Atlas of Genetics and Cytogenetics in Oncology and Haematology

Home   Genes   Leukemias   Solid Tumours   Cancer-Prone   Deep Insight   Case Reports   Journals  Portal   Teaching   

X Y 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 NA

Uniparental Disomy

This is an ambitious title to deal with.
Of course, UPD refers to the accidental presence of a chromosome pair or a chomosome segment derived from only one parent in a diploid individual.
In fact, the information on this subject has grown so large that Pub Med, the webb-site of the US National Library of Medecine, by now lists over 550 original titles not to mention the so-call related articles.

In the bulk of this material. I particulary like to stress the elegant contributions from Prs Lidia Larizza, Orsette Zuffardi and their colleagues on the role of parental chromosome 15 inversions in subsequent segmental deletions of that chromosome and their study of UBE3A mutations in AS.
I also want to mention the wealth of information and observations that we owe to Pr A Schinzel and his group and to Dr Dietrich Kotzot in this area.

Lidia, Albert, I thank you whole-heartedly, as well as the Organizing Committee and Dr Konstantin Miller for inviting me to address the Audience of this select ECA Meeting in Bologna. Thank you, indeed for your hospitality.


Slide 1

I thus started in the field at this most exciting period wich I call the Golden Years. Within two of these years, 1959 and 1960, the three major autosomal trisomies, G, E, and D, namely 21, 18 and 13 turned up along with three of the four more common sex chromosome anomalies. XXY, XXX, XO (the XYY would appear later), plus the first example of human chromosome mosaicism.

We all known the prestigious names of the Scientists listed here, wich include two illustrious pionneers of this Country, Marco Fraccaro and Paul Polani.
These achievements had been acquired at the cost of great efforts, particulary with respect to the cultures of the solid tissues or marrow samples, needed to offer the sight of our chromosomes.


Slide 2


Moorehead PS, Nowell PC, Mellman WJ, Batipps DM and Hungerford :

Chromosome preparations of leukocytes cultured from human peripheral blood

Exp Cell Research 1960, 20, 613

In this context, the opportunity of using a few drops of venous blood for short term culture and chromosome studies with Phytohemagglutinin for blast tranformation of monolymphocytes represented a boon to all interested personnel.

1960 is precisely the time when I joined the MGH in Boston and began setting up there a Cytogenetic laboratory.

Slide 3

Other notable discoveries occurred in the sixties in our field, including the sighting of some tiny deletions, but, just as happened in the early years of photography, the chromosomes appeared uniformely dark over a clear white background.

This is not to say that we could not see very interesting details such as they dislpayed here, with the Ph1 chromosome, a dicentric 17-18 E chromosome and a D/D translocation - the later congenital - in this example of the clonal pattern of a leukemic cell at the blast phase of CML.


Slide 4




Casperson T, Zech L, Johansson C, Modest EJ.
Identification of human chromosomes by DNA-binding fluorescent agents.
Chromosoma. 1970 30:215-27.


Couturier J, Dutrillaux B, Lejeune J.
Specific fluorescence of R and G bands in human chromosomes.
CR Acad Sci 1973 276:339-342

As I just said earlier, the specfic identification, as of 1969-1970 of individual chromosomes by fluorochromes pionnered thanks to Caperson, Zech et al. and also obtained by other banding procedures devised by Seabright, Dutrillaux and other major contributors changed the whole ball game.

And, it is under the benefit of so much technical sophistication that I shall now quickly review the cytogenetic results gained from the systematic studies of the product of first trimester spontaneous abortions.

They all pointed to the high rate of gamete aneuploidy, as a fact of observation which served as the basic of the UPD concept and suggested the fairly simple idea that, sometimes, somehow a diploid individual might be conceived or could develop, with one of the 23 chromosome pairs from one single parent.


Slide 5


50 % DU TOTAL :

41, 443-454, 1978.

