Efficiency of polymerase chain reaction assay for cystic fibrosis in single human blastomeres according to the presence or absence of nuclei*

Transcription

1 FERTILITY AND STERILITY Copyright lei 1993 The American Fertility Society Vol. 59. No.4. April 1993 Printed on acid-free paper in U.S.A. Efficiency of polymerase chain reaction assay for cystic fibrosis in single human blastomeres according to the presence or absence of nuclei* Jiaen Liu, M.D. t Willy Lissens, Ph.D.:j: Paul Devroey, M.D., Ph.D.t Ingeborg Liebaers, M.D., Ph.D.:j: Andre C. Van Steirteghem, M.D., Ph.D.t University Hospital, Dutch-speaking Brussels Free University (Vrije Universiteit Brussel), Brussels, Belgium Objective: To amplify by polymerase chain reaction (PCR) assay the region of the most common mutation of cystic fibrosis (CF) in human blastomeres. Design: Blastomeres were isolated from two- to eight-cell tripronucleate embryos. The nuclear status of blastomeres was recorded by light microscopy (LM) and by fluorescence microscopy after vital labeling with the fluorochrome Hoechst (H-33342; Sigma, Brussels, Belgium). In each blastomere the region around the ~F508 mutation site was amplified by two PCRs with nested primers. Setting: Research units of the Centres for Reproductive Medicine and Medical Genetics of the Dutch-speaking Free University of Brussels, Belgium. Results: The presence of a nucleus by LM in 118 of 160 blastomeres was always confirmed by fluorescence microscopy, and in 10 additional blastomeres the nucleus was only visible by fluorescence microscopy. In the PCR assay all blanks were negative and in nucleate blastomeres (assessed by LM) the amplification rate was 96%. After staining with Hoechst dye the percentage of amplification was 71 % or 91 % if PCR was performed 2 hours or 20 hours after coloration. Conclusion: Efficient pre implantation diagnosis for CF on blastomeres requires assessment of nuclear status by LM and vital staining on blastomeres that are anucleate by LM. Fertil Steril 1993;59:815-9 Key Words: Preimplantation diagnosis, cystic fibrosis, blastomere, fluorochrome Hoechst 33342, polymerase chain reaction assay, ~F508 mutation, decontamination Preimplantation genetic diagnosis became possible because human embryos can be obtained in vitro, blastomeres can be isolated from, e.g., eightcell embryos, and polymerase chain reaction (PCR) may detect gene mutations on minute amounts of DNA. Preimplantation diagnosis has been applied Received September 1, 1992; revised and accepted December 7,1992. * Supported by Belgian Fund for Medical Research; grant numbers , , and , Brussels, Belgium. t Centre for Reproductive Medicine. t Department of Medical Genetics. Reprint requests: Willy Lissens, Ph.D., Department of Medical Genetics, University Hospital, Vrije Universiteit Brussel, Laarbeeklaan 101, B-1090 Brussels, Belgium. clinically for pregnancies at risk for X-linked diseases (1) and cystic fibrosis (CF) (2). Nevertheless, there is further need for preclinical basic research before this technology can be considered sufficiently accurate and efficient to advocate widespread clinical application (3). Cystic fibrosis is the most common lethal autosomal recessive disease among whites. The most frequent mutation ~F508 involves the deletion of three base pairs (bp). Recently, we reported efficient and accurate PCR conditions to assay the normal and ~F508 CF allele on single sperm cells from an heterozygous carrier (4). The aim of this study was to assess this PCR assay on human blastomeres obtained after in vitro culture of tripronucleate non-~f508 oocytes. The presence or Vol. 59, No.4, April 1993 Lin et al. per for cystic fibrosis in human blastomeres 815

