DNA Polymorphism Analysis in Families with Recurrence of Free Trisomy 21

Constantinos G. Pangalos1, C. Conover Talbot, Jr.2, John G. Lewis2, Patricia A. Adelsberger2, Michael B. Petersen2, Jean-Louis Serre3, Marie-Odile Rethoré1, Marie-Christine de Blois1, Philippe Parent4, Albert A. Schinzel5, Franz Binkert5, Joelle Boue6, Elisabeth Corbin7, M. F. Croquette8, Simone Gilgenkrantz9, Jean de Grouchy10, M. F. Bertheas11, Marguerite Prieur1, Odile Raoul1, Francoise Serville12, J. P. Siffroi13, Francois Thepot14, Jerome Lejeune1, and Stylianos E. Antonarakis2

Am J. Hum. Genet. 51:1015-1027, 1992

Résumé :

• Introduction


• Patients, Materials, and Methods
• Patients
• Cytogenetic Analysis
• DNA Polymorphism Analysis


• Results and Discussion
• Category I Families
• Category 2 Families
• Category 3 Families
• Concluding Remarks








• Annexes

Haut

Introduction

Trisomy 21 is the most common known genetic cause of mental retardation. Free trisomy 21 accounts for approximately 95 % of cases (Giraud and Mattei 1975). Free trisomy 21 in more than one sibling occurs rarely; the recurrence risk in families with one affected child was estimated to be 1 % - 2 %, on the basis of empiric or prenatal data (Mikkelsen and Stene 1979; Daniel et al. 1982). The frequency of trisomy 21 in second-degree (uncle--aunt/nephew-niece combinations or third-degree (first cousins) relatives is not firmly established (Tamaren et al. 1983; Abuelo et al. 1986; Eunpu et al. 1986).

Possible explanations for the recurrence of free trisomy 21 in families include: (i) parental mosaicism, as has been reported by Harris et al. (1982), Uchida and Freeman (1985), and Nielsen et al. (1988); (ii) a genetic predisposition or factor that favors nondisjunction (for discussion of this hypothesis, see Alfi et al. 1980; Yokohama et al. 1981; De Voto et al. 1985); (iii) environmental predisposing factors, including hypothyroidism, toxins, etc. (Epstein 1989); (iv) the presence of double nucleolus organizing regions in acrocentric chromosomes (Jackson-Cook et al. 1985), however, this hypothesis has not been confirmed by other laboratories (Schwartz et al. 1989); and (v) chance alone.

Analysis of DNA polymorphisms on the long arm and the pericentromeric region of chromosome 21 in families with trisomy 21 can be used to establish the parental origin and meiotic stage of nondisjunction (Antonarakis et al. 1991, 1992; Sherman et al. 1991) in almost all families. The analysis of DNA polymorphism in the rare families with more than one individual affected with free trisomy 21 (47,+ 21) can provide valuable information far any possible genetic predisposing factor. In this paper we have collected and studied 22 families with more than one individual affected with free trisomy 21. The parental and meiotic origin of nondisjunction has been determined in all cases. The interpretation of the results is discussed below.

Haut

Patients, Materials, and Methods

Haut

Patients

A total of 22 families, each with more than one individual affected with free trisomy 21, were included in the study. Figure 1 shows the pedigrees of the families studied. The families were divided into the following three categories: (i) There were 12 nuclear families, each with two affected siblings (families RDS-01-RDS-O5, RDS-O7-RDS-11, RDS-13, and RDS-14) and one family (RDS-06) with three affected siblings. (ii) In four additional families (families RDS15-RDS-18), the individuals with trisomy 21 were second-degree relatives. They were all related through a male individual who was the father and sibling of the patients. (iii) In five families (families RDS-19-RDS-23), the individuals with trisomy 21 were third degree relatives, that is, they were offspring of siblings. The families of all three categories were collected from cytogenetic laboratories in France and Switzerland. The ages of mothers and fathers at the time of birth of each individual with Down syndrome are included in table 1.

