX, United Kingdom Received August 5, 1997; accepted November 6, 1997
In this paper, the construction, evaluation, and application of cDNA libraries from eight unfertilized oocytes and single four-cell-, seven-cell-, and blastocyst-stage embryos are described. Rapid, reproducible, and efficient procedures for the construction of PCR-based cDNA libraries from fewer than 10 cells were first developed in small populations of fibroblast cells. The human embryo libraries display complexities sufficient (between 105 and 106 clones) to represent the entire active gene population at these early stages of human development. The ubiquitous cytoskeletal elements, b-actin, keratin-18, and a-tubulin, were detected at the expected frequency. Sequencing of consecutively picked random clones, without selection, showed the presence of a variety of sequences, such as the human transposable element, LINE-1 and Alu repeat sequences, housekeeping genes, and tissue-specific genes, such as a-globin and FMR-1. In addition to cDNAs corresponding to known ESTs (expressed sequence tags) in the GenBank and dbEST databases, a high proportion of novel sequences were detected. Applications of the libraries to several areas of interest, such as expression of CpG-island-containing ‘‘tissue-specific’’ genes, developmental genes expressed in a stage-specific manner, and a search for monoallelic expression of imprinted genes, are described. The libraries are a valuable resource for the study of gene expression during human preimplantation development and obviate the need for research on the human embryos themselves. q 1997 Academic Press
INTRODUCTION
Molecular analyses of early human embryos are important and necessary for identifying the genes involved in normal development, their function, and the regulation of their expression. Such molecular analyses will lead to a greater understanding of the consequences, and ultimately the diagnosis, of abnormal function leading to congenital abnormality and inherited genetic disease. Although animal models, especially the mouse, 1
To whom correspondence should be addressed. Telephone: (/44) 171 242 9789 12280/2181. Fax: (/44) 171 404 6191. E-mail: J.
[email protected].
have provided us with much of our knowledge of early mammalian development, there are many differences among mammals. In the final analysis, it is necessary to study the human embryos themselves. In vitro fertilization (IVF) procedures for the treatment of infertility provide access to these early stages of human development. However, samples are rare and difficult to obtain, and usually only degenerate embryos, i.e., those unsuitable for transfer to the mother or freezing for later transfer, are available. In Britain, all research must be licensed by HFEA (Human Fertilisation and Embryo Authority). In addition, even when embryonic material is available, molecular analyses are extremely difficult because of the relatively small size of the embryo, which consists of only a few cells. The availability of human preimplantation embryo cDNA libraries would circumvent these difficulties and facilitate the study of expression of known genes, as well as allow the isolation and identification of novel stage-specific embryonic genes. To create libraries from a limited number of cells, it is necessary to amplify the cDNA corresponding to the mRNA population by a polymerase chain reaction (PCR) procedure (Saiki et al., 1988). The construction and differential screening of representative cDNA libraries made by the PCR approach have led to the identification and characterization of genes differentially expressed during mammalian embryonic development (Rothstein et al., 1992; Smith and Gridley, 1992; Hwang et al., 1996). So far, PCR-generated cDNA libraries based on the techniques established by Beliavsky et al. (1989) and Brady et al. (1990) have been made from mouse oocytes (Welsh et al., 1990; Revel et al., 1995), preimplantation mouse embryos (Weng et al., 1989; Temeles et al., 1994), postimplantation mouse embryos (Varmuza and Tate, 1992), human fetal tissues (Buraczynska et al., 1995; Jay et al., 1997), and quail embryo neural crest cells (Bevan et al., 1996). The limited availability of human preimplantation embryos requires modifications of previously established techniques to achieve sufficient sensitivity and efficiency. These modifications were devised and developed in the construction of libraries from cultured hu-
337
AID
GENO 5117
/
6r54$$$521
12-09-97 20:27:27
0888-7543/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
gnmal
338
ADJAYE ET AL.
man fibroblast cells before attempts were made to construct the embryo libraries. Once developed, we used the procedures to construct representative cDNA libraries from human preimplantation embryos. This paper describes the construction and characterization of human oocyte, four-cell-, seven-cell-, and blastocyst-stage embryo libraries, which have complexities of 2.0 1 106, 5.0 1 105, 3.5 1 106, and 2.5 1 106 clones, respectively. As far as we know, this is the first documentation of cDNA libraries prepared from human preimplantation embryos. Analysis of the cDNA libraries has revealed their potential as a limitless resource for the characterization of expression of specific genes and the isolation and identification of new genes active at these early stages, thus allowing further elucidation of the genetic program of development. Some applications of the libraries in the analysis of unexpected expression of tissue-specific genes, the expression of stage-specific developmental genes, and investigation of the regulation of X-chromosome inactivation and imprinting are described. MATERIALS AND METHODS Fibroblast cells. Cultured human fibroblasts were obtained from John Rainer, Institute of Child Health (ICH), and isolated in groups of 10, 50, and 100, in droplets of PBS under oil, using finely pulled glass pipettes. Fibroblasts were isolated in a tissue culture room separate from the laboratory. Embryo collection and lysis. Embryos derived by in vitro fertilization were donated for research by patients, with appropriate advice and consent, at the Assisted Conception Unit, King’s College Hospital (KCH), London. Samples are anonymized at the source so that, in the case of unexpected equivocal findings, parents cannot be traced. All embryos were placed in acid Tyrode’s solution (pH 2.0–2.4) and observed under a dissection microscope until the zona pellucida had just dissolved (between 