We had come to learn that one half or so of these aborted fetuses showed major chromosome anomalies, half of them as a trisomy, one fifth as an X-monosomy and one third as a polyploidy, mostly triploidies.


Slide 6


47, +16
32 %
47, +21
13 %
47, +15
10 %
47, +22
13 %

- CREASY M.R. & AL. : HUM. GENET. 31. 177-196. 1976.

And, among the trisomies, four of them largely prevailed, namely trisomy 16 in one third of the cases, and trisomies 21, 22 and 15, each accounting for about 10 % of the lot, thus making up altogether some two thirds of the trisomies observed in these abortuses.

And since, as a rule, meiotic mis-segregation must result in as many nullisomie as disomic gametes, it did not seem to me too far-fetched an idea to statistically envisage the following possibility, namely that coincidental fertilization and complementation of a nullisomic gamete by one disomic for a same chromosome might indeed occur, thus occasionally causing a diploid conceptus to derive one pair from only one parent.


Slide 7

Thus, on the basis of the figures documented for the rate of aneuploidy limited to these four autosomes and the X and making some asumptions wich I shall not her develop, it looked as follows :

for a toll of 20 % abortions of all conceptuses, one half of which were aneuploid, complementation at fertilization for these 5 member chromosomes causing UPD for one of these pairs might occur with the incidences reported here.

In brief, on those premisses, one might envisage 2 or 3 cases of putative UPD for one or the other of these 5 members every 10,000 births and even more when considering an abortion frequency of 50 % !


Slide 8

(as considered in 1979)

The birth of Mendelian non traditional inheritance

1) Homozygous traits inheritable from one carrier parent only

2) Father to son exceptional transmission of an X-linked trait

3) Xg(a-) daughters born to Xg(a+) legitimate fathers

4) Affected daughters born to recessiv X-linked carrier mothers

5) Duplication of chromosomal markers - morphological or molecular - present in only one parent

If that were to happen what might be the occasionnal consequences of deriving one chromosome from one parent only?


Slide 9

American Journal of Medical Genetics 6: 137-143 (1980)

A New Genetic Concept: Uniparental
Disomy and Its Potential Effect, Isodisomy

Eric Engel

Institute of Medical Genetics, Geneva University School of Medicine, Geneva, Switzerland

In recent years, cytogenetic studies of spontaneous abortion products have disclosed a relatively high frequency of aneuploid embryos. These karyotypic anomalies chiefly stem from meiotic errors affecting the distribution of the chromosomes in one of two gametes. This information not only implies the remarkable frequency of gonocyte aneuploidy but also reveals the pre

And thus, after many months of cogitation, I came to spend one night, from a saturday to a sunday, to put down a draft of this idea in writing.


Slide 10

Once in print and published, the idea slept in the medical literature for some years because, at the time of the publication,1980, the means to trace the parental origin of a chromosome were still limited, awaiting the analyses of DNA polymorphisms as schematically shown here.

Given the four constrated alleles of a particular locus in two parents, each one can be traced through appropriate enzyme restriction, gel electrophoresis and adequate marking.

In this diagram from our book, individual 3 has an allele from each parent, as normal and individuals 4 and 5 have only paternal alleles, two contrasted ones for individual 4i.e. heterodisomy and two identical ones for offspring No. 5, the so-called isodisomy. Also it is of note that if this duplicated allele was that of a recessive trait, the individual would be affected.