2 absence of a nucleus in each blastomere was carefully examined before the PCR assay. MATERIALS AND METHODS Embryos Fifty-five embryos were donated for research since the day after insemination three pronuclei (PN) were clearly observed. The use of these tripronucleate human oocytes for research purposes was approved by the Institutional Ethical Committee. Forty-two ofthem were frozen at the 3PN stage with propanediol as a cryoprotectant. Fresh and frozenthawed tripronucleate oocytes were cocultured on Vero cells with B2 medium for approximately 48 hours at 37 C in an atmosphere with 5% CO 2, 5% O 2, and 90% N2 (5). After 2 days of coculture, the embryos had cleaved to the two- to eight-cell stage. Isolation of Blastomeres The zona pellucida (ZP) of the embryos was removed by incubation for 2 to 5 minutes in acidic Tyrode's solution (ph 2.6). The disappearance of the ZP was followed under the stereo microscope. The embryos without zona were then washed several times in M2 medium containing 0.5% bovine serum albumin (BSA) fraction V (Sigma A9647; Pasture, Brussels, Belgium). The zona-free embryos were incubated for 10 to 20 minutes in M2 medium without Ca++ and Mg++ to loosen the intercellular contacts between the blastomeres. Each single blastomere was then transferred to a 20-ttL droplet of M16 medium with 0.5% BSA (6). Nuclear Status of the Single Blastomeres The presence or absence of a nucleus in each blastomere was examined under an Olympus inverted light microscope (LM, Olympus 1M; Olympus Optical Co., Ltd., Tokyo, Japan) at a magnification of 300. The nuclei of the blastomeres were also vitally labeled with the fluorochrome benzimidazole compound H and observed under the Olympus BH2 microscope (Olympus BH2; Olympus Optical Co., Ltd., Tokyo, Japan). A stock solution containing 50 ttg/ml ofh was made in M2 medium. Each blastomere was incubated for 10 minutes in a 30- to 50-ttL droplet of M2 medium with H The blastomere was then washed several times in M2 medium and transferred to the depressed well of a 76 X 26-mm glass slide containing 30 ttl of M2 with 0.5% BSA. The blastomeres were examined by flu816 Liu et al. per for cystic fibrosis in human blastomeres orescence microscopy for the presence of a nucleus using ultraviolet excitation filters (magnification X200). The presence of a nucleus is evidenced by a bright fluorescent spot in the blastomere (Fig. 1). The absence of fluorescence after labeling with H33342 is interpreted as a sign of an anucleate blastomere. After the nuclear assessment, the blastomeres were washed three to five times in M2 medium and then transferred to 25-ttL droplets of M16 medium with 0.5% BSA covered with lightweight paraffin oil and incubated at 37 C in 5% CO 2 in air for either 2 hours or 20 hours before PCR. Blastomere Preparation for PCR The blastomeres were rinsed several times in M2 medium without BSA and then transferred to a PCR tube. The type of PCR tube was dependent on the model of PCR equipment used: [1] microtubes (72699; Sarstedt, Niimbrecht, Germany) for the DNA Thermal Cycler (Perkin Elmer Cetus, Norwalk, CT) or [2] MicroAmp reaction tubes ( ; Perkin Elmer Cetus) for the Gene Amp PCR System Under close visual control a blastomere was transferred with a mouth-controlled finely drawn pipette into a PCR microtube containing 20 ttl distilled water; to prevent evaporation, 100 ttl of sterile light-weight paraffin oil (29436 SH; British Drug House, Dorset, United Kingdom) was layered above the water. When the MicroAmp tubes were used, the water was covered with 50 ttl paraffin oil. The last M2 washing droplet of each blastomere was used as a blank. The tubes were stored at -20 C or at -80 C until the time of PCR assay. Polymerase Chain Reaction Assay for the Region of CF Mutation af508 The frozen PCR tubes with a blastomere or a washing droplet were kept at 96 C for 15 minutes Figure 1 Fluorochrome staining of isolated blastomere with (right) or without nucleus (left) (original magnification was XlOO). Fertility and Sterility

3 in the PCR machine, rapidly cooled in ice, and then Table 1 Comparison Between LM and Hoechst Staining for the Presence of Nuclei in Single Human Blastomeres further processed. Complete PCR mixtures without DNA template were prepared and decontaminated No. of nucleate blastomeres found by incubation with 2 U of restriction enzyme Hinfl No. of (Amersham, Aylesbury, United Kingdom) at 37 C Embryo blastomeres LM Coloration for 3 hours; the restriction enzyme was then inac (100)' 2 (100) tivated by incubation at 65 C for 20 minutes. The (75) 3 (75) region around the LlF508 mutation site was amplified (100) 4 (100) (100) 4 (100) by two consecutive PCRs with nested primers. The (60) 3 (60) two PCRs and the analysis of the PCR products by (100) 5 (100) agarose and polyacrylamide gel electrophoreses were (57) 4 (57) (67) 5 (83) carried out according to the previously described (75) 3 (75) procedure (4) (100) 2 (100) (100) 3 (100) Statistics (100) 3 (100) (60) 