20 and 40 s). Great care was taken to remove any remaining cumulus cells. The collection of embryos into lysis buffer was carried out at the Assisted Conception Unit, KCH, as follows: Single embryos in 0.5 ml of PBS were added to 1.5 ml lysis buffer [0.8% Igepal (Sigma), 1 U/ml RNasin (Gibco BRL), 5 mM DTT (Gibco BRL)], centrifuged briefly at 12,000g, and overlaid with one drop of mineral oil (Sigma). Samples were snap frozen in liquid nitrogen and stored at 0707C. The treated samples were transferred, on dry ice, to the laboratory at ICH, where the reverse transcription (RT) reaction was carried out. Before reverse transcription (see below), samples were heated at 807C for 5 min and then transferred immediately to ice before addition of reverse transcription reagents. Great care was taken to avoid contamination. Preparation of reaction mixtures and addition or removal of aliquots were carried out in a fume hood, using aerosolresistant pipette tips. Reverse transcription. The use of an oligo(dT) linked to an EcoRI restriction enzyme sequence as a primer to initiate cDNA synthesis primes the mRNA molecules from their 3 * poly(A) sequence. The attached EcoRI sequence serves as a priming site for the PCR amplification step and, later, allows ligation and excision of cDNAs from the vector (see Fig. 1, steps 1 and 2). The reverse transcription reagents (11 reverse transcription buffer (Gibco BRL), 40 U MMLV reverse transcriptase/superscript (Gibco BRL), 0.9 mM each deoxynucleotide (Pharmacia), 5 mM DTT, 2.25 mg oligo(dT)–EcoRI and 2 U RNasin) were mixed on ice, and 3 ml was added to the cell lysate to produce a final volume of 5 ml. The reverse transcriptase enzyme was omitted from negative controls. Reverse transcription to produce cDNA was carried out at 377C for 1.5 h. Samples were then immediately returned to ice.
AID
GENO 5117
/
6r54$$$521
12-09-97 20:27:27
Poly(dG) tailing. Elongation of the 3* end of the first-strand cDNA (cDNA:mRNA hybrid) with deoxyguanosine residues [poly(dG) tailing] was carried out using terminal deoxynucleotidyl transferase (see Fig. 1, step 3). Under our conditions we estimate the addition of 10 to 15 nucleotides at the 3* ends of the cDNA sequences. However, it is possible that the enzyme may add G residues to the 3 * end of the mRNA but these will be lost during the amplification step. Tailing was carried out as recommended by the manufacturer (Gibco BRL) in a total volume of 10 ml. Library construction. The dG tailing allows the use of distinct 5* and 3* primers during PCR to amplify cDNA, followed by EcoRI digestion, size fractionation, and ligation into a vector of choice (see Fig. 1, steps 4 and 5). The oligo(dG)-tailed first-strand cDNA was amplified in a 50-ml reaction volume consisting of 10 mM each of dT-EcoRI and dCEcoRI primers (see Table 1), 0.5 mM each dNTP (Pharmacia), 5 units of cloned Pfu polymerase (Stratagene), and 11 PCR buffer (Perkin– Elmer). The reaction cycle consists of an initial denaturation at 967C for 10 min, 20 cycles of denaturation at 967C for 2 min, annealing at 587C for 2 min, and elongation at 727C for 4.5 min, and then a final elongation step at 727C for 10 min, using a Perkin–Elmer thermal cycler. The high GC content of the primers permits annealing temperatures of up to 587C during PCR and as such increases the specificity of priming. The final elongation step of 10 min is designed to produce full-length double-stranded cDNA bearing doublestranded EcoRI restriction sites at the 5* and 3* ends. The next step involves further amplification of selected longer length products (to obtain longer cDNAs in the library) from the first round of PCR (Beliavsky et al., 1989). Forty-five microliters of the first-round PCR product was resolved on a 1.2% agarose gel. Two gel slices of sizes 200–500 bp and ú500 bp were excised, and the cDNA was eluted using Qiagen elution columns (Qiagen, UK). The eluted cDNAs (50 ml) were reamplified in a 100-ml reaction volume under the same PCR conditions as before. Upon completion of amplification, the cDNAs were pooled, phenol–chloroform extracted, and precipitated with 1/10 vol of 3 M sodium acetate and 2.5 vol ethanol, and the pellet was resuspended in 50 ml distilled water. To enable ligation into the vector, the cDNAs were digested with EcoRI and resolved on a 1.2% agarose gel, and the cDNAs were excised, eluted, and precipitated as before. Approximately 200 ng of cDNA was ligated into Lambda ZAP II and pBluescript vector (Stratagene). The cDNAs ligated into Lambda ZAP II (four-cell and seven-cell embryos) were packaged in vitro and infected into the Escherichia coli strain XL1-Blue MRF* (Stratagene). The oocyte and blastocyst cDNAs were ligated into pBluescript and transformed into E. coli strain XL1-Blue MRF*, following the manufacturer’s instructions. Using the blue/colorless scoring system, approximately 90 to 95% of the plaques and colonies were colorless (recombinant). Finally, the Lambda ZAP II libraries were amplified and titered according to standard protocols (Stratagene), and they were stored after the addition of 10% DMSO (Sigma). The transformed bacteria representing the oocyte and blastocyst libraries were spread onto nitrocellulose filters previously layered onto agar plates containing ampicillin (50 mg/ml). The bacterial colonies, after overnight growth at 377C, were scraped off the filters and stored at 0807C after the addition of 10% glycerol (Sigma). PCR amplification of cDNAs directly from phage and bacterial colonies. Using the KS and SK primers (see Table 1) flanking the unique EcoRI site present in Lambda ZAP II and pBluescript, PCR amplifications typically consisted of an initial denaturation step at 947C for 10 min followed by 30 cycles of denaturation at 967C for 2 min, annealing at 557C for 2 min, elongation at 727C for 3 min, and then a final elongation step at 727C for 10 min. Library screening and nucleotide sequencing. To quantify the relative abundance of known transcripts, phage were plated and plaque lifts were performed as described in Adjaye et al. (1993). Gel-purified cDNA for specific genes to be used as probe (see Results for details of specific probes chosen) was labeled using [a-32P]dCTP and a random labeling kit (Pharmacia). Oligonucleotides derived from specific genes were also used as probes for screening (see Table 1) and were end-labeled with T4 polynucleotide kinase and [g-32P]dATP. Hybridizations were performed as described previously (Adjaye et al., 1995).