Slide 11

Am. J. hum. Genet. 42: 215-216, 1988

Editorial : Uniparental Disomy: A Rare Consequence of the High Rate of Aneuploidy in Human Gametes

Dorothy Warburton
Department of Genetics and development of Pediatrics, Columbia University. New York


Am. J. hum. Genet. 42: 217-225, 1988

Uniparental Disomy as Mechanism for Human Genetic Disease

J. Edward Spence, Ronald G. Perciccante, Guillian M. Greig. Huntington F. Willard. David H. Ledbetter. J. Fielding Hejtmancik, Marilyn S. Pollack, William E. O'Brien and Arthur L. Baudet
Howard Hughes Medical Institute, Institute of Molecular Genetics and Department of Microbiology and Immunology, Baylor College of Medicine, Houston: Mercy Hospital, Watertown, NY: and Department of Medical Genetics, University of Toronto, Toronto

And this is precisely the mechanism which helped these investigators to uncover the first thoroughly analyzed and described case of UPD. It was one involving maternal chromosome 7, responsible for cystic fibrosis in an unusually short girl who carried Gly542Ter mutation in her CFTR gene.

This article, of Beaudet's lab, with Ledbetter among the Authors and Spence as the Senior Author, was not only featuring the first case ever sighted of non-traditional recessive inheritance through reduction to homozygosity of the recessive mutant only carried by one of the two parents. It also offered a most comprehensive review of the possible mechanisms leading to the occurrence of UDP.

Interestingly enough the journal Science rejected this report, apparently for describing a situation too exceptional for a broad readership; and, while accepted for publication by the American Journal of Human Genetic, the accompanying editorial almost echoed the very reasons why the other major publications had turned down the article.


Slides 12


Recessive Disorders
UDP type
Pycnodysostosis 1 pat Gelb et al. (1998)

Junctional epidermolysis bullosa, Herlitz type

1 mat Pulkkinen et al. (1997)

Spinal muscular atrophy III (juvenil type)

5 pat Brzustowicz et al; (1994)
Complement deficiency of C4A+C4B 6 pat Welch et al. (1990)
Methylmalonic acidemia 6 pat Abramowicz et al. (1994)
Cystic fibrosis 7 mat Spence et al. (1988), Voss et al. (1989)
Osteogenesis imperfecta (COL1A2 mutation) 7 mat Spotila et al. (1992)
Cystic fibrosis and Kartagener syndrome 7 pat Pan et al. (1998)
Congenital chloride diarrhea 7 pat Hôglund et al. (1994)
Chylomicronemia, familial 8 pat Benlian et al. (1996)
Cartilage / hair hypoplasia 9 mat Sulisalo et al. (1997)
Beta-thalassemia major 11 pat Beldjord et al. (1992)
Complete congenital achromatopsia (rod monochr.) 14 mat Pentao et al. (1992)
Bloom syndrome (with Prader-Willi syndrome) 15 mat Woodage et al. (1994)
Hydrops fetalis alpha-thalassemia 16 pat N'go et al. (1993)
Duchenne muscular dystrophy X mat Quan et al. (1994)
Hemophilia A XY Vidaud et al. (1989)

EE (2/10/1998)




UDP type


Chediak-Higashi Syndrome

1 mat

Dufourcq-Lagelouse et al


Mapple Syrup Disease Type II

1 mat

Lebo et al


Congenital insensivity to pain, anhydrosis (CIPA)

1 pat

Miura et al


Herlitz junctional epidermolysis bullosa

1 pat

Takizawa et al


Mosaicism Rh+/Rh-

1 pat

Miyoshi et al


CIPA+Pyruvate kinase receptor deficiency

1 pat

Indo et al


Leber congenital amaurosis

1 pat

Thomson et al


Retinis, Usher type II

1 pat

Rivolta et al


Lactic acidosis (trifunctionnal protein deficiency)

2 mat

Spiekerkoetter et al





Pseudohermaphroditism (5-alpha reductase deficiency)

2 pat

Chavez et al


Retinis pigmentosa (MERKT)

2 pat

Thompson et al



4 mat

Yang et al


21-Hydroxylase deficiency

6 pat

Lopez-Guttierez et al


Cystic fibrosis

7 mat

Hehr et al


Leigh syndrome

9 mat

Tiranti et al


EE july 2003

It is precisely at this junction that I would like to review the list of some thirty or so different recessive conditions traced to this very mechanism over the last 14 years. Some of these have indeed been observed more than once.