4 (80) A paired t-test was done to compare the obser (60) 3 (60) (100) 2 (100) vation of nuclei in blastomeres by LM or vital la (100) 4 (100) beling with H The efficiency of DNA ampli (50) 2 (50) fication with or without H coloration was (50) 6 (75) (100) 2 (100) compared by means of a two-by-two comparison (50) 1 (50) with the two-tailed X2 test. Probabilities < 5% were (50) 1 (50) considered significant (75) 3 (75) (83) 6 (100) RESULTS (100) 6 (100) (86) 7 (100) (100) 4 (100) One hundred sixty blastomeres were obtained (71) 6 (86) (50) 3 (50) from 35 two- to eight-cell embryos that were gen (50) 4 (67) erated after coculture on Vero cells of 35 tripronu (100) 3 (100) cleated fresh or frozen-thawed human oocytes. The (75) 3 (75) (75) 4 (100) presence or absence of a nucleus in each of these (40) 2 (40) 160 blastomeres was recorded by LM and by fluo (71) 5 (71) rescence microscopy after vital labeling with the (63) 6 (75) Total: DNA-binding fluorochrome H (Table 1). A ± 20.6t ± 19.8t:j: clear and distinct nucleus was observed by LM in 77% of 160 blastomeres; the presence of a nucleus ' Values in parentheses are percents. t Mean percentage ± SD. in these 118 blastomeres was always confirmed by :j: There is a significant difference between LM and coloration a positive signal after coloration with the benzimi- (P < 0.01). dazole derivative. As indicated in Table 1, 10 additional nucleated blastomeres were observed after coloration. The percentage of positive nuclei after amplification. If the blastomeres were colored with Hoechst labeling is significantly higher than after Hoechst dye, the amplification rate was higher when, light microscopical observation. When left in culture after the staining procedure, the blastomeres were for 20 hours after the staining with Hoechst dye, kept in culture for 20 hours instead of 2 hours before only 12 of 87 nucleate blastomeres divided once PCR (91% amplification versus 73%, P < 0.05). or twice. These divided blastomeres were not used None of the anucleate blastomeres gave a positive for PCR. amplification signal if the anucleate status had been The results of the PCR assays for amplification confirmed by Hoechst dye. The blastomeres of the of the region of the most common CF mutation in remaining 20 embryos were only analyzed by LM isolated blastomeres is summarized in Table 2. The for the presence of nuclei. In 96 nucleate blastodecontamination procedure was efficient because meres, 92 (96%) were positive by PCR, and in 3 of none of the 274 blanks displayed an amplification 30 anucleate embryos a positive band was observed. signal. There was a clear relation between the pres- The latter finding confirms the 10% false-negative ence of a nucleus in the blastomere and the PCR anucleate blastomeres by LM. The positive ampli- Vol. 59, No.4, April 1993 Liu et al. per for cystic fibrosis in human blastomeres 817

4 Table 2 Efficiency of DNA Amplification in Single Blastomeres Stained blastomeres Non-stained blastomeres Nucleate Embryos Blastomeres Group I* Group IIt Anucleate Embryos Blastomeres Nucleate Anucleate Blanks (n = 35) (n = 160) (n = 41) (n = 75) (n = 32) (n = 20) (n = 126) (n = 96) (n = 30) (n = 274) Positive amplification 30 (73) [30/41] :\: 68 (91) [68/75] * per was performed 2 hours after coloration. t per was performed 20 hours after coloration. :\: Values are numbers with positive amplification with percents in parentheses and proportions in brackets. fication rate in unstained blastomeres was significantly higher than in blastomeres where PCR was done 2 hours after the Hoechst coloration. Amplification rates were not different if PCR was done without staining or 20 hours after coloration. DISCUSSION A possible approach for preimplantation diagnosis is to remove two blastomeres from eight-cell embryos (1, 7). The embryo can be discarded, if affected, or replaced in the uterus, if not affected. DNA amplification from a single blastomere is fraught with false-positive or false-negative results (2, 3, 8). An absolute prerequisite for DNA amplification in a single cell is the presence of a nucleus. Recent studies have reported that anucleate blastomeres may be present in apparently normal cleaving embryos after IVF (9). Therefore, we decided to evaluate the nuclear status of blastomeres isolated from two- to eight-cell embryos obtained after in vitro culture of Figure 2 Isolated blastomere inside per tube (original magnification was X100). The insert depicts a magnification X300 of the blastomere. 