gnmal
HUMAN PREIMPLANTATION EMBRYO LIBRARIES
339
Positively reacting plaques were confirmed by replating and rehybridization of the filters. The corresponding cloned cDNA was excised in vivo to release the pBluescript phagemid (Stratagene). Confirmation of the identity of these clones was also obtained by Southern blotting (Southern, 1975) using the same probe. Further isolation, amplification, restriction enzyme analysis, and sequencing of plasmid DNA followed established protocols (Sambrook et al., 1989). Sequence analyses were performed with computer programs implemented within the University of Wisconsin Genetics Computer Group software on the HGMP network.
RESULTS
In preliminary work, we used decreasing numbers of human fibroblasts as starting material to synthesize cDNA. By continual refinement of the procedures, we could finally reproducibly generate cDNAs (up to 1000 bp) using only 10 human fibroblast cells. Analysis of the expression library obtained from these cDNAs revealed a complexity of approximately 105 PFU (plaqueforming units) with approximately 90% recombinants. Insert sizes ranged between 350 and 1000 bp. To test the quality of the fibroblast libraries, we screened for the presence of the hypoxanthine phosphoribosyltransferase (HPRT) transcript by PCR and Southern blotting. Following its identification, sequencing of the HPRT-positive clone (over 270 bp within the coding sequence, nt 210 to nt 480) showed that no PCRassociated base misincorporation had occurred in this gene during the construction of the library. We noted that the proportion of HPRT transcripts in the larger size ranges decreases proportionately with the number of starting cells (100, 50, and 10) used to synthesize oligo(dT)-primed cDNA (data not shown).
FIG. 2. Size distribution of amplified cDNA. (a) Approximately 500 ng of cDNA was electrophresed on a 1.2% agarose/11 TBE gel prior to ligation into Lambda ZAP II. (b) cDNA inserts released by PCR amplification of randomly selected plaques. The numbers to the left indicate the position of the DNA size standards (1-kb ladder, Gibco BRL).
Therefore, we might expect to generate smaller cDNAs in libraries from single embryos with some clones representing the 3* untranslated regions of transcripts. We considered an alternative approach to clone mainly coding regions, using random hexamers rather than oligo(dT) to prime the cDNA synthesis. However, although cDNA synthesized by random hexamers generally yields clones representing the coding sequence of genes, this approach is found to be suitable only when mRNA has been isolated and not when total RNA is being used as template, because contamination with genomic DNA and cDNA derived from ribosomal RNA is more frequent (Weng et al., 1989). After optimization of our procedures on cultured fibroblast cells, we proceeded to construct cDNA libraries from eight human oocytes and single four-cell-, seven-cell-, and blastocyst-stage embryos. The libraries have complexities of 2.0 1 106, 5.0 1 105, 3.5 1 106, and 2.5 1 106 clones, respectively. Figure 1 outlines the successive steps of the cDNA amplification and cloning procedure used to construct the libraries in Lambda ZAP II for the four-cell and seven-cell libraries or pBluescript (SK) for the oocytes and blastocyst libraries. The cDNA insert sizes were determined in two ways—by agarose gel electrophoresis prior to ligation into the vector and by PCR amplification of randomly picked plaques and colonies from the libraries using the KS and SK primers (see Table 1) flanking the unique EcoRI site present in Lambda ZAP II and pBluescript. The insert sizes clustered around 300 bp but included fragments of 500 bp and above (Fig. 2). Qualitative Analysis of the Libraries
FIG. 1. Schematic representation of PCR-based library construction.
AID
GENO 5117
/
6r54$$$521
12-09-97 20:27:27
The cDNA used for constructing the library was amplified by PCR, which might favor certain sequences over others. To test whether the libraries showed a proportional representation of transcribed genes in a human cell, we estimated the clone frequency of the cytoskeletal elements, b-actin, a-tubulin, and keratin18, and of the repetitive sequence, LINE-1. In addition, the clone frequency for 18S ribosomal RNA was assessed to determine the level of ribosomal RNA contamination. Clone frequencies were determined by ‘‘plaque lifts’’ and colony hybridization. The oligonucleotide and
gnmal
340
ADJAYE ET AL.