Slide 13

In some studies, particularly the first ones listed here, one can get some idea of the condition of reduction to homozygosity to recessive traits as compared to that of classical biparental recessive inheritance which ranges from up to 2 to 4 % in the series with more than 50 cases tested.

And, it is as much as I shall now devote to this aspect of non-traditional inheritance in UPD.


Slide 14

Nature 1989. 342: 281-5

Genetic imprinting suggested by maternal heterodisomy in nondeletion Prader-Willi syndrome.

Nicholls RD, Knoll JH, Butler MG, Karam S, Lalande M.

Howard Hughes Medical Institute; Havard Madical School, Boston, Massachusetts.

I now turn to another major player in the field of UPD, brought into action by Rob Nicholls et al, the phenomenon of genomic imprinting.


Slide 15

But, to bring that up, let me first refer to the well know and significant observation of a tiny 15q11q13 deletion in the Prader-Willi syndrome, by David Ledbetter and colleagues in 1981.

It did take wonderful eyes to detect such a small, albeit most important cytogenetics detail!


Slide 16

It was Rob Nicholls and colleagues' merit to establish that in the rarer cases of PWS without the tiny deletion, a chromosome pair 15 looked pink, painted exclusively of maternally segregating alleles and markers (!!). Thus, in these instances, these rarer cases showed maternal UPD 15, along with the lack of a paternal chromosome 15.

The obvious lesson to it was that an intact second maternal 15 could not substitute successfully for the missing paternal one. Therefore, in this instance, although normal looking, the second maternal chromosome 15 was lacking the genetic expression of a proper paternal one. Why was it so ? Indeed this very observation was to serve at the introduction of a still poorly understood phenomenon, genomic imprinting.


Slide 17


Definition : Genomic Imprinting

the epigenetic modification of certain genes through methylation as a function of their parental origin

    >> an "imprinted" gene is often considered to be an inactived gene

    >> the result is functional hemizygosity (maternal or paternal) for some allelic pairs

    >> imprint "relaxation" normally occurs early in gametogenesis

We can see on it a fairly simple reminder of the definition of genomic imprinting.


Slide 18

With time and patience, it was recognized that the imprinting disruption caused by the possession of a UPD pair could intervene as a cause of some previously known syndrome as well as a help in delineating some new ones.

UPDs, maternal or paternal, for chromosomes 6, 7, 11 and 15 have occured in a variable proportion of the listed syndromes, while both maternal and paternal UPD 14 each delineated a new syndrome.

Three other pairs came under suspicion of exercising harmful effects through a similar mechanism, although such an interference appears less and less certain for maternal chromosome 2, still quite likely for maternal chromosome 16 and definite for chromosome 20, both paternal and maternal, a topic in full evolution.


Slide 19



MAPS AT 20q13.3


MAPS AT 20q11.2

1) LIU et al 2000
2) EVANS HK et al 2001

Both paternal and maternal chromosome 20 show an imprinting mark, which, on the maternal side, allows sensitivity to parathormone and, on the paternal side, expresses a protein essential for embryofetal neurologic development.


Slide 20

This slide shows what proportion of some well defined syndromes might be caused by a given uniparental pair proven responsible for disrupting the normal imprinting process.


Slide 21

The figures on slide 21 lend support to some extrapolation to evaluate the baseline frequency of a few of the viable UPDs involved as a cause disease.

Thus, if the PWS phenotype is in general viable and knows a clinical frequency of 1 in 20,000 live births, and if maternal UPD 15 serves as an etiology for some 25 % of these cases, one may infer that maternal UPD 15 occurs around once every 80,000 live births. And so on for several other UPDs causing a proportion of syndromic conditions of reasonably well documented overall frequencies.


