Download Increasing the denaturation temperature during the first cycles of

Molecular Human Reproduction vol.2 no.3 pp. 213-218, 1996
Increasing the denaturation temperature during the first cycles of
amplification reduces allele dropout from single cells for
preimplantation genetic diagnosis
Pierre F.Ray1 and Alan H.Handyside
Human Embryology Laboratory, Institute of Obstetrics and Gynaecology, Royal Postgraduate Medical School,
Hammersmith Hospital, Du Cane Road, London W12 ONN, UK
^o whom correspondence should be addressed
Single cell polymerase chain reaction (PCR) for preimplantation genetic diagnosis (PGD) needs to be highly
efficient and accurate. In some single cells from human embryos presumed to be heterozygous for the AF508
deletion causing cystic fibrosis (CF), we recently observed random amplification failure of one of the two
parental alleles following nested PCR. To investigate allele dropout (ADO), we have examined two different
lysis protocols and the effect of altering the denaturation temperature in the primary PCR using single
lymphocytes heterozygous for AF508 or for two fi-thalassaemia mutations IVS 1 nt 1 (G/T) and 5 (G/C) using
a nested PCR protocol to amplify the 5' region of the p*-globin gene. Amplification rates were high after lysis
in either water or lysis buffer and at all denaturation temperatures studied (> 92%). With a typical denaturation
temperature (93°C), ADO was detected at both loci. When the denaturation temperature was lowered to
90°C, however, ADO increased substantially and conversely by raising the denaturation temperature to 96°C
during the first 10 cycles ADO was reduced but not eliminated. ADO was also reduced with cells in lysis
buffer. We suggest that ADO may be caused by a combination of inefficient denaturation and degradation of
one of the genomic alleles in the first cycles of PCR. For autosomal recessive conditions in which both parents
are carrying the same mutation, ADO would not cause serious misdiagnosis. For compound heterozygotes
or autosomal dominant conditions, however, extensive testing of the amplification protocol with single
heterozygous cells and individual calibration of each thermocycler for the effect of denaturation temperature
on ADO is essential before clinical application.
Key words: allele dropout/polymerase chain reaction/preimplantation genetic diagnosis/single cell
Introduction
DNA amplification from single cells, typically using nested
polymerase chain reaction (PCR), has been applied widely
for fine genetic mapping over short distances using single
spermatozoa (Li et al, 1988; Boehnke et al, 1989), the
detection of low levels of viral or bacterial pathogens (Bej
and Mahbubani, 1992; Zimmerman et al, 1993) and for
examining clonal genetic alterations in tumour progression
(Trumper et al, 1993). AJI important clinical application has
been for preimplantation genetic diagnosis (PGD) in which
single cells removed from human preimplantation embryos
following in-vitro fertilization (TVF) are analysed for a specific
genetic defect causing an inherited disease (Coutelle et al,
1989; Holding and Monk, 1989). Embryos are biopsied early
on the third day post-insemination (day 3) at the 6-10 cell
stage and nested PCR used for genetic analysis of one or
two single cells. Unaffected embryos are then selected and
transferred to the uterus later the same day. This allows couples
at risk of transmitting a particular gene defect to avoid
the possibility of a termination following diagnosis later in
pregnancy after amniocentesis in the second or chorion villus
sampling (CVS) in the first trimester of pregnancy
(Handyside, 1993).
For PGD, single cell PCR needs to be highly efficient and
© European Society for Human Reproduction and Embryology
allow accurate detection of the genetic defect causing the
disease. With cystic fibrosis (CF), several nested PCR protocols
have been reported for amplification of regions of exon 10 of
cystic fibrosis transmembrane regulatory (CFTR) gene from
single cells for detection of the predominant AF508 deletion
(Strom et al., 1990; Lesko et al, 1991; Liu et al, 1992). Our
protocol amplifies the whole of exon 10 with high efficiency,
and detection of the AF5O8 deletion by heteroduplex formation
was completely accurate in a series of single cells and
blastomeres disaggregated from cleavage stage embryos (Lesko
et al, 1991). A singleton pregnancy was established after PGD
and the baby girl confirmed at birth to be unaffected (Handyside
et al, 1992). Since then four more pregnancies and singleton
births, all confirmed unaffected, have been achieved in a total
of 16 PGD cycles (Ao et al, 1996).