TABLE 1 Oligonucleotide and cDNA Probes Used in the Library Construction and Screening Primers/gene
Sequence (5*–3*)
Position
Reference
dT-EcoRI dC-EcoRI Keratin-18a Keratin-18b b-Actin a-Globin a-Tubulin KS SK 18S rRNA
GTCTCGAGCGGAATTCCGGCC(T)13 GATCGGAATTCCGG(C)13 CCCATCCCTGATCCAGCAG CTGAGGCATTAAGCCAGCAG CCCAAGTCCACACAAGGGGAGG Last exon Full-length clone TCGAGGTCGACGGTATC CGCTCTAGAAACTAGTGGATC
N/A N/A 6360–6393 6236–6256 1508–1529 598–833 1–1246 N/A N/A 940–1190
Kulesh and Oshima (1988) Kulesh and Oshima (1988) Ponte et al. (1984) Proudfoot and Maniatis (1980) Cowan et al. (1983) Stratagene Stratagene McCallum and Maden (1985)
cDNA probes used for these analysis are described in Table 1. All positive hybridizations were confirmed by plating out and rehybridization. Further confirmation of the identity of the clones was obtained by Southern
blotting and sequencing. The clone frequency of b-actin determined by screening 15,000 plaques was 0.05%. The clone frequencies of a-tubulin and keratin-18 determined by screening 30,000 plaques were in each case
TABLE 2 Features of Sequenced Clones and Results of BLAST Search Clone
Size (bp)
DNA homology (BLASTN)
Accession No.
1 2 3
153 200 192
4 5 6
138 163 300
7 8 9 10 11 12 13 14 15
103 230 261 250 361 231 94 231 160
16 17 18
143 142 98
19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
313 247 140 226 232 251 140 269 139 142 144 141 140 143 137 171 300
Human b-tubulin Novel Mouse embryonic carcinoma F9-cell cDNA EST Human fetal brain EST Novel Human fetal lung EST, similar to polyadenylate binding protein Human LINE-1a Human fetal brain EST Human placenta EST Human fetal liver EST Human FMR-1 Novel Novel c-erbB3-receptor tyrosine kinase c-AMP-dependent transcription factor ATF-4 Human keratinocyte EST Human skeletal muscle EST Lung carcinoma EST, similar to human erythrocyte adducin b-subunit Human Alu RNA transcript Novel Human skeletal muscle EST Human mitochondrial EST from skeletal muscle Novel Human 18S ribosomal RNAa Novel Novel Novel Human fetal lung EST Human breast basic conserved protein-1 Novel Human skeletal muscle EST Human fetal brain EST Human fetal brain EST Glucose-regulated protein-78 Novel
% Identity
Overlap
X00734
99
153
D28656 W36298
79 91
192 138
T62932
93
177
M80343 W36298 C17730 T87870 L29074
91 96 93 92 74
103 230 261 250 206
S61953
81
42
N73650 D29455 F17041 AA129349
89 85 96 86
120 143 142 53
M87914
82
287
F15854 F15772
83 92
100 105
X03205
62
251
AA402038 X64707
63 60
80 81
F17041 W36298 W36297 T27747
96 96 97 99
140 143 137 171
Note. Plasmid DNA was used as the template for cycle sequencing as described under Materials and Methods. Clones 1–6, 7–22, 23– 32, and 33–35 are derived from the oocyte-, 4-cell-, 7-cell-, and blastocyst-stage libraries, respectively. All clones have a tract of adenine residues present at the 3* end of the cDNA immediately 5* to the EcoRI site of pBluescript. a Clones 7 and 24 lack the corresponding polyadenylation signal (AAUAAA) sequence (Birnstiel et al., 1985) present upstream of the poly(A) tract.
AID
GENO 5117
/
6r54$$$521
12-09-97 20:27:27
gnmal
341
HUMAN PREIMPLANTATION EMBRYO LIBRARIES
TABLE 3 Compositions of cDNA Libraries as Determined by Random Clone Sequences No.
Gene category
1 2 3 4 5 6 7 8 9
Database match–Human Mitochondrial Repetitive elements Ribosomal RNA Brain Skeletal muscle Inducible genes Other nuclear genes Database match–Nonhuman Novel genes
Oocyte
GENO 5117
/
6r54$$$521
Seven-cell
Blastocyst
1
1 2 1
2 2
1 1 1
2 1 2
6
2
3
5
1
0.006%. LINE-1 repetitive sequences were present at a frequency of 0.6%, and ribosomal RNA clones were present at a frequency of 0.03% following the screening of 15,000 plaques in both cases. This level of contamination of the cDNA libraries with ribosomal RNA is acceptable because we are generating libraries with complexities on the order of 105 to 106 independent clones. The tissue-specific a-globin gene is also represented among the expressed sequences in the four-cell cDNA library at a frequency of 0.003%, determined by screening of 30,000 plaques. To isolate and identify other transcripts present in the oocyte, four-cell-, seven-cell-, and blastocyst-stage libraries, randomly picked consecutive clones of up to 400 bp were processed for sequencing as described under Materials and Methods. Analyses of sequence homologies involved the use of the Basic Local Alignment Search Tool (BLAST) algorithm (Altschul et al., 1990) to search the GenBank, EMBL, and dbEST databases. After the BLAST search, all the cDNAs were compared against GenBank by FASTA to determine whether significant matches were missed because of the use of BLASTN for the initial database search. The features and the results from the BLAST searches of 35 cDNA clones analyzed are listed in Table 2. The poly(A) signal (AAUAAA), which directs polyadenylation of mRNAs, is present