Slide 22

So far we have in this lecture followed two leads, one looking at the UPDs recognized as the cause of recessive traits, the other as a cause of malformations through the normal process of genomic imprinting.

At this junction, in guise of more systematic approach, we can review, as shown here, the 47 possibilities of UPD for wholesale chromosomes, namely 22 paternal and 22 maternal pairs for the autosomes as well as 3 more pairs for the sex chromosomes, one maternal XX and two paternal ones, namely XX or XY.


Slide 23


1 2 4 6 7
8 9 10 12 13
14 15 16 17 20
21 22     X
1 2 5 6 7
8 11 13 14 15
16 21  22   XY
3 5 11 18 19
3 4 9 10 12
17 18 19 20 X

On this next slide, we show somewhat arbitrarily the chromosome numbers, maternal or paternal, which have contributed a monoparental pair in the make up of one purely and uniformly diploid genome, assuming that the available information allowed an exclusion of the mosaic compounded by an aneuploid component.

This review is comprised of 18 maternal and 14 paternal numbers, for a total of 32. Thus some 15 numbers are still currently without inclusion in a uniparental pair, if we disregard paternal 20 and paternal X, so far only noted in an aneuploid mosaic context.


Slides 24


 1987  21 mat  Créau-Goldberg et al
 1988   7 mat  Spence et al, Voss et al
 1989  15 mat  Nicholis et al
 1989  XY  Vidaud et al
 1990   6 pat  Weich et al
 1991  11 pat  Grundy et al
 1991   4 mat  Lindenbaum et al
 1991  14 mat  Temple et al
 1991  14 pat  Wang et al
 1991  15 pat  Malcolm et al
 1992  16 mat  Benett et al
 1993  21 pat  Blouin et al
 1993  16 pat  N'Go et al
 1994  22 mat  Schinzel et al
 1994   5 pat  Brzustowicz et al
 1994   7 pat  Höglund et al
 1995  2 mat  Harrison et al
 1995  10 mat  Jones et al
 1995  13 mat  Stallard et al
 1995  13 pat  Slater et al
 1995  22 pat  Miny et al
 1996   8 pat  Benlian et al
 1996   6 mat  Van den Berg Loonen
 1997   1 mat  Pulkkinen et al
 1997   8 mat  Piantadina et al
 1997   9 mat  Sulisalo et al
 1997   X mat  Quan et al
 1998   1 pat  Gelb et al
 1998  20 mat  Chuboda et al
 1999  17 mat  Genuardi et al
 2002   2 pat  Thomson et al
 2002  12 mat  Von Eggling et al

EE july 2003

Both these slides show the pace at which these uniparental pairs were uncovered since the first ones were identified.

We only see a few in the first decade following publication of the concept. Many more are documented in the 5 years from 91 to 95 and still quite a few are observed in the last 7 years till now, to the best of my knowledge.

It will be interesting to see which others will be detected in the forthcoming years to finally assume that those never seen are, may-be, lethal.


Slide 25

I would like to devote the rest of my talk to some peculiar machanisms of UPD formation. I have selected these examples because, to me, they illustrate some incredible twists of Nature.

I first aim at showing the role of some so-called non homologous or homologous Robertsonian translocations or centric fusions of acrocentric chromosomes.

You see here, at first glance, a non-homologous balanced translocation which, through an adjacent meiotic separation, produces a disomic gamete. This segregant, upon fertilization, generate a trisomic conceptus. If UPD must result, of two possible new hits, one will take off the singly inherited number, leaving behind a UPD pair made of one free and one attached acrocentric chromosome.

According to Lisa Shaffer and colleagues, this will happen in 0,6 % or so of prenatally diagnosed Robersonian translocations but the toll will rise to about 4 % when prospecting a cohort of phenotypically abnormal carriers.
On the other hand, two thirds of the bearers of homologous centric fusions will display a uniparental pair for the involved number.


Slide 26

On ths slide, precisely, a pattern of homologous centric fusion for chromosome 22 is found in a woman who aborts ten times in a row before producing a normal female offspring who, in turn, in due time will abort seven times.


Slide 27

Dealing in more details into this siuation, we see that eggs with the segregation of this homologous centric fusion can, upon fertilization, only produce monosomic or trisomic 22 inviable abortion products !
The only healthy offspring must have resulted from gamete complementation or, more likely, from the very early embryonic loss of parental 22. Such a luck in this case will not occur at the next generation in spite of 7 trials ending in as many abortions.


Slide 28

In the next example, an homologous 13/13 centric fusion or an isochromosome 13q (or an isodicentric 13) is found in a balanced woman withot a maternal 13. She thus examplifies a case of paternal UPD 13.

She, in turn, produces a balanced male offspring born after 5 spontaneous abortions. This balanced offspring carries the same 13/13 fusion as his mother, thus harboring a maternal UPD 13 without a traceable paternal13.

Here, amazingly, UPD 13 has taken place over two generations, once of paternal and once maternal origin, while the other parental 13 has not made its way in the embryonic cells ! A true miracle !


Slide 29

On this slide, we see how a parental 13/14 evidently non homologous Robertsonian translocation in a father ends up into an isochromosome 14 in a son with the syndrome of maternal UPD 14! An adjacent segregation has resulted into a nullisomy 14 in a paternal gamete whose deletion has been apparently patched up by duplication of the maternal 14 into an isochromosome, after fertilization.


Slide 30

And, since isochromosomes for acrocentrics have now just been mentioned, let me show, again from literature, some examples of UPD resulting from the presence of two isochromosomes per balanced individual genomes, namely one for each arm of a biarmed chromosome such a number 1, 2 (twice), 4, 7 or 9. Besides, in the case of Eggerding et al, the short arm isochromosome 7 was paternal ad the isochromosome for the long arm was maternal in origin. Most remarkable, is not it?


Slide 31

Sometime the UPD does not involve the whole of a chromosome and remains confined to a segment of a pair as it arises from a somatic crossing over between two homologous non-sister chromatids. When interstitial, the segmental UPD results from two symmetrical breaks, which are shown here as the result of an interchromatid kiss ! Mitotic segregation of the duplicated chromosomes, thereafter leads to mosaicism with one native and one reshuffled balanced cell line.


Slide 32

In other instances the segmental UPD is terminal and results from a single symmetrical break in each of two homologous non-sister chromatids, as seen here. Mosaicism involving two somatic cell types also results from this.


Slide 33

On this slide are presented examples of both types of segmental UPD, terminal or interstitial, as found for various chromosomes, 4, 6, 7, 11, 14, 20.

Some were discovered because of reduction to homozygosity causing recesive traits, while others involved imprinted domains and disrupted them.


Slide 34

Here are briefly reviewed some modes of UPD formation for more commonly affected member chromosomes.


Slide 35

This summary slide attempts to compile the information developed in this presentation.


Slide 36

This slide shows a source to find more information from a book written witn my friend and Colleague Stylianos Emmanuel Antonarakis which was published in 2002 by Liss-Wiley in New York.


Slide 37

My last slide is a symbol of my indebtedness to the many Authors who gave so much life to so simple an idea.

In this composite picture the dwarf sitting on the shoulders of the giant is the personn who sees the farthest.


My thanks go to Mr. Jean-Claude Malgouyres for assistance in preparing the graphic material for this lecture.




Written2003-12Eric Engel
Department of Medical Genetics, Development, University of Geneva, Geneva, Switzerland

© Atlas of Genetics and Cytogenetics in Oncology and Haematology
indexed on : Tue Jan 17 17:44:14 CET 2017

Home   Genes   Leukemias   Solid Tumours   Cancer-Prone   Deep Insight   Case Reports   Journals  Portal   Teaching   

X Y 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 NA

For comments and suggestions or contributions, please contact us