To confirm the diagnoses in these later PGD cycles and to
investigate their accuracy, blastomeres from human embryos
which were not transferred were disaggregated and each single
cell amplified (P.Ray, R.M.L.Winston and A.H.Handyside, in
preparation). With presumed homozygous normal or affected
embryos the accuracy of the diagnosis was high. In only four
out of 366 cells would a serious misdiagnosis have occurred
because of a homozygous normal or heterozygous carrier
diagnosis in an otherwise uniformly homozygous affected
213
P.F.Ray and A.H.Handyside
embryo (presumably as a result of contamination). More
striking, however, was the relatively frequent failure to amplify
either the normal (8%) or affected (10%) parental allele,
apparently randomly, in blastomeres from heterozygous carrier
embryos. In these cases, where both partners carry the AF508
mutation, this random failure of amplification of one of the
alleles, allele dropout (ADO), would not result in a serious
misdiagnosis since CF is autosomal recessive. For compound
heterozygotes or autosomal dominant conditions, however, this
could result in the transfer of an affected embryo.
Although preferential amplification of a particular allele
sometimes occurs with conventional PCR, often because of
size differences in the amplified fragments (Walsh et al., 1992),
random ADO has not been reported with non-limiting target
DNA. It seems likely, therefore, that ADO is specific for single
cell PCR or at least very low target copy number. With only
two target molecules present in a single diploid cell, ADO
could arise from strand breaks or other DNA damage during
preparation. Alternatively, access to the target genomic
sequence by the primers and Taq polymerase may be restricted
by, for example, adjacent G/C rich regions reducing denaturation efficiency or different degrees of folding perhaps related
to the stage of the cell cycle. In any case, ADO is most likely
to arise in the initial cycles of the primary PCR before the
number of target molecules is increased by the process of
amplification.
To investigate the phenomenon of ADO and to distinguish
between these possibilities, the effects of different lysis conditions and denaturation temperatures during the first few cycles
of primary amplification with the outer primers have been
examined using single heterozygous lymphocytes. To eliminate
locus-specific effects, both nested PCR for AF508 detection
and nested PCR for amplification of part of exon 1 and the 5'
end of the intervening sequence 1 (IVS 1) of the p*-globin
gene were tested (P.F.Ray, J.S.Kaeda, J.Bingham, T.Vulliamy,
I.Dokal, I.Roberts, R.M.L.Winston and A.H.Handyside, in
preparation), in this case, with single lymphocytes from a child
affected with P-thalassaemia major and heterozygous for both
mutations IVS 1 nt 1 (G/T) and 5 (G/C).
Materials and methods
Preparation of single lymphocytes
Unclotted blood (3 ml) in EDTA was gently pipetted onto a layer
of Ficoll-paque (Pharmacia, St Albans, Herts., UK), lymphocytes
separated by centrifugation and washed twice in phosphate-buffered
saline (PBS) supplemented with 10 mg/ml bovine serum albumin
(BSA; Sigma, Poole, UK). Using a mouthtube, a finely pulled glass
pipette and a dissecting microscope, lymphocytes were loaded from
a dilute suspension in PBS + BSA and single lymphocytes distributed
to microdrops of the same medium under silicon oil. After checking
that a single lymphocyte had been transferred correctly to each
microdrop, each of them was transferred to 0.5 ml thin-walled
microcentrifuge tubes, loaded with either 10 ul double distilled
water (ddH2O) or 5 uJ of lysis buffer (200 mM KOH, 50 mM DTT)
(Li et al., 1991), again using a mouth controlled pipette. Samples
in ddH2O were frozen to -20°C and prior to PCR, thawed and
denatured for 5 min at 95°C. Samples in lysis buffer were either
214
stored at -20°C or used immediately by heating to 65°C for 10 min.
The alkaline lysis buffer was subsequently neutralized by adding
modified PCR buffer containing 5 ul of neutralizing buffer (900 mM
Tris HC1, pH 8.3, 300 mM KC1, 200 mM HC1). To control for
contamination, media from drops that had contained cells were
also sampled.
Nested polymerase chain reaction (PCR)
Primers for the CF nested PCR were as previously published, resulting
in an amplified fragment of 154 or 151 base pairs (bp) depending on
the presence or absence of the AF508 deletion (Handyside et al.,
1992). Conditions for the denaturation temperature with the outer
primers were modified as follows: initial denaturation at 95°C for 3
min then denaturation at 90, 93 or 96 (first 10 cycles) + 94°C (12
cycles) (to avoid excessive denaturation of Taq polymerase).