in 33 of the 35 clones presented in Table 2. However, the absence of the poly(A) signal in the remaining two clones (7 and 24; Table 2) does not exclude mRNA-directed cDNA synthesis, because the poly(A) signal may be further 5* to the cDNA sequence obtained. The cDNAs listed in Table 2 show nucleotide sequence identity of 60 to 99% with genes that already appear in databases. On the basis of database searches the 35 cDNAs in Table 2 are classified into nine groups as shown in Table 3. Seven groups, derived from 23 of the cDNAs (65.7% of the total), consist of matches to the following human sequences: mitochondrial, repetitive elements (LINE-1 and Alu repeats), ribosomal RNA, brain, skeletal muscle, inducible genes, and other nuclear genes. Of these, one cDNA (clone 22; 2.9%) represents a mitochondrial gene, two cDNAs (clones 7 and 19; 5.7%) represent repetitive elements, one cDNA (clone 24; 2.9%) represents ribosomal RNA, five cDNAs (clones 4, 8, 11,
AID
Four-cell
12-09-97 20:27:27
1
32, and 33; 14.3%) represent brain genes, with clones 4, 8, and 32 matching the same EST (W36298). Three cDNAs (clones 17, 21, and 31; 8.6%) represent muscle genes, and one cDNA (clone 34; 2.9%) represents an inducible gene. Another cDNA clone (clone 3) of the 35 randomly picked cDNA sequences so far analyzed matched a mouse embryonic carcinoma EST, not previously shown to be present in human cDNA libraries. Eleven of the 35 cDNAs (clones 2, 5, 12, 13, 20, 23, 25, 26, 27, 30, and 35; 31.4%) had less than 10% sequence identity (insignificant matches) to both human and nonhuman entries in all the databases. These cDNAs are therefore regarded as potential novel genes. DISCUSSION
The construction and evaluation of PCR-based cDNA libraries from eight human oocytes and single human four-cell-, seven-cell-, and blastocyst-stage embryos are described. Cultured human fibroblasts were used initially to optimize the sensitivity of cDNA synthesis and establish efficient, reliable, and reproducible library construction from a few cells before the construction of the libraries from precious human preimplantation embryos was commenced. The measured complexities of the oocyte, four-cell-, seven-cell-, and blastocyst libraries were 2.0 1 106, 5.0 1 105, 3.5 1 106, and 2.5 1 106 clones, respectively, and indicate that the libraries are representative, given that the human genome is estimated to consist of 50,000 to 100,000 genes (Adams et al., 1991). The complexities of our libraries are comparable to those for human fetal tissues as reported by Buraczynska et al. (1995) and for 10 mouse oocytes as reported by Revel et al. (1995). This reproducibly high yield of clones is made possible by the single-tube procedure outlined in Fig. 1. Important aspects pertinent to this methodology of library construction are as follows. (i) The cloned sequences are short, reflecting the preference of PCR procedures for the amplification of smaller molecules. Nevertheless, these short 3*-UTR cDNA sequences act as an efficient source of ESTs (Adams et al., 1991), and they may be extended into the
gnmal
342
ADJAYE ET AL.
5*-coding regions by their use as primers in 5*-RACE (Frohman et al., 1988) with mRNA isolated from an appropriate tissue. (ii) Ribosomal and repetitive sequences are present. Because ribosomal RNA typically constitutes 80 to 85% of total RNA, it is important to avoid a high level of ribosomal cDNAs in the libraries. This was achieved by the use of oligo(dT) as primer in the reverse transcription reaction. We showed that the clone frequency of the human 18S ribosomal RNA (McCallum and Maden, 1985) is 0.03%, thus demonstrating that ribosomal DNA contamination is negligible in our libraries. Similarly, we have shown that the repetitive sequences do not saturate the libraries, in that LINE-1 clones (Dombroski et al., 1991) are present at a frequency of 0.6%. (iii) Genomic DNA contamination can be controlled and is kept at a very low frequency by the use of the oligo(dT) primer to select mRNA for first-strand cDNA synthesis. Because 33 of 35 randomly sequenced clones contain the polyadenylation signal (AAUAAA), we can say that at least 94% of our clones are derived from mRNA. (iv) Carrier RNA in the form of E. coli or yeast ribosomal RNA is usually added to precipitations of cDNA or RNA when present in low concentrations (McConnel and Watson, 1986; Taylor and Piko, 1987; Temeles et al., 1994). However, our procedures do not include the use of exogenous carrier RNA, thus avoiding altogether the possibility of clones derived from carrier RNA. (v) The highly thermostable Pfu DNA polymerase, unlike Taq DNA polymerase, possesses 3* to 5* exonuclease proofreading activity that enables the polymerase to correct nucleotide-misincorporation errors, and thus our libraries will have fewer errors than those generated with Taq DNA polymerase.