Haut

Cytogenetic Analysis

Chromosomal analysis of blood lymphocytes was performed on all individuals with trisomy 21 and on their parents. Chromosome banding was performed by the RHG or GTG technique (Dutrillaux and Lejeune 1971; Seabright 1971). In order to detect mosaicism, a total of 30 metaphases per individual with trisomy 21 were analyzed, as well as 200 metaphases in each of their parents. Mosaicism in our sample is defined as the presence of at least two trisomic cells in 200 metaphases examined. No cytogenetic heteromorphisms were studied since a considerable number of DNA polymorphisms were analyzed, including several pericentromeric markers (see below).

Haut

DNA Polymorphism Analysis

The parental origin of the supernumerary chromosome 21 and the meiotic stage of the nondisjunction were detected by using DNA polymorphic markers. The following DNA polymorphisms were used after Southern blot hybridization (Southern 1975): D21S13, D21S110, D21S11, D21S8, D21S111, D21S82, D21S3, D21S112, D21S113, MX1, and COL6A1. Description of the probe-enzyme combinations, the detection method of these polymorphic markers and their mapping position on the long arm of human chromosome 21 can be found in Warren et al. (1989) and Petersen et al. (1991b). In addition, the following DNA polymorphisms were used after PCR amplification (Saiki et al. 1915) and detection of the alleles due to short sequence repeats (SSR) by PACE: D21S215 (21GT14) Warren et al. (1992), D21S120 (Burmeister et al. 1990), D21S192 (Van Camp et al. 1991), D21S213 (21-GTO5) and D21S212 (21-GT10) (A. C. Warren and S. E. Antonarakis, unpublished data), D21S210 (21-CT12) Warren et al. (1992), IFNAR (McInnis et al. 1991), D21S156 (Lewis et al. 1990), and HMG14 (Petersen et al. 1990). All of these polymorphisms are due to (GT)n dinucleotide repeats, except IFNAR, which is due to a (TAAA)n repeat in the poly(A) tail of an Alu sequence. Description of the detection method and the scoring of polymorphic alleles per family can be found elsewhere (Petersen et al. 1991a). Several of the markers used are considered pericentromeric in the long arm of the chromosome (Antonarakis et al. 1992). These markers are D21S215, D21S120, D21S13, D21S192, and D21S172. The last four markers show approximately 6 % recombination with a rare chromosome 21-specific polymorphism of alphoid sequences (Jabs et al. 1991), while marker D21S215 shows no recombination with the alphoid polymorphism (Warren et al. 1992). All pericentromeric markers can be used to determine the meiotic origin of nondisjunction. Not all the DNA markers were determined in all families.

Haut

Results and Discussion

Table 1 presents the genotypes of the DNA polymorphisms examined in the members of all families. This table also presents the parental and meiotic origin of each trisomy 21 and the presence of crossovers in chromosomes 21 that participated in nondisjunction.

Haut

Category I Families

There are 13 families that belong to the first category, in which the individuals with trisomy are siblings (preliminary analysis of families RDS-01 and RDS-02 has been described by Pangalos et al. 1988). Cytogenetic analysis showed that, in 3 of these 13 families, there was parental mosaicism in blood leukocytes. In families RDS-09 and RDS-10 there was a 2 % maternal mosaicism for trisomy 21 (46,XX/47,XX +21), while in family RDS-02 there was a 2 % paternal mosaicism for trisomy 21 (46,XY/47,XY + 21). Analysis of DNA polymorphisms confirmed that the parental origin of the trisomy 21 in individuals DS1 and DS2 of family RDS-02 and individual DS2 of family RDS-10 was from the parent with the mosaicism. Moreover, in family RDS-09, the DNA analysis revealed a chromosome 21 present in the individual with trisomy 21 that was not detected in the DNA of blood from either parent (see markers D21S82 and D21S112 in table 1 and D21S212 in table 1 and fig. 2), confirming the results from the cytogenetic analysis. Furthermore, analysis of DNA polymorphisms revealed potential mosaicism in two other families (RDS-13 and RDS-14) of this category, since polymorphic alleles for chromosome 21 markers found in the offspring with trisomy 21 were not present in the parents (see markers D21S156 for family RDS-13 and D21S112 for family RDS-14 in table 1; the mosaicism in family RDS-14 is probably maternal, since the "new" polymorphic allele for DNA marker D21S112 was not found in the paternal grandparents). Mosaicism that remains undetected by cytogenetic analysis can therefore be recognized after DNA analysis; however, there are cases in which mosaicism was only detected by cytogenetic analysis and was not confirmed by DNA analysis. Theoretically, mosaicism can never be detected by DNA analysis if the polymorphic alleles in the parental trisomic cells are identical.