Primers for amplification of the (J-globin sequence were as follows:
Outer primers: 5'-TAAGCCAGTGCCAGAAGAGCC
5' -GCA ATC ATTCGTCTGTTTCCC A
Inner primers: 5'-GGCATATAAGTCAGGGCAGAG
5'-CCTTAAACCTGTCTTGTAACCTTGGTA
IVS 1 nt 1 (G/T) nullifies a natural BsiYl restriction site. The reverse
inner primer has been modified to create a Kpnl site at IVS 1 nt 5
when amplifying from normal DNA which is not present if the IVS
1 nt 5 (G/C) mutation is present.
Cycling conditions were: (i) initial denaturation at 95°C for 3 min;
(ii) 22 cycles with the outer primers, denaturation at 90, 93 or 96
(first 10 cycles) + 94°C (12 cycles) for 45 s, annealing at 55°C for
45 s and extension at 72°C for 90 s; and (iii) 2 \i\ of the outer
reaction product amplified with the inner primers for 30 cycles after
an initial denaturation at 95°C for 3 min, denaturation at 94°C for
45 s, annealing at 56°C for 45 s and extension at 72°C for 60 s.
Reaction volumes were 50 u.1 overlaid with 50 u.1 of mineral oil.
Outer amplification mix contained 5 |il of lysis buffer (200 mM
KOH, 50 mM DTT), 5 ul neutralizing buffer (900 mM Tris HC1,
pH 8.3, 300 mM KC1, 200 mM HC1) and potassium-free 10XPCR
buffer (25 mM MgCl2, 1 mg/ml gelatine, 100 mM Tris HC1, pH 8.3),
0.8 U.M of each primer, 0.2 mM of each nucleotide and 1 IU Taq
DNA polymerase (AmpliTaq; Perkin Elmer, Warrington, Cheshire,
UK). Inner amplification mix contained 0.8 U.M of each inner primer,
0.2 mM of each nucleotide, 1 IU of AmpliTaq (Perkin Elmer), 5 ul
of 1 OX PCR buffer II (Perkin Elmer) and 3 ul of 25 mM MgCl2
(Perkin Elmer).
Stringent precautions were taken to minimize contamination. Gloves
and dedicated gowns were worn during the preparation of single cells
and setting up the primary PCR. All reagents were filtered with
0.22 u.m filters and aliquoted in small volumes. Separate sets of
micropipettors were used for the two amplification steps and filtered
pipette tips were used throughout. PCR mix and aliquots were
prepared in a class II flow cabinet located in a first 'clean lab'. The
product of the first PCR was transferred to the secondary PCR
mix in a separate air flow cabinet located in a different lab. Gel
electrophoresis was performed in a third area located in a different
building. No amplified product was brought back to the first lab.
Gel electrophoresis and mutation detection
The amplification products were run on 10% polyacrylamide gels
stained with ethidium bromide and viewed with a UV transilluminator.
Amplification of both the normal and AF5O8 alleles from the
single heterozygous lymphocytes was detected by the presence of a
heteroduplex band (Figure 1). Samples with no heteroduplex band
were re-analysed by adding amplified DNA of known genotype to
determine which allele had been amplified (Handyside et al., 1992).
Reduced allele dropout in single cell PCR
Denaturation Temperature and Allele Drop-Oui
Dt°
M 1 2 3 4 5 6 7 8 9 10 B B
Heteroduplex
Homoduplex
90°C
93°C
96°C
Figure 1. Polyacrylamide gel electrophoresis of polymerase chain reaction (PCR) product after amplification of single blastomeres
heterozygous at the AF508 locus at three different denaturation temperatures. The absence of the heteroduplex band is indicative of allele
dropout (ADO). M = molecular marker (pBR322 Haelll digested); 1-10 = single lymphocyte; B = blank control.
(a)
c
o
(b)
100 1
100'
801
SO-
0
60-
SO
1
40-
|
"5.
a
I
20-
28
20-
O
Q
<
90
93
96
90
93
96
90
Denaturation temperature (°C)
Figure 2. Amplification rate ( - • - ) and frequency of allelic dropout (ADO) (-D-) at denaturation temperatures of 90, 93 and 96°C. Panel
(a) with the cystic fibrosis primers and water lysis; panel (b) with CF primers and alkaline lysis; and panel (c) fi-globin primers and lysis
buffer. Error bars are SEM.
Heteroduplex
Homoduplex
Figure 3. Polyacrylamide gel electrophoresis of PCR product after amplification with the inner primers of different ratios of CF AF508
normal and mutant alleles amplified with the outer primers. M = 100 bp molecular marker, tracks 1 = 1:1 dilution of normal/mutant; tracks
2 = 1:2 dilution; tracks 3 = 1:4; tracks 4 = 1:16; track 5 = 1:256.
For detection of the IVS 1 mutations in the (3-globin amplified
products, 8 u.1 aliquots were digested for 16 h with either 10 IU of
BsiYl (Boehringer Mannheim UK Ltd., Lewes, East Sussex, UK) at
55°C for the detection of the ntl mutation or 10 IU of Kpnl
(Pharmacia) at 37°C for the detection of the nt 5 mutation. The base
pair substitution at IVS I nt 1 (G/T) nullifies a natural BsiYl
215
P.F.Ray and A.H.Handyside
Table I. Amplification rate (Amp.) and frequency of allelic dropout (ADO) with the cystic fibrosis or P-globin primers at different denaturation temperature,
using a water or alkaline lysis protocol
Denaturation
temperature
90
93
96
Cystic fibrosis (water protocol)
50
62
78
Cystic fibrosis (lysis buffer)
Amp. (%)
ADO(%)
92
97
94
70
44
14
40
40
40
fi-globin lysis buffer
Amp. (%)
ADO(%)
95
97
97
21
13
5
30
32
26
Amp. (%)
ADO (%)
93
94
92
46
33
4
n = number of single lymphocytes analysed.
restriction site, and a mismatch in the reverse inner primer was
deliberately introduced to create a new Kpnl site when amplifying
the normal nt 5 allele which is abolished by the IVS 1 nt 5 (G/C)
mutation. Homozygous normal cells show a single band reduced in
size to 173 bp, or 184 bp when restricted with BsiYl or Kpnl,
respectively. Heterozygous cells give rise to both undigested (208 bp)
and digested bands (173/184 bp). Products of known genotype were
digested at the same time to control for partial digestion. Analysis
of both mutations in the same amplified product also acted as an
internal control.
Definitions
Amplification rate or efficiency was calculated as the proportion of
single cells from which at least one allele amplified.
The rate or frequency of ADO was calculated as the number
of samples in which only one allele amplified divided by the total
number in which one or both allelcs amplified.
Similarly, with the P-globin primers, nine cells amplified the
normal allele for IVS 1 nt 1 (and therefore the affected allele for
IVS 1 nt 5) against 14 which amplified the affected allele at
IVS 1 nt 1 (normal at IVS 1 nt 5).
To test the possibility that ADO is not the result of complete
amplification failure but reduced amplification of one allele to
a level below the threshold for detection, we amplified mixtures
of different proportions of the outer amplification products
from homozygous normal and affected cells (AF508) (Figure
3). With increasing ratios from 1:1 to 1:256 of normal:affected
outer amplification product the intensity of the heteroduplex
band decreased markedly. However, even at the lowest ratio
tested of 1:256 a very faint heteroduplex band could still be
detected by careful inspection of the gel.
Discussion
Results
A total of 310 single lymphocytes heterozygous for AF508
(190 lysed in ddH2O and 120 in alkaline lysis buffer) and
88 single lymphocytes heterozygous for mutations at both IVS
1 nt 1 and 5 of the p-globin gene were analysed. The efficiency
of amplification and frequency of ADO under the different
conditions tested are shown in Table I and Figure 2.
Amplification rates were high after lysis in either ddH2O
or lysis buffer and at all denaturation temperatures studied (5»
92%). With a typical denaturation temperature (93°C) during
the primary PCR, ADO was detected at both loci and ranged
between 13-44%. When the denaturation temperature was
lowered to 90°C, ADO increased substantially (21-70%).
Conversely at 96°C during the first 10 cycles, ADO was
reduced but not eliminated (4-14%). ADO was also reduced
by use of a lysis buffer irrespective of the denaturation
temperature used. Contamination occurred in eight out of 56
(14%) control blanks analysed for AF508. As all the cells
amplified here are heterozygotes, contamination can only result
in a small reduction in the occurrence of ADO. Therefore, we
do not think it compromises the results presented here. With
the P-globin primers, 0/24 blanks amplified. Typical amplifications with CF primers at the three denaturation temperatures
tested are shown in Figure 1.