onic development, as they appear in most, if not all, cells (Kazazian et al., 1988; Trelogan and Martin, 1995; Sinnett et al., 1992). The 313-bp Alu transcript derived from the four-cell library (clone 19, Table 2) shares 82% identity over an overlap of 287 bp with the published Alu sequence, suggesting that this clone could be a new member of the Alu subfamilies (Jurka and Milosavljevic, 1991). It is possible that transcription, as indicated by the presence of these repetitive sequences in the cDNA libraries, is related to instability of transposable elements in embryonic cells. The movement of these elements is medically significant; cases of hemophilia A, muscular dystrophy, breast cancer, and colon cancer have been associated with the insertion of the human L1 transposons (Kazazian et al., 1988). The finding that a large proportion of the embryo transcripts (14.3%) are homologous to brain transcripts is not surprising considering that, of the 50,000 to 100,000 genes expressed in the human genome, up to 30,000 are expected to be expressed in the brain (Adams et al., 1991). Indeed, up to 25% of all genetic diseases affect neurological functions. A high level of expression of brain transcripts in preimplantation embryos suggests that these transcripts are important during human development. For example, clone 11, derived from the fourcell library, shares a sequence identity of 74% over a 206-bp overlap with the human fragile X mental retardation (FMR1) gene (Verkerk et al., 1991, 1993). There are a number of transcripts suggestive of new members of gene families bearing homologies of 60 to 86% with entries in the databases (see Table 2). These genes may be similar, but not necessarily identical, to genes of known function in humans. The sequence homologies include proto-oncogene c-erbB3 (Katoh et al., 1993), ATF-4, a cyclic-AMP-dependent transcripWe have shown that the libraries are representative tion factor (Estes et al., 1995), and a mouse embryonal of the mRNA population of human cells. The clone fre- carcinoma EST (Nishiguchi et al., 1994). quency data show that actins are expressed more abunFinally, there are clones for which no significant simdantly than the keratins and tubulins, as expected, and ilarities are found in the databases. These are referred also correlate with the electron microscopy revelation to as ‘‘novel’’ genes (31.4% of total). Characterizations of microtubules, microfilaments, and 10-nm filaments of the full-length clones of these sequences and their (intermediate filaments) in unfertilized oocytes and temporal and spatial expression patterns in fetal develblastomeres of all cleavage stages (Coonen et al., 1993). opment are in progress. The assembly of identified genes (normally expressed in tissues as diverse as brain, liver, lung, placenta, Applications of the Stage-Specific cDNA Libraries pancreas, muscle, and heart) in the oocyte-, four-cell-, The inherent limitation of the amount of biological seven-cell-, and blastocyst-stage embryo libraries (Tamaterial has previously restricted the ability of investible 2) gives a clear indication of the complexity of the libraries. However, it is important to note that very gators to carry out research in gene expression during preimplantation development in humans. We are currare cDNAs may not be recovered in our libraries. Interesting findings emerging from the analysis of the rently using the embryonic cDNA libraries in several sequenced clones include evidence for expressed repeti- areas of research. tive sequences; an abundance of transcripts also found Zygotic gene activation. The earliest stages of develin brain; sequences with homologies to oncogenes, tran- opment in mammals are partly regulated by materscription factors, and stem-cell genes; and novel tran- nally inherited information, and the transition from scripts with no homology to sequences in the databases. maternal to embryonic control of development is of conThese categories are discussed further below. siderable interest. cDNAs from embryos of different Repetitive elements account for 5% of the human ge- stages coupled with subtractive hybridization (Lisitsyn nome. Transposition of LINE-1 and Alu sequences is et al., 1993) and/or an adaptation of differential display likely to have occurred during meiosis or in early embry- (Liang and Pardee, 1992; Zimmermann and Schultz,
AID
GENO 5117
/
6r54$$$521
12-09-97 20:27:27
gnmal
HUMAN PREIMPLANTATION EMBRYO LIBRARIES
1994) using the embryo libraries will allow identification of new stage-specific gene activation. Expression of ‘‘tissue-specific’’ CpG-island-containing genes. The conservation of CpG islands throughout evolution depends on these sequences remaining unmethylated, because methylated CpG sites gradually mutate to TpG (Cross and Bird, 1995). We have previously shown that CpG-island-containing tissuespecific genes are expressed during early human development, and we proposed that the expression of CpGisland genes at these stages prevents methylation and, hence, mutation of these sequences (Daniels et al., 1997a). One of the expressed genes is a-globin, a red blood cell-specific, CpG-island-containing gene. This tissue-specific gene was detected in the four-cell cDNA library at a frequency of 0.003%, consistent with the expression in the embryos themselves. Keratin-18, another tissue-specific gene, was also detected. X-inactivation and imprinting. The embryonic libraries, together with genomic libraries prepared from parental samples (cumulus cells and sperm) corresponding to the individual embryo generating the cDNA library, are being utilized in the study of X-inactivation and genomic imprinting. We have previously demonstrated that both parental XIST alleles are expressed in human preimplantation embryos (Daniels et al., 1997b), thus unlinking preferential parental allele expression with imprinted paternal X-inactivation in the human trophoblast (Goto et al., 1997). Following sexing of the embryonic libraries (Daniels et al., 1997b), maternal XIST allele expression can be confirmed in male embryo cDNA libraries. In addition, the libraries are being screened for imprinted (monoallelic) expression of known imprinted genes during preimplantation development. Following confirmation of the expression of the imprinted gene in the libraries, polymorphism(s) within the 3* end of the coding sequence or UTR of the gene is sought in the parental genomic DNA libraries for use as a marker(s) to distinguish between mRNA derived from each parental copy of the gene. In addition to their intrinsic academic interest, these studies will have medical implications. The understanding of gene function and developmental expression programs during preimplantation development in humans is a promising area for study, justified on both medical and scientific grounds. Critical genetic decisions (e.g., X-chromosome inactivation, imprinted gene activity), which irreversibly affect the genetic potential of the individual in later life, are made during these very early stages of development. The data we have presented clearly demonstrate that our embryonic libraries will be invaluable for these studies. Access to Libraries Queries regarding the use of these libraries should be directed to Dr James Adjaye.
AID
GENO 5117
/
6r54$$$521
12-09-97 20:27:27
343
ACKNOWLEDGMENTS The use of embryos is covered by HFEA License No. R0063 to Dr. Virginia Bolton. This work was supported by the Medical Research Council.