It is of interest to note that the mosaic individuals in families RDS-09, RDS-13, and RDS-14 are probably themselves the products of meiotic nondisjunction, since they each have, in some of their cells, three different alleles at certain chromosome 21 loci tested. Far example, the data suggest that the mother in family RDS-09 contains, in some of her cells, alleles 2, 3, and 5, for polymorphic marker D21S212 (see table 1).

In summary, in the set of 13 nuclear families with two siblings with trisomy 21, we detected five cases (38 %) of parental mosaicism as the cause of the recurrence of trisomy 21. The mean maternal age for the first offspring with Down syndrome in these five families was 28.6 years, and the mean paternal age was 29.3 years. The presence of parental mosaicism for trisomy 21 has been previously shown to occur in families with more than one sibling with free trisomy 21 (Harris et a1.1982; Uchida and Freeman 1985; Nielsen et al. 1988). In those studies, the mean maternal age far the first offspring with Down syndrome in 11 families with more than one offspring with Down syndrome and parental mosaicism was 24.6 years.

In eight category 1 families, parental mosaicism was not detected. In these eight families there were 17 individuals with trisomy 21. The parental origin of the extra chromosome 21 in all 17 cases with Down syndrome was maternal. In all families (RDS-O1, RDS03--RDS-O8, and RDS-11 ), there was concordance of the parental origin of the trisomy 21 for the affected siblings. The mean maternal age for the first affected offspring in these eight families was 33.6 years, and the mean paternal age was 34.5 years. These ages are not different from the mean maternal or paternal ages in large series of families with trisomy 21. Analysis of pericentromeric DNA polymorphisms revealed that the nondisjunction had occurred in maternal meiosis I in 10 (58.8 %) of the 17 cases, while in 7 (41.2 %) of the 17 cases the error had occurred in maternal meiosis II. There is an excess of meiosis-II errors in this small sample compared with the observed 23 % among maternal meiosis errors in the study of 200 families with one child with free trisomy 21 (Antonarakis et al. 1992); however, the difference is not statistically significant (?2 = 2.48). In three families (RDS-03, RDS-06, and RDS-07), all offspring with Down syndrome were the result of meiosis I error, while in two families (RDS-04 and RDS-11) both offspring with Down syndrome were the result of meiosis II error. However, in the remaining three families (RDS-01, RDS-05, and RDS-08), the trisomy 21 in the first affected individual was due to meiosis I error, and the trisomy in the second affected individual was the result of meiosis II error. There are two families with affected DZ twins (or, in theory, polar-body twins). In one family (RDS-04) the trisomy 21 in both affected twins was due to maternal meiosis II errors. In the other family (RDS-08) the trisomy 21 in one affected twin was the result of maternal meiosis I error, while the trisomy 21 in the second affected twin originated from a maternal meiosis II error (see DNA marker D21S215 of table 1). In all seven cases with maternal meiosis II errors, crossovers have been observed in the chromosomes 21 that participated in nondisjunction. These results exclude the possibility of postzygotic (mitotic) error as the cause of these trisomies. In 9 of the 10 cases of maternal meiosis I errors in which enough DNA polymorphic markers on the long arm of chromosome 21 have been studied, crossovers have been observed in four cases, while in the remaining five cases no crossovers have been detected. This is in agreement with the proposed hypothesis of reduced recombination in meiosis I in trisomy 21 (Warren et al. 1987), which has been subsequently confirmed by Sherman et al. (1991) and Antonarakis et al. (1992).

In summary, in these eight families with two affected siblings and no paternal mosaicism, there is no apparent difference from the usual families with one affected child. We therefore presume that the recurrence of individuals with trisomy 21 in the same nuclear family is the result of chance alone. Assuming that the frequency of trisomy 21 in the population is 1/700 liveborn, we expect that 1/490 000 families with two children will have two affected individuals with trisomy 21 by chance. In conclusion, this study suggests that parental mosaicism is an important and frequent cause of recurrent trisomy 21 in nuclear families, since it has been found in about 40 % of the families; however chance alone accounts for the remaining 60 % of the families.