ADO was not allele-specific at any of the denaturation
temperatures tested. After amplification with the CF primers
36 cells had amplified only the normal allele and 31 only the
affected allele (16716 at 90°C, 15/10 at 93°C and 5/5 at
96°C had amplified the normal/affected allele respectively).
216
This study clearly demonstrates that random failure of amplification of one parental allele, allele dropout (ADO), can
occur following nested PCR with some single heterozygous
lymphocytes with both sets of primers for amplification of
exon 10 of the CFTR gene and detection of the AF508 deletion
causing CF, and primers for amplification of part of the Pglobin gene and detection of IVS 1 nt 1 (G/T) and nt 5 (G/C)
mutations causing P-thalassaemia.
There was a strong correlation between the rate of ADO
and the denaturation temperature used during the first few
cycles of amplification although the overall amplificate rate
was not affected (Figure 2). The use of more stringent
lysis conditions, using a lysis buffer originally developed for
amplification from single sperm (Li et al., 1991), also decreased
the occurrence of ADO at all three denaturation temperatures
tested. These two points stress the need for efficient separation
of the native double-stranded DNA to allow adequate amplification of both alleles.
ADO was first detected following PGD in a series of couples
at risk of having children with CF because both partners carry
the AF508 deletion (Ray et al, 1994, 1995). The proportion
of heterozygous carrier embryos diagnosed in these cases, on
the basis of analysis of one or two biopsied blastomeres, was
lower than the expected ratio of 1:2:1 with homozygous normal
and affected embryos. We therefore disaggregated the embryos
which were not transferred to confirm the diagnosis and
analyse multiple single cells from each embryo. This demonstrated that, although single cell analysis of homozygous
normal or affected embryos was accurate and highly consistent,
Reduced allele dropout in single cell PCR
in carrier embryos only one parental allele amplified in about
20% of cells, equally distributed between the parental alleles.
With blastomeres, however, it is known that amplification
efficiencies tend to be lower (Pickering et al, 1992) and some
ADO could be related to the high levels of chromosomal
mosaicism which have been detected at these stages by
fluorescence in-situ hybridization (Harper et al, 1995; Munn6
et al, 1994).
For an autosomal recessive disease such as CF, ADO should
not cause a serious diagnosis in which an affected embryo is
misdiagnosed and transferred, at least in cases in which both
partners are carrying the same mutation. In this situation,
unaffected heterozygous carrier embryos would be diagnosed
either as homozygous normal or affected and the latter excluded
from transfer. In cases in which the partners are carrying
different mutations, however, it is essential to detect the
heterozygous state for both mutations. Both misdiagnoses of
CF to date have been in couples of this kind (Harper and
Handyside, 1994) and ADO could have been a contributory
cause. Similarly, for dominantly inherited conditions, it will
be essential to minimize the incidence of ADO. For these even
the minimal rate of ADO achieved here, around 5%, may not
be an acceptable reduction in risk.
The reason for ADO is not known. The occurrence of ADO
with both sets of primers for different genes seems to rule out
any explanation based on the quality of the primers or locusspecific effects resulting from, for example, polymorphic
sequences. Amplification failure from one or both target
sequences in single haploid or diploid cells has been welldocumented (Li et al, 1988; Boehnke et al, 1989) and
amplification failure of a highly repeated sequence specific for
the Y chromosome from a single blastomere is the most likely
cause of the misdiagnosis of sex in an early case of PGD for
X-linked disease (Handyside and Delhanty, 1993). Previous
work with single cells, however, prior to application for PGD
has failed to detect ADO with a number of different primer
sets and cycling conditions including those used here for
detection of the AF508 deletion in CF (Lesko et al, 1991;
Strom et al, 1991).
Although preferential amplification of a particular allele
sometimes occurs with conventional PCR, often because of
size differences in the amplified fragments (Walsh et al, 1992),
ADO has not been reported with non-limiting target DNA. It
seems likely, therefore, that ADO is specific for single cell
PCR or at least very low target copy number. With only two
target molecules present in a single diploid cell, ADO could
result from strand breaks or other DNA damage during
preparation for PCR. Some ADO or indeed complete amplification failure may be caused in this way and the reduction
observed with lysis using an alkaline lysis buffer could in part
reflect protection from, for example, endogenous nucleases.
However, the strong correlation between the initial denaturation
temperature and ADO indicates that most ADO originates in
the initial cycles of the primary PCR.
amplified, the relative amounts of both products were normally
distributed with a small shift towards preferential amplification
of the larger, normal allele. In contrast, ADO (around 5%)
appeared to be random between the two alleles as demonstrated
here. If the probability of the genomic target sequence to
denature is < 1, one allele could begin to amplify in an earlier
cycle than the other. Assuming that short amplification products
have an equal and high probability of amplification, quantitative
differences in the relative amounts of amplified fragments
from both alleles are likely to be maintained in subsequent
cycles and could account for this distribution. However, for
one allele apparently to fail to amplify after complete nested
amplification because the amount of amplified product is
undetectable, our mixing experiments suggest that the difference would have to be of the order of 1:250, or less, equivalent
to amplification failure for the first eight cycles of the primary
PCR. This seems unlikely and is not compatible with the very
high efficiency of amplification of at least one allele irrespective
of the lysis or cycling conditions.
We therefore propose that ADO is caused by a combination
of suboptimal PCR conditions and degradation of genomic
target sequences during the initial cycles. In addition, to explain
the high efficiency of amplification of at least one allele, we
suggest that successful denaturation, primer annealing and
extension may inhibit or prevent degradation of the allele
being amplified, during that cycle. A correlation between
complete amplification failure and ADO would still be predicted but would be much less pronounced and perhaps not
detectable without analysis of a large series of single cells. In
support of this hypothesis, preliminary experiments in which
single cells with all PCR reagents minus dNTPs have been
subjected to five cycles at a suboptimal denaturation temperature (90°C), dNTPs added and nested amplification completed
with optimal denaturation temperature (96°C) in the first 10
cycles have demonstrated that the rate of ADO is equivalent
to nested amplification at the suboptimal denaturation temperature. This strongly suggests that some genomic targets have
been destroyed in the preliminary cycles before amplification
under optimal conditions commenced.
If this early degradation of genomic targets is caused by
residual cellular endonucleases or degradative enzymes from
other sources it could explain failure to detect ADO in
previous studies because of subtle differences in the preparation
of cells or purity of the reagents used. Also, differences in the
temperature calibration of different thermal cyclers could have
a pronounced effect on denaturation efficiency which in turn
would affect the probability of degradation before amplification had commenced from each allele. Preliminary work
with single heterozygous cells is therefore essential to establish
the optimal denaturation temperature to minimize ADO without
compromising amplification efficiency because of excessive
degradation of the Taq polymerase. Further work is necesssary
to explore this possibility and investigate whether inhibitors
of DNases, for example, can reduce or eliminate the incidence
of ADO.
Recent quantitative analysis of
by fluorescent PCR suggests that
fication and ADO occur following
heterozygous cells (Findlay et al,
Aknowledgements
the primary CF product
both preferential ampliamplification from single
1995). When both alleles
We thank Professor Robert M.L.Winston who generously supported
this work and Professor Norman Arnheim for helpful comments.
217
P.F.Ray and A.H.Handyside
References
Ao, A., Ray, P., Lesko, J. el al. (1996) Clinical experience with preimplantation
diagnosis of the AF508 deletion causing cystic fibrosis. Prtnat. Diagn., 16,
137-142.
Bej, A.K. and Mahbubani, M.H. (1992) Applications of the polymerase chain
reaction in environmental microbiology. PCR. Methods Appl, 1, 151-159.
Boehnke, M., Amheim, N., Li, H. and Collins, F.S. (1989) Fine structure
genetic mapping of human chromosomes using the polymerase chain
reaction on single sperm: experimental design considerations. Am. J. Hum.
Genet., AS, 21-32.
Coutelle, C , Williams, C , Handyside, A. et al. (1989) Genelic analysis of
DNA from single human oocytes: a model for preimplantation diagnosis
of cystic fibrosis. Br. Med. J., 299, 22-24.
Findlay, I., Ray, P., Quirke, P. et al. (1995) Allelic drop-out and preferential
amplification in single cells and human blastomeres: implications for
preimplantation diagnosis of sex and cystic fibrosis. Hum. Reprod., 10,
1609-1618.
Handyside, A.H. (1993) Diagnosis of inherited disease before implantation.
Reprod. Med. Rev. 2, 51-61.
Handyside, A.H., Lesko, J.G., Tan'n, J.J. et al. (1992) Birth of a normal girl
after in vitro fertilization and preimplantation diagnostic testing for cystic
fibrosis. N. Engl. J. Med, 327, 905-909.
Handyside, A.H. and Delhanty, J.D.A. (1993) Cleavage stage biopsy of human
embryos and diagnosis of X chromosome-linked recessive diseases. In
Edwards, R.G. (ed.). Preconception and Preimplantation Diagnosis of
Human Genetic Disease. Cambridge University Press, Cambridge, pp.
239-270.
Harper, J.C. and Handyside, A.H. (1994) The current status of preimplantation
diagnosis. Curr. Obstet. Gynecol., 4, 143-149.
Harper, J.C, Coonen, E., Handyside, A.H. et al. (1995) Mosaicism of
autosomes and sex chromosomes in morphologically normal, monospermic
preimplantation human embryos. Prenat. Diagn., 15, 41—49
Holding, C. and Monk, M.(1989) Diagnosis of beta-thalassaemia by DNA
amplification in single blastomeres from mouse preimplantation embryos.
Lancet, II, 532-535.
Lesko, J., Snabes, M., Handyside, A. H. and Hughes, M.(199I) Amplification
of the cystic fibrosis DF5O8 mutation from single cells: applications toward
genetic diagnosis of the preimplantation embryo. (Abstr.) Am. J. Hum.
Genet., 49 (4), 223.
Li, A., Gyllenstein, U.B., Cui, X. et al. (1988) Amplification and analysis of
DNA sequences in single human sperm and diploid cells. Nature, 335,
414-419.
Li, H., Cui, X. and Amheim, N. (1991) Analysis of DNA sequence variation
in single cells. In Methods: A Companion to Methods in Enzymologv.
Vol. 2. Academic Press, New York, USA pp. 49-59.
Liu, J., Lissens, W.. Devroey, P. et al. (1992) Efficiency and accuracy of
polymerase-chain-reaction assay for cystic fibrosis allele delta F5O8 in
single cell. Lancet, 339, 1190-1192.
Munne\ S., Weier, H.U., Grif6, J. and Cohen, J. (1994) Chromosome mosaicism
in human embryos. Biol. Reprod., 51, 373-379.
Pickering, SJ., McConnell, J.M., Johnson, M.H. and Braude, P.R.(I992)
Reliability of detection by polymerase chain reaction of the sickle cellcontaining region of the p^-globin gene in single human blastomercs. Hum.
Reprod., 7, 630-636.
Ray, P.F., Winston, R.M.L. and Handyside, A.H. (1994) Single cell analysis
for diagnosis of cystic fibrosis and Lesch—Nyhan syndrome in human
embryos before implantation. Miami (Abstract) In Bio/Technology, Short
Reports, Advances in Gene Technology: Molecular Biology and Human
Genelic Disease 55, 46.
Ray, P.F., Winston, R.M.L. and Handyside, A.H. (1995) Elimination of allele
drop-out (ADO) in single cell analysis for diagnosis of cystic fibrosis (CF).
(Abstr.) Hum. Reprod., 10, (Abstract Book 2), 131.
Strom, CM., Verlinsky, Y., Milayeva, S. et al (1990) Preconception genetic
diagnosis of cystic fibrosis [letter]. Lancet, 336. 306-307
Strom, CM., Rechitsky, S. and Verlinsky. Y. (1991) Reliability of gender
determination using the polymerase chain reaction (PCR) for single cells.
J. In Vitro Fertil. Embryo. Transfer, 8, 225-229.
Trumper, L.H., Brady, G., Bagg, A. et al. (1993) Single-cell analysis
of Hodgkin and Reed-Stemberg cells: molecular heterogeneity of gene
expression and p53 mutations. Blood, 81, 3097-3115.
Walsh, P.S., Erlich, H.A. and Higuchi, R. (1992) Preferential PCR amplification
of alleles: mechanisms and solutions. PCR Methods Appl., 1, 241-250.
218
Zimmerman, K., Pischinger, K. and Mannhalter, J. W. (1993) Rapid
nonradioacuve detection of HIV-1 RNA from a single-cell equivalent by
reverse transcription PCR with nested primers. Biotechniques. 15, 806-808.
Received on August 15, 1995: accepted on November 22, 1995