REFERENCES Adams, M. D., Kelley, J. M., Gocayne, J. D., Dubnick, M., Polymeropoulos, M. H., Xiao, H., Merril, C. R., Wu, A., Olde, B., Moreno, R. F., Kerlavage, A. R., McCombie, W. R., and Venter, J. C. (1991). Complementary DNA sequencing: Expressed sequence tags and human genome project. Science 252: 1651–1656. Adjaye, J., Marsh, P. J., and Eagles, P. A. M. (1993). Immunological properties and cDNA sequence analysis of an intermediate-filament-like protein from squid neuronal tissue. J. Cell Sci. 106: 1283–1290. Adjaye, J., Plessmann, U., Weber, K., and Dodemont, H. (1995). Characterisation of neurofilament protein NF70 messenger RNA from the gastropod Helix aspersa reveals that neuronal and nonneuronal intermediate filament proteins of cerebral ganglia arise from separate lamin-related genes. J. Cell Sci. 108: 3581–3590. Altschul, S. F., Gish, W., Miller, W., Myers, E. W., and Lipman, D. J. (1990). Basic local alignment search tool. J. Mol. Biol. 215: 403– 410. Beliavsky, A., Vinogradova, T., and Rajewsky, K. (1989). PCR-based cDNA library construction: General cDNA libraries at the level of a few cells. Nucleic Acids Res. 17: 2919–2932. Bevan, S. G., Southey, M. C., Armes, J. E., Venter, D. J., and Newgreen, D. F. (1996). Spatiotemporally exact cDNA libraries from quail embryos: A resource for studying neural crest development and neurocristopathies. Genomics 38: 206–214. Birnstiel, M. L., Busslinger, M., and Strub, K. (1985). Transcription termination and 3* processing: The end is in site. Cell 41: 349– 359. Brady, G., Barbara, M., and Iscove, N. (1990). Representative in vitro cDNA amplification from individual hematopoietic cells and colonies. Methods Mol. Cell. Biol. 2: 17–25. Buraczynska, M. J., Van Keuren, M. L., Buraczynska, K. M., Chang, Y. S., Crombez, E., and Kurnit, D. M. (1995). Construction of human embryonic cDNA libraries: HD, PKD1 and BRAC1 are transcribed widely during embryogenesis. Cytogenet. Cell Genet. 71: 197–202. Coonen, E., Dumoulin, J. C., and Ramaekers, F. C. (1993). Intermediate filament protein expression in early developmental stages of the mouse: A confocal scanning laser microscopy study of in vitro fertilized and in vivo cultured pre-implantation mouse embryos. Histochemistry 99(2): 141–149. Cowan, N. J., Dobner, P. R., Fuchs, E. V., and Cleveland, D. W. (1983). Expression of human alpha-tubulin genes: Interspecies conservation of 3* untranslated regions. Mol. Cell. Biol. 3: 1738– 1745. Cross, S. H., and Bird, A. P. (1995). CpG island and genes. Curr. Opin. Genet. Dev. 5: 309–314. Daniels, R., Lowell, S., Bolton, V., and Monk, M. (1997a). Expression of tissue-specific genes in human preimplantation embryos. Hum. Reprod. 12: 2251–2256. Daniels, R., Zuccotti, M., Kinis, T., Serhal, P., and Monk, M. (1997b). XIST expression in human oocytes and preimplantation embyos. Am. J. Hum. Genet. 61: 33–39. Dombroski, B. A., Mathias, S. L., Nanthakumar, E., Scott, A. F., and Kazazian, H. H., Jr. (1991). Isolation of an active human transposable element. Science 254: 1805–1808. Estes, S. D., Stoler, D. L., and Anderson, G. R. (1987). Normal fibroblasts induce the C/EBP beta and ATF-4 bZIP transcription factors in response to anoxia. Exp. Cell Res. 220: 47–54. Frohman, M. A., Dush, M. K., and Martin, G. R. (1988). Rapid production of full-length cDNAs from rare transcripts: Amplification
gnmal
344
ADJAYE ET AL.
using a single gene-specific oligonucleotide primer. Proc. Natl. Acad. Sci. USA 85: 8998–9002. Goto, T., Wright, E., and Monk, M. (1997). Paternal X-chromosome inactivation in human trophoblastic cells. Mol. Hum. Rep. 3: 77– 80. Harrison, S. M., Dunwoodie, S. L., Arkell, R. M., Lehrach, H., and Beddington R. S. P. (1995). Isolation of novel tissue-specific genes from cDNA libraries representing the individual tissue constituents of the gastrulating mouse embryo. Development 121: 2479– 2489. Hwang, S., Benjamin, L. E., Oh, B., Rothstein, J. L., Ackerman, S. L., Beddington, R. S. P., Solter, D., and Knowles, B. B. (1996). Genetic mapping and embryonic expression of a novel, maternally transcribed gene Mem3. Mamm. Genome 7: 586–590. Jay, P., Diriong, S., Taviaux, S., Roeckel, N., Mattei, M-G., Audit, M., Berge-Lefranc, B-L., and Berta, P. (1997). Isolation and regional mapping of cDNAs expressed during early human development. Genomics 39: 104–108. Jurka, J., and Milosavljevic, A. (1991). Reconstruction and analysis of human Alu genes. J. Mol. Evol. 32: 105–121. Katoh, M., Yazaki, Y., Sugimura, T., and Terada, M. (1993). c-erbB3 gene encodes secreted as well as transmembrane receptor tyrosine kinase. Biochem. Biophys. Res. Commun. 