Haut

Category 2 Families

In this category of four families, the individuals with Down syndrome are second-degree relatives. In all of these families the individuals with trisomy are related through a male individual (families RDS-15-RDS-18 of fig. 1). In families RDS-16 and RDS-17, the parental origin of nondisjunction in the Down syndrome of the third generation (designated "DS2" in the appropriate pedigrees in fig. 1) was maternal, and, therefore, the nondisjunction originated in unrelated individuals in those pedigrees. Data on the meiotic origin of nondisjunction in these families are included in table 1.

In families RDS-15 and RDS-18 the parental origin of nondisjunction in the Down syndrome of the third generation (designated DS3 for pedigree RIDS-15 and DS1 in pedigree RDS-18 in fig. 1) was paternal, and, therefore, the nondisjunction apparently originated in individuals in those pedigrees who were related. In these fathers (individuals "Fa" of pedigree RDS-15 and "Fa" of pedigree RDS-18 in fig. 1), no mosaicism has been observed either by cytogenetic or DNA analysis. The relationship of the origin of nondisjunction in families RDS-15 and RDS-18 can he attributed to chance alone; however, the fact that paternal nondisjunction for trisomy 21 is rare (about 5 %; Antonarakis et al. 1991; Sherman et al. 1991) in the general population suggests that these two families may be different from the ordinary families with trisomy 21. In family RDS-1 S, DNA polymorphism analysis ofpericentromeric markers showed that the paternal nondisjunction of individual DS3 occurred in the second meiotic division. Further analysis of DNA polymorphisms in 21q suggested that there was no recombination in the chromosomes that participated in nondisjunction. The presence of two chromosomes identical at all polymorphic loci analyzed that originate from one parent can be explained by (i) meiosis II error without a crossover event in the preceding meiosis I; (ii) paternal mosaicism that has not been discovered by the cytogenetic and DNA analysis, or (iii) mitotic error. In the last case the origin of trisomy 21 is somatic, involving the paternal chromosome. The paternal chromosome 21, which is present in two copies in individual DS3 of family RDS-15, is identical at the pericentromeric region to one of the grandmaternal chromosomes that participated in the maternal nondisjunction that causes trisomy 21 in the monozygotic twins DS1 and DS2. In family RD-18 the trisomy 21 in individual DS1 was due to an error in meiosis I in the paternal germ cells. DNA was not available from all members of this pedigree in order to study the nature of the chromosomes 21 that participated in the two nondisjunction events.

Haut

Category 3 Families

In this category of five families, the individuals with Down syndrome are third-degree relatives, that is, their parents are siblings. In pedigrees RDS-19, RDS-22, and RDS-23 the parents of the individuals with Down syndrome are brothers and sisters, while in pedigrees RDS-20 and RDS-21, the parents of the individuals with Down syndrome are brothers (see fig. 1). In four pedigrees, namely RDS-20-RDS-23, the parents in which nondisjunction had occurred were not blood relatives, and, therefore, the occurrence of two individuals with Down syndrome in these extended pedigrees can be attributed to chance. In family RDS-19 the parents in whom nondisjunction had occurred were a brother and sister. The analysis of pericentromeric DNA markers in this pedigree showed that the error for individual DS1 was in maternal meiosis I, while for individual DS2 the error was in paternal meiosis II. A mitotic error in the latter case has been excluded, since crossover events have been detected in chromosomes 21 that participated in the paternal nondisjunction. Although a predisposing factor to nondisjunction cannot be excluded in this family, chance alone also can be the explanation of the recurrent Down syndrome. It is of interest that, among the nine individuals with Down syndrome studied in this category, there was an excess of paternally derived trisomy 21 (two of nine cases), but the sample is too small to derive any conclusions.

Haut

Concluding Remarks

The aim of the study was to detect a possible genetic predisposing factor in trisomy 21. We therefore chose and collected 22 families with two affected individuals, in order to maximize the possibility of detecting such a genetic predisposition by using the powerful and unequivocal analysis of DNA markers on chromosome 21. With the exception of parental mosaicism in the relatively small sample studied, no other major genetic predisposing factor has been identified, and chance alone seems to be the main reason for the recurrence of free trisomy 1 within families.

Figure I. - Pedigrees of families with two or more individuals with free trisomy 21. Blackened squares indicate males affected with Down syndrome. Blackened circles indicate affected females. Unblackened squares and circles indicate unaffected males and females, respectively. Numbers inside unblackened symbols indicate total number of unaffected siblings. The diamond enclosing a square indicates a male fetus, and the diamond enclosing a circle indicates a female fetus. Small dots indicate abortuses or miscarriages, for which no other information is known. Fa = father; Mo = mother; DS = Down syndrome; NS = normal sibling. The abbreviations for the origin of the supernumerary chromosome 21 are as in the legend to table 1.

Figure 2. - Representative autoradiogram of the study of the origin of the extra chromosome 21 in individuals with Down syndrome. The alleles for DNA dinucleotide repeat marker D21S212 (21-GT10) are shown. The father (Fay) has alleles 1 and 2; the mother (Mo) shows alleles 3 and 4. The first offspring with Down syndrome (DS1) has alleles, 2, 3, and 5, while the second offspring with Down syndrome (DS2) has alleles 1, 3, and 5. Allele 5 comes from the mother, who cytogenetically shows mosaicism for trisomy 21.

Haut

References

Abuelo D, Darsel-Bowers G, Busch W, Pueschel S, Pezzullo J (1986) Risk for trisomy 21 in offspring of individuals who have relatives with trisomy 21. Am J Hum genet 25:365-369

Alfi O, Chang R, Azen S (1980) Evidence for genetic control of nondisjunction in man. Am J Hum Genet 32:477-483

Antonarakis SE, Petersen MB, McInnis MG, Adelsberger PA, Schinzel A, Binkert F, Pangalos G, et al (1992) The meiotic stage of nondisjunction in trisomy 21: determination using DNA polymorphisms. Am J Hum genet 50:544-550

Antonarakis SE, and the Down syndrome collaborative group (1991) Parental origin of the extra chromosome in trisomy 21 using DNA polymorphism analysis. N Engl J Med 324:872-876

Burmeister M, Cox DR, Myers RM (1990) Dinucleotide repeat polymorphism located at D21 S120. Nucleic Acids Res 18:4969

Daniel A, Stewart L, Saville T, Brookwell R, Paull H, Purvis-Smith S, Lang-Po-Tang PR (1982) Prenatal diagnosis in 3000 women far chromosome, X-linked and metabolic disorders. Am J Med Genet 11:61-75

Devoto M, Prosperi L., Dagna-Bricarelli F, Coviello DA, Groci G, Zelante L, Ferranti G, et al (1985) Frequency of consanguineous marriages among parents and grandparants of Down syndrome patients. Hum Genet 70:256-258

Dutrillaux B, Lejeune J (1971) Sur une nouvelle technique d'analyse du caryotype humain. C R Acad Sci 273:2638-2640

Epstein CJ (1989) Down syndrome (trisomy 21). In: Scriver CR, Beaudet AL, Sly WS, Valle D (eds) The metabolic basis of inherited disease, 6th ed. McGraw-Hill, New York, pp 291-326

Eunpu DL, McDonald DM, Zackai EH (1986) Trisomy 21: rate in second degree relatives. Am J Med Genet 25:361-363

Giraud F, Mattei JF (1975) Aspects epidemiologues de la trisomy 21. J Genet Hum 23:1-30

Harris DJ, Begleiter ML, Chamberlin J, Haukins L, Magenis RE (1982) Parental trisomy 21 mosaicism. Am J Hum Genet 34:125-133

Jabs EW, Warren AC, Taylor EW, Colyer CR, Meyers DA, Antonarakis SE (1991) Alphoid DNA polymorphisms for chromosome 21 can be distinguished from those on chromosome 13 using probes homologous to both. Genomics 9:141-146

Jackson-Cook CK, Flannery DB, Corey LA, Nance WE, Brown JA (1985) Nuclear organizer region variants as a risk factor for Down syndrome. Am J Hum Genet 37:1049-1061

Lewis JG, Weber JL, Petersen MB, Slaugenhaupt SA, Kwitek A, May PE, Warren AC, et al (1990) