Human Reproduction, Vol.28, No.7 pp. 2010 –2020, 2013
Advanced Access publication on April 30, 2013 doi:10.1093/humrep/det132
ORIGINAL ARTICLE Reproductive genetics
Tetraploidy in hydatidiform moles
Linda Sundvall 1, Helle Lund 2, Isa Niemann 3, Uffe Birk Jensen 1,4,
Lars Bolund 1, and Lone Sunde 1,4,*
1
Department of Biomedicine, Aarhus University, Aarhus DK-8000, Denmark 2Institute of Pathology, Aalborg Hospital, Aalborg University
Hospital, Aalborg DK-9000, Denmark 3Department of Gynaecology and Obstetrics, Aarhus University Hospital, Skejby, Aarhus DK-8200,
Denmark 4Department of Clinical Genetics, Aarhus University Hospital, Skejby, Aarhus DK-8200, Denmark
*Correspondence address. Tel: +45-7845-5560; Fax: +45-8678-3181; E-mail: [email protected] and [email protected]
Submitted on December 3, 2012; resubmitted on April 5, 2013; accepted on April 9, 2013
study question: How does tetraploidy develop in hydatidiform moles (HMs), and what is the frequency of the different origins?
summary answer: Most molar pregnancies with tetraploid cells appear to be produced by somatic endoreduplications, while a minority
originate from a tetraploid zygote. The frequency of zygotic tetraploidy was estimated to be 0.7%.
what is known already: The parental origin of the genome in tetraploid HMs has only been evaluated in a few cases, most showing
three genome sets from the father (PPPM). Estimates of the proportion of HMs that are tetraploid vary between 2 and 28%.
study design, size, duration: From 1986 to 2010, unfixed samples of clinically suspected molar pregnancies were forwarded to the
Danish Mole Project. For this cohort study 442 samples fulfilled the following criteria for inclusion: macroscopic appearance of HM and ≥10 vesicular chorionic villi with a diameter of ≥1 mm.
participants/materials, setting, methods: Of 403 karyotyped samples, 21 cases disclosed ≥2 tetraploid metaphases.
The 21 cases were scrutinized by karyotyping, flow cytometry (FC) and DNA-marker analysis.
main results and the role of chance: Among 20 HMs, 3 showed the genotype PPPM: one with the sex chromosomes XXYY
and two with XXXY, indicating that they originated in tetraploid zygotes. In 14 androgenetic, one likely androgenetic and two mosaics, the tetraploid cells likely developed by endoreduplications of diploid cells. One case did not fulfil the histopathological criteria for HM.
limitations, reasons for caution: As an inclusion criterion was the macroscopic observation of vesicular chorionic villi, some
non-molar hydropic placentas may have been included and some early moles may have been excluded.
wider implications of the findings: In future, studies to determine that an HM is tetraploid and discriminate cases of mosaicism
and to deduce the origin of the tetraploidy must use the techniques of karyotyping, DNA-marker analysis and FC in combination.
study funding/competing interest(s): No external funding was sought to support this study. None of the authors has any
conflict of interest to declare.
Key words: hydatidiform mole / genetics / tetraploidy / genotype / fertilization
Introduction
Hydatidiform mole (HM) is a pathological condition observed in 1/
500–1000 pregnancies. The morphological classification is performed
using the criteria originally suggested by Szulman and Surti (1978a,b;
Sebire, 2010). Complete HMs (CHMs) are characterized by
absence of embryo, trophoblastic hyperplasia and hydatidiform degeneration of all villi, from oedema to central cistern formation. Partial
HMs (PHMs) should display moderate trophoblastic hyperplasia in a
mixture of normal and oedematous villi, and embryonic/fetal tissue
may be present.
Though the aetiology for HM is unknown, it has long been postulated
that the morphology is correlated with an excessive amount of paternal
DNA (Surti et al., 1986; Slim and Mehio, 2007). The majority of CHMs
are diploid and contain two genome sets from the father and no maternal
set (PP) (Kajii and Ohama, 1977; Kovacs et al., 1991; Niemann et al.,
2008), whereas PHMs are commonly characterized by diandric monogynic triploidy (PPM) (Vassilakos et al., 1977; Szulman and Surti, 1978a,b).
The parental origin of the genome in tetraploid HMs and in tetraploid
conceptuses with vesicular villi has been analysed using chromosomal
heteromorphism and protein polymorphism (Sheppard et al., 1982;
Surti et al., 1986; Vejerslev et al., 1987), and DNA minisatellite
& The Author 2013. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved.
For Permissions, please email: [email protected]
Tetraploidy in hydatidiform moles
2011
2012
markers (Lawler et al., 1991). Recently, the benefits of short tandem
repeat (STR) analysis was highlighted (Murphy et al., 2012). In most
cases triandric tetraploidy (PPPM) was found.
Efforts to estimate the prevalence of tetraploidy in HMs using flow
cytometry (FC) or image cytometry have resulted in quite varying frequencies (2 –30%) (Lage et al., 1992; Berezowsky et al., 1995; Fukunaga
et al., 1995; Rua et al., 1995; Osterheld et al., 2008). Tetraploid cells may
originate in a tetraploid conceptus (zygotic tetraploidy) in which case all
cells should be tetraploid. However, contaminating maternal cells may
give rise to the observation of diploid cells; moreover, tetraploid cells
are frequently observed in chorionic villi, both in uncultured and cultured
cells (Noomen et al., 2001). Thus, the detection of tetraploidy in HMs is
challenging and the definition of tetraploidy needs clarification.
In this study our aim was to characterize the nature of tetraploidy in
HMs and to estimate the frequency of zygotic tetraploidy in HMs.
The study was conducted as a systematic genetic characterization of
21 placentas that showed two or more tetraploid metaphases at
karyotyping, identified among 403 consecutively collected placentas suspected to be molar pregnancies. Our study demonstrates the need for
DNA marker analysis in conjunction with ploidy analysis by FC or traditional karyotyping.
Sundvall et al.
photographed and we subsequently chose to analyse all these metaphases
(Supplementary data, Table SIII). For presentation, simplified versions of
the karyotypes were created: metaphases with chromosome counts 40 –
48 and 82 – 93 were characterized as ‘di’ and ‘tetra’, respectively. If there
was no sign of aneuploidy concerning the sex chromosomes, the composition of sex chromosomes was assumed to be identical in all cells.
DNA marker analysis of conceptuses and mothers was performed using a
panel of 16 STR markers (AmpFlSTR Identifierw) (Supplementary data, Table
SIV), according to the manufacturer’s instructions. Ten nanograms of template DNA was used in each reaction. We frequently noticed minor peaks
in the histograms generated by analysis of the conceptus, which corresponded to the peaks observed in the maternal DNA. We interpreted this
as an indication of maternal contamination (‘contamination peaks’). The fraction of maternal cells in the sample from the conceptus as estimated by the
size of ‘contamination peaks’ was never higher than 10%.
FC was performed on fresh tissue, using red blood cells from trout and
chicken as controls (Vindeløv et al., 1983). To allow the identification of artefacts generated by the controls, each sample was analysed both with and
without controls. The gain was adjusted in order to make 8n peaks (peaks
indicating nuclei with octaploid DNA contents) detectable. In two samples
with 8n peaks, the analysis was repeated with settings allowing a 16n peak
to be detected.
Morphology
Methods
Since 1986 unfixed samples clinically suspected to be HMs have been collected in the Danish Mole Project. The project is approved by the regional Research Ethics Committees and registered by the Danish Data Protection
Agency. All patients gave informed consent. The departments of Gynaecology and Obstetrics in the western part of Denmark (Jutland) forward representative samples from the evacuated tissue to the local pathologists
(formalin fixed) and to the Mole Project (unfixed). Only fresh samples containing at least 10 vesicular villi with a diameter ≥1 mm are eligible for inclusion. In the period April 1986 to December 2010, 442 samples fulfilled the
morphological criteria for inclusion. Of these, 403 were karyotyped. In
order to screen for tetraploidy, we arbitrarily decided to focus on conceptuses showing at least two metaphases with a tetraploid chromosome
count. Twenty-one placentas fulfilled this criterion, and of these 20 fulfilled
the histopathological criteria for HM. Some observations in the cohort of
HMs entering the Danish Mole Project before June 2003 have been published
previously (Niemann et al., 2007; Sunde et al., 2011).
Genetics
Karyotypes were made on uncultured and cultured cells (Vejerslev et al.,
1991). We generally intended to analyse 5 and 10 metaphases from uncultured and uncultured cells, respectively. However, in some cases fewer analysable metaphases were obtained and in some cases more metaphases were
Histopathological diagnoses made by the local pathologists (the ‘microscopic’
diagnoses) were collected from the Danish National Pathology Registry and
Data Bank and from patient records. For ‘macroscopic’ classification, the
entire unfixed sample forwarded to the genetic laboratory was transferred
to Petri dishes with sterile phosphate-buffered saline. The tissue was inspected
with the naked eye and using a dissection microscope (×25). A sample was
classified as macro-CHM when all villi appeared vesicular and no embryonic/fetal parts were observed. A sample was classified as macro-PHM when
some pieces of tissue contained both vesicular villi and villi that appeared
normal. The sample was classified as ‘twinning’ if vesicular or normal appearing
villi only were observed in separate pieces of tissue, but twinning was not
observed among the 21 conceptuses suspected of tetraploidy. Non-villous
tissue (solid tissue, blood clots, etc.) was removed before performing
genetic analyses to minimize maternal contamination.
Results
Twenty-one placentas with vesicular chorionic villi showing at least two
tetraploid metaphases were subjected to detailed genetic analyses.
In 15 cases, androgenesis (PP) was suspected. In 14 cases DNA marker
analysis disclosed exclusively paternal alleles in several loci in the conceptus. Thirteen cases showed universal homozygosity (P1P1) while one
case showed heterozygosity (P1P2) (Fig. 1A and B). In case 263 maternal
Figure 1 DNA marker analysis. The X-axis shows the florescent marker on the allelic ladder of loci amplified and the Y-axis the signal intensity of the PCR
product. (A) Case 180. Marker D7S820: the conceptus (likely zygotic diploid) shows the genotype P1P1. (B) Case 568. Marker D8S1179: the conceptus
(likely zygotic diploid) shows the genotype P1P2. (C) Case 648. Marker D3S1358: in the conceptus (likely zygotic diploid), the paternal peak is significantly
higher than the maternal peak, indicating mosaicism PP/PM. A minor peak in the conceptus is identical to a maternal peak, indicating maternal contamination
(MC). Marker TH01: three different peaks in the conceptus indicate that the paternal genome in the two cell lines originated in two or three spermatozoas
(P1P2/PxM). ‘Px’ indicates that we were not able to determine if the paternal DNA originated from two or three spermatozoas. (D) Case 804. Marker
D21S11: in the conceptus (likely zygotic diploid) the paternal peak is significantly higher than the maternal peak, indicating mosaicism PP/PM. Only one
peak was not identical to a maternal peak, suggesting that the paternal genome in the two cell lines originated in one spermatozoa (P1P1/P1M).
(E) Case 190. Marker D13S317: in the conceptus (likely zygotic tetraploid), the three different peaks in the conceptus indicate that the paternal
genome in the tetraploid cells originated in two or three spermatozoas (P1P2PxM). (F) Case 811. Marker CSF1PO: in the conceptus (likely zygotic
diploid) the heights of the paternal peak (P) and the maternal peak (M) are similar, indicating equal contributions from the two parents (PM).
Tetraploidy in hydatidiform moles
2013
Figure 2 FC of cases 568 and 641. The X-axis shows fluorescence intensity (proportional to the nuclear DNA content) and the Y-axis the number of
particles (nuclei). (A and B) Case 568 with and without chicken (c) and trout (t) red blood cell controls. (C) Chicken and trout red blood cells (controls).
(D and E) Case 641 with and without chicken (c) and trout (t) red blood cell controls. (F) Case 641 with chicken (c) and trout (t) red blood cell controls,
analysed with a lower gain setting. Consequently, it was not possible to specify or visualize the c peak, because it could not be discriminated from the debris
signal 2n: fluorescence as in diploid human cells in G1 phase, 4n: fluorescence as in diploid human cells in G2/M phase and in tetraploid human cells in G1
phase. 8n: fluorescence as in tetraploid human cells in G2/M phase and in octaploid human cells in G1 phase. tx2, tx3 and tx4: aggregates of 2, 3 and 4 trout
red blood cells, respectively.
DNA was not available; however DNA marker analysis revealed homozygosity in all of 16 loci. In all of these 15 cases, the FC histogram showed a
distinct peak in the 2n area (i.e. indicative of diploidy), followed by a peak
corresponding to cells with the double amount of DNA, 4n. The size of
the 4n peak varied within a wide range between cases. In Fig. 2, panels A
and B illustrate a sample with a 4n peak in the lower end of the range. The
largest 4n peak was seen in case 641, which would have been classified as
tetraploid, if ploidy was determined by the DNA contents of the highest
2014
Sundvall et al.
Figure 3 FC of cases 648, 804, 420 and 811. The X-axis shows fluorescence intensity (proportional to the nuclear DNA content) and the Y-axis the
number of particles (nuclei). (A) Case 648 with chicken (c) and trout (t) red blood cell controls. (B) Case 804 with chicken (c) and trout (t) red blood
cell controls. (C) Case 420 with chicken (c) and trout (t) red blood cell controls. (D) Case 811 with chicken (c) and trout (t) red blood cell controls.
2n: fluorescence as in diploid human cells in G1 phase, 4n: fluorescence as in diploid human cells in G2/M phase and in tetraploid human cells in G1
phase. 8n: fluorescence as in tetraploid human cells in G2/M phase and in octaploid human cells in G1 phase. tx2: aggregates of 2 trout red blood cells.
peak, only (Fig. 2D–F). An 8n peak was discernible in most, but not all,
cases (Fig. 2. and Supplementary data, Table SIII). In case 641, reducing
the gain disclosed a peak in the 16n field (Fig. 2F).
In two cases, mosaicism with two diploid cell lines (PP/PM) was suspected. Case 648: the DNA marker analysis suggested mosaicism with
one heterozygous androgenic and one biparental cell line (P1P2/PxM).
(’Px’ indicates that we were not able to determine if the paternal DNA
originated from two or three spermatozoas.) In addition, in all informative loci minor peaks corresponding to all peaks in the maternal
sample were noted, indicating maternal contamination (Fig. 1C). The frequency of the maternal cells was estimated to be 5–10%. At FC, peaks of
2n, 4n and potentially 8n were seen (Fig. 3A). Karyotyping showed both
cells with the karyotype di,XY and cells with the karyotype tetra, XXYY,
indicating that the 2n peak could not be explained by maternal contamination, only. Case 804: the DNA marker analysis suggested mosaicism
with one homozygous androgenic and one biparental cell population,
the paternal allele in the 2-cell populations being identical (P1P1/P1M)
(Fig. 1D). At FC, most cells presented in the 2n region (Fig. 3B).
In three cases, tetraploidy (P1P2PxM)* was considered: cases 420, 578
and 190. The DNA marker analysis showed indications of the genotype
P1P2PxM (see Fig. 1E). In the FC histograms the majority of nuclei presented in the 4n region, indicative of tetraploidy. In addition, in all
three cases small peaks appeared in the 2n and 8n regions (Fig. 3C). Karyotyping of case 420 showed tetra,XXXY and karyotyping of case 578
showed tetra,XXYY; neither of these cases displayed diploid metaphases. In case 190 most metaphases had the tetra,XXXY karyotype.
However, both when analysing uncultured cells and when analysing cultured cells, occasional metaphases with the di,XX karyotype were
noted, the diploid metaphases being more frequent in cultured cells
(Supplementary data, Table SIII). The DNA marker analysis of case
190 showed diminutive maternal contamination peaks with heights
roughly corresponding to the 2n peak observed in FC (Fig. 1E). Thus, it
is highly likely that the 2n peaks and the diploid metaphases were
explained by maternal contamination.
In one case, a biparental genome (PM) was suspected: case 811. The
DNA marker analysis indicated a balanced biparental contribution to
the genome: in 3 loci the conceptus had 1 allele and in 13 loci the conceptus had 2 alleles. In all loci showing 2 alleles, one allele was identical with a
maternal allele, whereas the other could be paternal, the heights of the
peaks indicating equal contributions from the two parents. In nine loci,
2015
Tetraploidy in hydatidiform moles
both the mother and the conceptus were heterozygous and in six of
these, the conceptus had only one of the maternal alleles, showing that
the peaks corresponding to maternal peaks could not be explained by
maternal contamination (Fig. 1F). In the FC histogram, the majority of
cells were present in a 4n peak, followed by a small 8n peak. In addition,
a 2n peak and a small 16n peak were observed (Fig. 3D). Karyotyping of
case 811 showed metaphases with the tetra,XXXX karyotype, no
diploid metaphases were observed.
Morphology
All of the 14 diploid androgenetic conceptuses and the diploid, suspected
androgenetic conceptus were classified as ‘macro-CHM’ at the inspection of unfixed tissue with the naked eye and under the dissection microscope. At the histopathological inspection, 12 of these cases were
diagnosed as CHM, 2 as PHM and one as HM. All of the three conceptuses with the P1P2PxM genotype were classified as ‘macro-CHM’ and
as CHM at the macroscopic and histopathological examinations. Also
for the 2 PP/PM mosaics, the two evaluations agreed on the classification: one (case 648) was classified as macro-CHM and CHM and the
second (case 804) was classified as macro-PHM and PHM. For case
811 where the genetic analysis indicated a balanced biparental
genome, it was noted at the macroscopic examination of unfixed
tissue that all vesicles were small (1 mm). The histopathological inspection classified case 811 as a degenerated placenta. Slides from paraffin
blocks were revised by one of the authors (H.L., pathologist): this analysis
showed predominantly uniform ‘round’ hydropic villi with no trophoblastic hyperplasia or cisterns, and in some areas, the trophoblasts
were almost lost. The revising pathologist thus confirmed the histopathological diagnosis made by the primary pathologist.
The results of the genetic and morphological analyses are summarized
in Table I. For full karyotypes, see Supplementary data, Table SIII.
Discussion
Among 403 karyotyped pregnancies clinically suspected to be hydatidiform moles, we found 21 with two, or more, tetraploid metaphases.
Based on karyotyping, DNA marker analysis and FC, 3 of the 20 cases
with a histopathological diagnosis of HMs, appeared to be tetraploid
with the genotype PPPM, indicating zygotic tetraploidy. In 15 cases the
genomes appeared to be entirely androgenic (PP), while 2 cases
showed indications of mosaicism (PP/PM). In these cases, the tetraploid
cells are likely to have been derived from post-zygotic endoreduplication(s), i.e. post-zygotic tetraploidy.
For case 811, which did not fulfil the morphological criteria for HM,
both karyotyping and the FC analysis suggested tetraploidy: karyotyping
revealed only tetraploid metaphases and FC showed a distinct peak
within the 4n region being larger than the 2n peak, and followed by a
8n peak. DNA marker analysis revealed a balanced biparental genotype
(PM). This conceptus likely arose by fertilization of one oocyte by one
spermatozoon. As no diploid metaphases were noted, one may speculate that tetraploidy was generated by endoreduplication of the genome
in the zygote, and that the 2n peak in FC histogram was caused by maternal contamination. However, the very small size of the ‘contamination
peaks’ in the DNA marker analysis showed that maternal contamination
was not extensive in this sample, indicating that (most of) the 2n peak was
generated by diploid cells from the conceptus. Thus, it is more likely that
the endoreduplication took place post-zygotically, and the diploid cells in
the conceptus, by chance or selection, did not present as metaphases.
Characterizing tetraploidy
The recorded frequency of tetraploidy depends upon the methods used
and the criteria chosen for diagnosing tetraploidy.
Karyotyping
With the use of karyotyping, exclusively, we would have diagnosed tetraploidy in three cases (420, 578 and 811), as uncultured and cultured cells,
revealed tetraploid metaphases, only. In case 190 ‘tetraploidy and maternal contamination’ would have been suggested. The remaining 19 cases
would have been classified as diploid, ignoring the tetraploid metaphases
due to the well known tendency for tetraploidization in placental cells
(Noomen et al., 2001). In tetraploid cells with an uneven number of Y
chromosomes, such as tetra,XXXY for cases 190 and 420, karyotyping
can identify zygotic tetraploidy. However, when a tetraploid conceptus
has an even number of Y chromosomes, post-zygotic endoreduplication
of a diploid conceptus cannot be excluded. For a conceptus displaying
diploid and tetraploid metaphases and no Y chromosome, karyotyping
cannot discriminate between maternal contamination of a zygotic tetraploid conceptus and a conceptus originating from a diploid zygote
encompassing post-zygotic tetraploid cells due to endoreduplication.
For the detection of zygotic tetraploidy, karyotyping therefore comes
with risks of both false-negative and false-positive results. Furthermore,
karyotyping can only be done on dividing cells and it only allows analysis of
a limited number. Thus, in the case of multiple cell lines, frequency estimates are uncertain. In addition karyotyping harbours a risk of ‘subjectivity’. To allow karyotyping, ‘suitable’ metaphases must be chosen,
prompting the technician to avoid metaphases with overlapping chromosomes. This could lead to an underestimation of the number of tetraploid
metaphases. On the other hand, the technician may chose to document
abnormal or interesting metaphases, which could lead to photographing
‘too many’ tetraploid metaphases. If the cells are cultured before analysis, the risk of error is further increased due to the risk of some cells
growing more rapidly than others. This introduces the possibility of a
few contaminating maternal cells outgrowing the cells from the conceptus. This may be the cause of the diploid cells in case 190 being more frequently in the cytogenetic analysis of cultured cells than in uncultured
cells.
DNA marker analysis
If tetraploidy in HMs was characterized by DNA marker analysis only,
and the diagnosis was made each time it revealed a ratio between paternal and maternal contribution to the genome of 3:1, four cases (190, 420,
578, 648) would have met the criteria for tetraploidy. This definition
would have excluded case 811 from being diagnosed as a tetraploid
HM, in accordance with the expectation from the histopathological classification of this conceptus. However, due to the inability to identify
mosaicism when the two cell lines are equally frequent, DNA marker
analysis comes with a risk for misclassification of diploid mosaics,
PP/PM, as tetraploid, PPPM, as illustrated by case 648 (Fig. 1C).
Furthermore, as this strategy relies on the assumption that all tetraploid HMs have the genotype PPPM, this method would not allow the detection of tetraploid HMs that does not comply with that assumption.
Lawler et al. (1991) found equal paternal and maternal contributions
2016
Table I Results of genetic and morphologic analyses.
Case
Morphologya
........................
Macro
Simplified karyotype; uncultured cellsb
Simplified karyotype; cultured cellsb
‘Combined karyotype’c
FC-ploidyd
Genotypee
Mode of formationf
Micro
..........................................................................................................................................................................................................................................................
180
CHM
CHM
Di,XX[25]
Tetra,no Y[7]/di,XX[1]
Tetra,no Y[7]/di,XX[26]
Di
P1P1
Post. tetra.
217
CHM
CHM
Tetra,XXXX[2]/di,XX[8]
Di,XX[7]
Tetra,XXXY[2]/di,XX[15]
Di
P1P1
Post. tetra.
224
CHM
CHM
Tetra,no Y[3]/di,XX[18]
Di,XX[3]
Tetra,no Y[3]/di,XX[21]
Di
P1P1
Post. tetra.
227
CHM
HM
Tetra,XXXX[2]/di,XX[15]
tetra,no Y[1]/di,XX[3]
Tetra,XXXX[3]/di,XX[18]
Di
P1 P 1
Post. tetra.
230
CHM
CHM
Tetra,no Y[2]/di,XX[12]
Di,XX[9]
Tetra,no Y[2]/di,XX[21]
Di
P1P1
Post. tetra.
400
CHM
PHM
Di,XX[10]
tetra,no Y[3]/di,XX[5]
Tetra,no Y[3]/di,XX[15]
Di
P1P1
Post. tetra.
543
CHM
CHM
Tetra,no Y[2]/di,no Y[7]
NA
Tetra,no Y[2]/di,no Y[7]
Di
P1P1
Post. tetra.
641
CHM
CHM
Tetra[1]
Tetra,XXXX[9]/di,XX[2]
Tetra,XXXX[10]/di,XX[2]
Tetra
P1P1
Post. tetra.
686
CHM
CHM
Tetra,no Y[1]/di,XX[4]
Tetra,XXXX[9]
Tetra,XXXX[10]/di,XX[4]
Di
P1P1
Post. tetra.
778
CHM
PHM
Tetra,XXXX[2]/di,XX[3]
Tetra,XXXX[1]/di,XX[9]
Tetra,XXXX[3]/di,XX[12]
Di
P1P1
Post. tetra.
803
CHM
CHM
Tetra,no Y[3]/di,XX[3]
Di,XX[10]
Tetra,no Y[3]/di,XX[13]
Di
P1P1
Post. tetra.
813
CHM
CHM
Tetra,no Y[3]/di,XX[4]
Di,XX[10]
Tetra,no Y[3]/di,XX[14]
Di
P1P1
Post. tetra.
835
CHM
CHM
Tetra,no Y[1]/di,XX[4]
Tetra,XXXX[3]/di,XX[7]
Tetra,XXXX[4]/di,XX[11]
Di
P1P1
Post. tetra.
568
CHM
CHM
Di,XX[5]
Tetra,XXXX[7]/di,XX[3]
Tetra,XXXX[7]/di,XX[8]
Di
P1P2
Post. tetra.
263
CHM
CHM
Tetra,no Y[3]/di,XX[18]
NA
Tetra,no Y[3]/di,XX[18]
Di
HOh
Post. tetra.
648
CHM
CHM
Tetra,[1]/di,XY[1]
Tetra,XXYY[11]
Tetra,XXYY[12]/di,XY[1]
Di
P1P2/PxM
Post. tetra.
804
PHM
PHM
Di,XX[5]
Tetra,XXXX[2]/di,XX[8]
Tetra,XXXX[2]/di,XX[13]
Di
P1P1/P1M
Post. tetra.
190
CHM
CHM
Tetra,XXXY[11]/di,XX[1]
Tetra,XXXY[6]/di,XX[5]
Tetra,XXXY[17]/di,XX[6]
Tetra
P1P2PxM
Zyg. tetra
420
CHM
CHM
Tetra,XXXY[3]
Tetra,Y[3]
Tetra,XXXY[6]
Tetra
P1P2PxM
Zyg. tetra
578
CHM
CHM
Tetra,XYY[4]
Tetra,XXYY[10]
Tetra,XXYY[14]
Tetra
P1P2PxM
Zyg. tetra
Tetra,XXXX[10]
Tetra,XXXX[12]
Tetra
PM
Post-tetra
811
g
CHM
g
Deg. plac. Tetra,XXXX[2]
a
Sundvall et al.
Macro: the classification made by inspection with the naked eye and using a dissection microscope (×25). CHM: all chorionic villi appeared vesicular and no embryonic/fetal parts were observed. PHM: some pieces of tissue contained both vesicular
and normally appearing villi. Micro: histopathological diagnoses.
b
In the simplified karyotypes, metaphases with chromosome counts 40 –48 were characterized as ‘di’ and metaphases with counts 82 –93 were characterized as ‘tetra’. All metaphases with chromosome counts other than ‘di’ or ‘tetra’ were
re-analysed in order to differentiate between ‘real abnormalities’ and artefacts due to loss of chromosomes during preparation. Metaphases that did not display a Y chromosome, where the quality did not allow a more detailed analysis, are marked
‘no Y’. In order to classify an abnormality as ‘real’, this had to be observed in at least two metaphases. Metaphases with 49 –81 chromosomes with a quality that did not allow reliable analysis, and metaphases with chromosome counts ,40 were
ignored. For details, see Supplementary data, Table SIII.
c
The ‘combined karyotype’ was composed by combining the information in the simplified karyotypes for uncultured and cultured cells.
d
Di: the FC histogram was indicative of diploidy, i.e. the highest peak appeared in the 2n range. tetra: the FC histogram was indicative of tetraploidy, i.e. the highest peak appeared in the 4n range.
e
Genotype: the parental origin of the genome determined by DNA marker analysis and comparing the identity and the number of alleles, as well as the heights of the peaks representing the alleles in the conceptus, with those in the mother. The
genotypes for the tetraploid cells are only given for zygotic tetraploid cases. P1P1: Homozygosity in all loci. In at least three loci, no alleles were identical to any of the maternal alleles. P1P2: Heterozygosity in at least three loci. In at least three loci, no
alleles were identical to any of the maternal alleles. P1P2/PxM and P1P1/P1M: in a likely diploid HM, DNA-marker analysis suggested mosaicism, one cell line having biparental genome and the other having androgenetic genome originating in two or
three spermatozoas, and one spermatozo, respectively (Sunde et al., 2011). P1P2PxM: in a likely tetraploid HM, at least three loci showed three different alleles (peaks), of which one was identical to a maternal allele and two were not identical with a
maternal allele. One of the peaks representing an allele not identical with a maternal allele had a height indicating two copies of that allele. ‘Px’ indicates that we were not able to determine if the third paternal genome set originated from a third
spermatozo or from duplication of one of two different spermatozoas. For details, see Supplementary data, Table SIV.
f
Post tetra.: post-zygotic tetraploidy. Zyg. tetra.: zygotic tetraploidy.
g
The macroscopic examiner felt obliged to classify case 811 as CHM, as all chorionic villi appeared vesicular, however, she had expressed doubt if this conceptus was an HM, as all vesicles were small (1 mm). Deg. plac.: degenerated placenta.
h
HO: homozygosity in all loci (maternal DNA not available).
2017
Tetraploidy in hydatidiform moles
to a tetraploid PHM, illustrating that in some cases this assumption may
not be correct.
When a substantial number of unlinked loci are analysed, DNA
marker analysis allows reliable identification/exclusion of maternal contamination. Although we meticulously removed maternal tissue from the
chorionic villi, the observation of ‘contamination peaks’ in most of the
DNA marker analyses indicates that there were likely maternal cells in
all samples.
Flow cytometry
Estimation of ploidy using FC has advantages over karyotyping as it does
not require living cells and as it allows analysis of a large number of cells.
The criteria for diagnosing tetraploidy by FC is often set by a peak within
the tetraploid region (4n) holding a certain amount of cells, followed by
an 8n peak. The fraction of cells ‘required’ in the 4n peak to make the
diagnosis ‘tetraploidy’ has been set to 10, 15, 20 and 25%, respectively
(Lage et al., 1992; Berezowsky et al., 1995; Fukunaga et al., 1995; Rua
et al., 1995).
Using image cytometry, Osterheld et al. defined a sample as tetraploid
when ‘a main peak was found between DNA index 1.70 and 2.30’, i.e. at
least 50% of cells in the 4n region (Osterheld et al., 2008). The definitions
used seem to have little influence of the fraction of HMs classified as tetraploid, as the observed frequencies of tetraploidy in these studies were 28,
2, 24, 13, and 20%, respectively.
In our material, the three zygotic tetraploid HMs, cases 190, 420 and
578, showed a 4n peak that was higher than the 2n peak and followed by
an 8n peak. That is, these cases would have been classified as tetraploid
using any of the published criteria for diagnosing tetraploidy using cytometry, including the more strict criteria made for image cytometry. A
more strict criteria should be used when the fraction of maternal cells
in the sample analysed deliberately is minimized, as in the study by
Osterheld et al. (2008) and in the present study, compared with
studies where FC was performed, using maternal cells as internal controls (Lage et al., 1989; Berezowsky et al., 1995; Fukunaga et al., 1995;
Rua et al., 1995). However, also cases 641 and 811 that likely contained
tetraploid cells due to post-zygotic tetraploidization would have been
classified as tetraploid, using this criterion.
There are several challenges in using FC for discriminating between
zygotic tetraploidy and post-zygotic tetraploidy: first, the 4n peak can
contain a combination of signals from diploid cells in the G2/M phase
and tetraploid cells. Thus, defining tetraploidy by a certain size of the
4n peak comes with the risk of making severe overestimations of the
frequency of tetraploidy. For that reason most authors require that an
8n peak is present in order to diagnose tetraploidy. However, as both
zygotically and post-zygotically generated tetraploid cells may enter
the G2/M phase, both can generate (or add to) an 8n peak. Thus in placental tissue, identifying an 8n peak does not aid in the differentiation
between zygotic and post-zygotic tetraploidy. Further as one can
never exclude that the sample contains maternal cells, the presence of
a 2n peak cannot be used to prove post-zygotic tetraploidy. And as
one (e.g. tetraploid) cell line may outgrow another (e.g. diploid) cell
line in vivo, zygotic tetraploidy cannot be proven, even in the absence
of a 2n peak.
The optimal way of identifying zygotic
tetraploidy in HMs
In karyotyping, the identification of 92 chromosomes and an uneven
number of Y chromosomes should establish a valid diagnosis of zygotic
tetraploidy. In other cases, to distinguish post-zygotically endoreduplicated cells from those originating in tetraploid zygotes and to identify mosaicism (e.g. to discriminate PP/PM from PPPM) and maternal
contamination, karyotyping/FC must be supplemented with DNA
marker analysis. In our material, two of the three zygotic tetraploid conceptuses (cases 190 and 420, both: tetra,XXXY) would have been diagnosed correctly by karyotyping alone, whereas in one case (case 578:
tetra,XXYY) zygotic tetraploidy was only evident after adding the
results of the DNA marker analysis. Interpretation of DNA marker analysis without an independent estimate of ploidy, may lead to errors, especially in the case of mosaicism.
The phenotype of tetraploid vesicular
placentas
Usually the phenotype correlates with the genotype: CHMs mostly being
diploid androgenetic (PP) and PHMs being diandric triploid (PPM) (Kajii
and Ohama, 1977; Vassilakos et al., 1977; Kovacs et al., 1991). We
observed the phenotype CHM in three conceptuses with the genotype
PPPM; however, combining with published observations, it appears that
conceptuses with the PPPM genotype can have a wide spectrum of phenotypes (HA, PHM and CHM; see Table II). Morphological diagnoses can
Table II Phenotypes of tetraploid conceptuses with the PPPM genotype.
Sex
chromosomes
Phenotype
Gestational age at
evaluation (weeks)
.............................................................................................................................................................................................
Sheppard et al. (1982)
XXXX
HA
11
Surti et al. (1986), case 1
XXXY
PHM
21
Surti et al. (1986), case 2
XXXY
PHM
11.5
Vejerslev et al. (1987)
91,XXXX,-4
CHM
16
a
Murphy et al. (2012)
XXXY
HM
10
Present work, case 190
XXXY
CHM
16.1
Present work, case 420
XXXY
CHM
12.2
Present work, case 578
XXYY
CHM
14.7
a
Sub-typing on morphological features, PHM versus CHM, was problematic
2018
be difficult to make on moles evacuated early (Sebire, 2010); however,
no obvious correlation with gestational age is seen among these cases.
Thus, the determining factor for the phenotype of PPPM conceptuses
is at present unknown.
In one placenta showing vesicular chorionic villi, but not fulfilling the
histopathological criteria for a hydatidiform mole, we found indications
of tetraploidy. However, the DNA marker analysis showed a balanced
biparental origin of the genome, making it more likely that the tetraploid
cells arose by a post-zygotic endoreduplication. Tetraploidy with a
balanced biparental genome has been reported in an HM (Lawler
et al., 1991) and in hydropic placentas (Fukunaga, 2004). It is thus tempting to suggest that the vesicular appearance of this placenta was caused by
the presence of the tetraploid cells. However, as placentas displaying
vesicles have been selected for this study, and as endoreduplication is frequent in non-molar placentas, too (Cera et al., 1992; Berezowsky et al.,
Sundvall et al.
1995; Noomen et al., 2001), it is also possible that the phenotype in this
placenta was caused by other factors.
Sex chromosomes
If zygotic tetraploidy arose by trispermy, we would expect that the four
potential karyotypes: 92,XXXX/92,XXXY/92,XXYY/92,XYYY
would occur in the ratios 1:3:3:1. If zygotic tetraploidy arose by dispermy,
with one paternal genome set being duplicated, we would expect that the
four potential karyotypes: 92,XXXX/92,XXXY/92,XXYY/92,XYYY
would occur with ratios 1:1:1:1. Of the published eight tetraploid HMs
with one maternal and three paternal genome sets (including the three
cases from the present study), two cases had the sex chromosomes
XXXX, five cases had XXXY, one case had XXYY and no case had
XYYY (Table II) (Sheppard et al., 1982; Surti et al., 1986; Vejerslev
Figure 4 Modes of fertilization in HMs with tetraploid cells. The most likely modes of fertilization in HMs with diploid and/or tetraploid cells; zygotic
tetraploidy: an oocyte is fertilized by three spermatozoa (P1, P2 and Px). Alternatively the oocyte could be fertilized by two spermatozoa, in which case
a third paternal genome set (Px) would be generated by endoreduplication of one of the paternal pronuclei. The fusion of the four pronuclei results in a
tetraploid zygote P1P2PxM. By mitotic cell divisions a conceptus is formed containing tetraploid cells, exclusively (cases 190, 420 and 578). ‘Px’ indicates
that we were not able to determine if the paternal DNA originated from two or three spermatozoas. Post-zygotic tetraploidy: an oocyte may be fertilized
by a single spermatozoon, followed by two endoreduplications of a paternal pronucleus. Before fusion of the four pronuclei, an erroneous cell division
isolates two pronuclei in one cell and the other two pronuclei in a second cell. When the two pronuclei in each cell melt together, two diploid cells are
created (diploidization P1P1/P1M). If the PM cell line degenerates, a diploid, androgenetic, universally homozygous conceptus is generated. A subsequent
endoreduplication of one (or more) P1P1 cells forms a conceptus with both diploid and tetraploid cells: P1P1/P1P1P1P1 (cases 180, 217, 224, 227, 230, 400,
543, 641, 686, 778, 803, 813, 853 and most likely 263). If two or three spermatozoas fertilize the oocyte, a similar mechanism can give rise to an androgeneous, heterozygous conceptus with diploid and tetraploid cells, P1P2/P1P1P2P2 (case 568). *If the PM cell line is not lost, the conceptus will be a mosaic
(PP/PM) from the 2-cell stage, and with post-zygotic endoreduplications up to four different cell lines, PP/PPPP/PM/PPMM, may co-exist (cases 804 and
648) (in the figure, the PPMM cell line is not illustrated).
2019
Tetraploidy in hydatidiform moles
et al., 1987; Murphy et al., 2012). Though the numbers are too small to
draw firm conclusions, they give a weak indication that trispermy is more
likely than dispermy + endoreduplication and/or that fertilization with
spermatozoas with X chromosomes are more likely to lead to tetraploid
HMs than fertilization with spermatozoas with a Y chromosome.
Modes of fertilization
studies on genotypes in diploid, mosaic and tetraploid HMs, and correlation with the phenotypes may shed light on fertilization (e.g. the role of
spermatozoas carrying X and Y chromosomes) and on very early cell development (e.g. the role of the timing of the first cell division).
Supplementary data
We observed two main variants of allelic composition in conceptuses
with tetraploid cells. The two variants are likely related to different
modes of fertilization. In Fig. 4 we summarize the most likely modes of fertilization in HMs with diploid and/or tetraploid cells.
Supplementary data are available at http://humrep.oxfordjournals.org/.
† In rare cases, tetraploidy arise at fertilization, i.e. zygotic tetraploidy. An
oocyte is fertilized by three spermatozoa, alternatively by two spermatozoa with a subsequent endoreduplication of one paternal pronucleus. The fusion of the four pronuclei results in a tetraploid
zygote P1P2PxM. Mitotic cell divisions result in a conceptus containing
tetraploid cells, exclusively.
† More frequently, tetraploidy arise after fertilization, i.e. post-zygotic
tetraploidy. Post-zygotic tetraploidy can in principle arise by simple
endoreduplication of the nucleus in any diploid cell, which would
result in a conceptus with both diploid and tetraploid cells.
However, as diploid HMs may arise from a zygote with four pronuclei
(Sunde et al., 2011), several possibilities do exist. If it is correct that
both diploid and tetraploid HMs originate in zygotes with four pronuclei, the timing of the first cell division is likely what determines
the fate of the conceptus: if the first cell division is late, the four pronuclei fuse before the cell division (zygotic tetraploidi). If the first
cell division occurs early it can form two cells, each with two pronuclei
(zygotic diploidy and mosaicism). Mosaicism can be reduced to a nonmosaic state, if one cell line is lost. And the diploid cell line(s) can give
rise to tetraploid cells by post-zygotic endoreduplication.
The authors wish to thank patients and the departments of gynaecology
and pathology in Jutland, Denmark for providing samples and clinical data
to the project. Flow cytometry was performed at the FACS Core Facility,
The Faculty of Health Sciences, Aarhus University, Denmark. Laboratory
technicians Bodil Schmidt and Anette Thomsen are thanked for skilled
laboratory work. And Sol Britt Sundvall is specially thanked for formatting
and editing of figures.
The frequency of zygotic tetraploidy
Among the 403 consecutively collected, clinically suspected molar pregnancies, we identified zygotic tetraploidy in 3 HMs (0.7%). One of our
limitations is the fact that an inclusion criterion for this study was that conceptuses must display the macroscopic appearance of an HM. For this
reason, some early moles may have been excluded and, as illustrated
by the case 811, some non-molar cases may have been included. Moreover, if the placental tissue was considerably contaminated with maternal
cells, diploid cells could have overgrown the placental cells, which in turn
could have prevented us from detecting a tetraploid conceptus. Thus, we
may have slightly underestimated or overestimated the frequency of
zygotic tetraploidy. The study of Lawler et al. (1991) is to our knowledge
the only previous study systematically examining HMs for ploidy and parental origin of the genome. In this study one tetraploid PHM was identified among 201 HMs (0.5%). Thus at present, the most valid estimate of
the frequency of zygotic tetraploidi among conceptus with the molar
phenotype seems to be ‘less than 1%’.
Conclusion
The majority of HMs with tetraploid cells appear to have developed by
somatic endoreduplication of diploid cells, while a minority originated
from tetraploid zygotes. To compare reports, we have to agree on a
common terminology and on common criteria for tetraploidy. Further
Acknowledgements
Authors’ roles
Lone Sunde and Linda Sundvall conceived the ideas of this study and
designed the project. Linda Sundvall has drafted the manuscript. Lone
Sunde and I.N. contributed to the acquisition of samples and data.
U.B.J. performed flow cytometric analyses and formatting of related
figures. H.L. was the revising pathologist. Lone Sunde and Linda Sundvall
did interpretation of data obtained from karyotypes, flow cytometry and
DNA marker analysis, and wrote the manuscript. H.L., I.N., U.B.J. and
L.B. performed critical revision of the manuscript. All authors have
given their final approval of present version to be published.
Funding
No external funding was sought for this study.
Conflict of interest
None declared.
References
Berezowsky J, Zbieranowski I, Demers J, Murray D. DNA ploidy of
hydatidiform moles and nonmolar conceptuses: a study using flow and
tissue section image cytometry. Mod Pathol 1995;8:775 – 781.
Cera G, Fruttero A, Rua S, Comino A, Abrate M. Flow cytometric studies in
spontaneous abortions. Applications in the medico-legal practice. Forensic
Sci Int 1992;54:167 – 175.
Fukunaga M. Immunohistochemical characterization of p57Kip2 expression
in tetraploid hydropic placentas. Arch Pathol Lab Med 2004;128:897– 900.
Fukunaga M, Endo Y, Ushigome S. Flow cytometric and clinicopathologic
study of 197 hydatidiform moles with special reference to the
significance of cytometric aneuploidy and literature review. Cytometry
1995;22:135 – 138.
Kajii T, Ohama K. Androgenetic origin of hydatidiform mole. Nature 1977;
268:633– 634.
Kovacs BW, Shahbahrami B, Tast DE, Curtin JP. Molecular genetic analysis of
complete hydatidiform moles. Cancer Genet Cytogenet 1991;54:143– 152.
2020
Lage JM, Weinberg DS, Yavner DL, Bieber FR. The biology of tetraploid
hydatidiform moles: histopathology, cytogenetics, and flow cytometry.
Hum Pathol 1989;20:419 – 425.
Lage JM, Mark SD, Roberts DJ, Goldstein DP, Bernstein MR, Berkowitz RS. A
flow cytometric study of 137 fresh hydropic placentas: correlation
between types of hydatidiform moles and nuclear DNA ploidy. Obstet
Gynecol 1992;79:403 – 410.
Lawler SD, Fisher RA, Dent J. A prospective genetic study of complete
and partial hydatidiform moles. Am J Obstet Gynecol 1991;164:1270– 1277.
Murphy KM, Descipio C, Wagenfuehr J, Tandy S, Mabray J, Beierl K,
Micetich K, Libby AL, Ronnett BM. Tetraploid partial hydatidiform mole:
a case report and review of the literature. Int J Gynecol Pathol 2012;
31:73 – 79.
Niemann I, Petersen LK, Hansen ES, Sunde L. Differences in current clinical
features of diploid and triploid hydatidiform mole. BJOG 2007;
114:1273– 1277.
Niemann I, Bolund L, Sunde L. Twin pregnancies with diploid hydatidiform
mole and co-existing normal fetus may originate from one oocyte. Hum
Reprod 2008;23:2031 – 2035.
Noomen P, van den Berg C, de Ruyter JL, Van Opstal D, Los FJ. Prevalence of
tetraploid metaphases in semidirect and cultured chorionic villi. Fetal Diagn
Ther 2001;16:129 –132.
Osterheld MC, Caron L, Chaubert P, Meagher-Villemure K. Combination of
immunohistochemistry and ploidy analysis to assist histopathological
diagnosis of molar diseases. Clin Med Pathol 2008;1:61 –67.
Rua S, Comino A, Fruttero A, Abrate M. DNA flow cytometric analysis of
abortions. A simple method for detection of triploidy and tetraploidy in
the trophoblastic cells. Pathologica 1995;87:107– 111.
Sundvall et al.
Sebire NJ. Histopathological diagnosis of hydatidiform mole: contemporary
features and clinical implications. Fetal Pediatr Pathol 2010;29:1 – 16.
Sheppard DM, Fisher RA, Lawler SD, Povey S. Tetraploid conceptus with
three paternal contributions. Hum Genet 1982;62:371– 374.
Slim R, Mehio A. The genetics of hydatidiform moles: new lights on an ancient
disease. Clin Genet 2007;71:25 – 34.
Sunde L, Niemann I, Hansen ES, Hindkjaer J, Degn B, Jensen UB, Bolund L.
Mosaics and moles. Eur J Hum Genet 2011;19:1026– 1031.
Surti U, Szulman AE, Wagner K, Leppert M, O’Brien SJ. Tetraploid partial
hydatidiform moles: two cases with a triple paternal contribution and a
92,XXXY karyotype. Hum Genet 1986;72:15 – 21.
Szulman AE, Surti U. The syndromes of hydatidiform mole. I. Cytogenetic
and morphologic correlations. Am J Obstet Gynecol 1978a;131:665–671.
Szulman AE, Surti U. The syndromes of hydatidiform mole. II. Morphologic
evolution of the complete and partial mole. Am J Obstet Gynecol 1978b;
132:20– 27.
Vassilakos P, Riotton G, Kajii T. Hydatidiform mole: two entities. A
morphologic and cytogenetic study with some clinical consideration. Am
J Obstet Gynecol 1977;127:167 – 170.
Vejerslev LO, Dissing J, Hansen HE, Poulsen H. Hydatidiform mole: genetic
origin in polyploid conceptuses. Hum Genet 1987;76:11 – 19.
Vejerslev LO, Sunde L, Hansen BF, Larsen JK, Christensen IJ, Larsen G.
Hydatidiform mole and fetus with normal karyotype: support of a
separate entity. Obstet Gynecol 1991;77:868 –874.
Vindeløv LL, Christensen IJ, Nissen NI. Standardization of high-resolution
flow cytometric DNA analysis by the simultaneous use of chicken and
trout red blood cells as internal reference standards. Cytometry 1983;
3:328 – 331.