818 Liu et al. per for cystic fibrosis in human blastomeres 0(0) [0/32] 92 (96) [92/96]11 3 (10) [3/30] 0(0) [0/274] Group I is significantly different from group II (P < 0.05). II The amplification of unstained nucleate blastomeres is significantly higher than in group I of the colored blastomeres (P < 0.05). tripronucleate oocytes donated for research. The combination of LM and vital labeling with Hoechst dye indicated that in 82% of 160 blastomeres a nucleus was present and that in 5% ofthese the nucleus was only visible after coloration with H The PCR assay for amplification of the LlF508 mutation of CF includes a decontamination step with the restriction enzyme Hinfi. As reported earlier for single sperm cells (4), the decontamination is very efficient because none of the 274 blanks had any sign of amplification. The overall amplification rate of nucleate blastomeres was 89.6% (190 of 212 blastomeres). Staining with the dye decreases the PCR amplification efficiency especially if the procedure is done 2 hours after coloration. Indeed, the efficiency reaches 91 % when PCR is done 20 hours after staining and 96% in blastomeres found to be nucleate after LM only. These amplification rates are lower than our results on single and two sperm cells (4) but are similar to the 87% efficiency reported by Verlinsky et al. (2) on 15 blastomeres and polar bodies. Furthermore, these amplification rates compare favorably with the 45% amplification rate for the sickle-cell containing region of the,b-globin gene in single human blastomeres (10). We do not think that loss of blastomeres are causing amplification failure because in the MicroAmp reaction tubes the presence of a blastomere can be confirmed under the inverted microscope (Fig. 2). This study enables us to recommend that biopsied blastomeres be carefully examined for the presence of a nucleus by LM and that vital staining be done only on blastomeres that are anucleate by LM. The stained blastomeres have to be incubated for 20 hours before the DNA analysis. Acknowledgments. We thank the scientific and technical staff of the Centre for Reproductive Medicine and the Department of Fertility and Sterility

5 Genetics for expert technical assistance; Mrs. Cecile Janssenswillen for the coculture; Mrs. Ria Maes and Mr. Etienne Van den Abbeel for freezing and thawing the embryos; Mrs. Viviane De Wolf for typing the manuscript; Mr. Michael Whitburn of the Language Centre for proofreading the text in English; and Mrs. Marie-Paule Derde, Ph.D., for providing statistical advice. REFERENCES 1. Handyside AH, Kontogianni EH, Hardy K, Winston RML. Pregnancies from biopsied human preimplantation embryos sexed by V-specific DNA amplification. Nature 1990;344: Verlinsky Y, Rechitsky S, Evsikov S, White M, Cieslak J, Lifchez A, et al. Preconception and preimplantation diagnosis for cystic fibrosis. Prenat Diagn 1992;12: Trounson AL. Preimplantation genetic diagnosis-counting chickens before they hatch? Hum Reprod 1992;7: Liu J, Lissens W, Devroey P, Van Steirteghem A, Liebaers I. Efficiency and accuracy of polymerase-chain -reaction assay for cystic fibrosis allele tj.f508 in single cell. Lancet 1992;339: Menezo Y, Guerin J-F, Czyba J-C. Improvement of human early embryo development in vitro by coculture on monolayers of vero cells. Bioi Reprod 1990;42: Hogan B, Costantini F, Lacy E. Manipulating the mouse embryo, a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, Hardy K, Martin KL, Leese HJ, Winston RML. Human preimplantation development in vitro is not adversely affected by biopsy at eight-cell stage. Hum Reprod 1990;5: Navidi W, Arnheim N. Using PCR in preimplantation genetic disease diagnosis. Hum Reprod 1991;6: Winston NJ, Braude P, Pickering SJ, George MA, Cant A, Johnson M. The incidence of abnormal morphology and nucleocytoplasmic ratios in 2-, 3-, and 5-day human pre-embryos. Hum Reprod 1991;6: Pickering SJ, McConnell JM, Johnson MH, Braude PR. Reliability of detection by polymerase chain reaction of the sickle cell-containing region of the f)-globin gene in single human blastomeres. Hum Reprod 1992;7: Vol. 59, No.4, April 1993 Liu et a1. per for cystic fibrosis in human blastomeres 819