192(3): 1189–1197. Kazazian, H. H., Jr., Wong, C., Youssoufian, H., Scott, A. F., Philips, D. G., and Antonarakis, S. E. (1988). Nature 332: 164–166. Kulesh, D. A., and Oshima, R. G. (1988). Cloning of the human keratin 18 gene and its expression in nonepithelial mouse cells. Mol. Cell. Biol. 8: 1540–1550. Liang, P., and Pardee, A. B. (1992). Differential display of eucaryotic messenger RNA by means of the polymerase chain reaction. Science 257: 967–971. Lisitsyn, N. A., Lisitsyn, N. M., and Wigler, M. (1993). Cloning the difference between two complex genomes. Science 259: 946–951. McCallum, F. S., and Maden, B. E. (1985). Human 18S ribosomal RNA sequence inferred from DNA sequence. Variations in 18S sequences and secondary modification patterns between vertebrates. Biochem. J. 232(3): 725–733. McConnel, J., and Watson, C. J. (1986). Construction of a representative cDNA library from mRNA isolated from mouse oocytes. FEBS Lett. 195: 199–203. Nishiguchi, S., Joh, T., Horie, K., Zou, Z., Yasunaga, T., and Shimada, K. (1994). A survey of genes expressed in undifferentiated mouse embryonal carcinoma F9 cells: Characterisation of low-abundance mRNAs. J. Biochem. 116(1): 128–139. Ponte, P., Ng, S. Y., Engel, J., Gunning, P., and Kedes, L. (1984). Evolutionary conservation in the untranslated regions of actin mRNAs: DNA sequence of a human beta-actin cDNA. Nucleic Acids Res. 12: 1687–1696. Proudfoot, N. J., and Maniatis, T. (1980). The structure of a human alpha-globin pseudogene and its relationship to alpha-globin gene duplication. Cell 21: 537–544. Revel, F., Renard, J-P., and Duranthon, V. (1995). PCR-generated cDNA libraries from reduced numbers of mouse oocytes. Zygote 3: 241–250.
AID
GENO 5117
/
6r54$$$521
12-09-97 20:27:27
Rothstein, J. L., Johnson, D., DeLoia, J. A., Skowronski, J., Solter, D., and Knowles, B. (1992). Gene expression during preimplantation mouse development. Genes Dev. 6: 1190–1201. Saiki, R. K., Gelfand, D. H., Stoffel, S., Scharf, S. J., Higuchi, R., Horn, G. T., Mullis, K. B., and Erlich, H. A. (1988). Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 239: 487–491. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989). ‘‘Molecular Cloning: A Laboratory Manual,’’ 2nd ed., Cold Spring Harbor Laboratories Press, Cold Spring Harbor, NY. Sinnett, D., Richer, C., Deragon, J-M., and Labuda, D. (1992). Alu RNA transcripts in human embryonal carcinoma cells. J. Mol. Biol. 226: 689–706. Smith, D. E., and Gridley, T. (1992). Differential screening of a PCRgenerated mouse embryo cDNA library: Glucose transporters are differentially expressed in early postimplantation mouse embryos. Development 116: 555–561. Southern, E. M. (1975). Detection of sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98: 503–517. Taylor, K. D., and Piko, L. (1987). Pattern of mRNA prevalence and expression of B1 and B2 transcripts in early mouse embryos. Development 101: 877–892. Temeles, G. L., Ram, P. T., Rothstein, J. L., and Schultz, R. M. (1994). Expression patterns of novel genes during mouse preimplantation embryogenesis. Mol. Reprod. Dev. 37: 121–129. Trelogan, S. A., and Martin, S. L. (1995). Tightly regulated, developmentally specific expression of the first open reading frame from LINE-1 during mouse embryogenesis. Proc. Natl. Acad. Sci. USA 92: 1520–1524. Varmuza, S., and Tate, P. (1992). Isolation of epiblast-specific cDNA clones by differential hybridization with polymerase chain reaction-amplified probes derived from single embryos. Mol. Reprod. Dev. 32: 339–348. Verkerk, A. J., Pieretti, M., Sutcliffe, J. S., Fu, Y. H., Kuhl, D. P., Pizzuti, A., Reiner, O., Richards, S., Victoria, M. F., Zhang, F., Eussen, B., Van Ommen, G. J. B., Blonden, L. A. J., Riggins, G. J., Chastain, J. L., Kunst, C. B., Galjaard, H., Caskey, C. T., Nelson, D. L., Oostra, B. A., and Warren, S. T. (1991). Identification of a gene (FMR-1) containing a CGG repeat coincident with a breakpoint cluster region exhibiting length variation in fragile X syndrome. Cell 65: 905–914. Verkerk, A. J., de Graaff, E., de Boulle, K., Eichler, E. E., Konecki, D. S., Reyniers, E., Manca, A., Poustka, A., Willems, P. J., Nelson, D. L., and Oostra, B. A. (1993). Alternative splicing in the fragile X gene FMR-1. Hum. Mol. Genet. 2: 399–404. Welsh, J., Liu, J-P., and Efstratiadis, A. (1990). Cloning of PCRamplified total cDNA: Construction of a mouse oocyte cDNA library. Genet. Anal. Technol. Appl. 7: 5–17. Weng, D. E., Morgan, R. A., and Gearhart, J. D. (1989). Estimates of mRNA abundance in the mouse blastocyst based on cDNA library analysis. Mol. Reprod. Dev. 1: 233–241. Zimmermann, J. W., and Schultz, R. M. (1994). Analysis of gene expression in the preimplantation mouse embryo: Use of mRNA differential display. Proc. Natl. Acad. Sci. USA 91: 5456–5460.
gnmal
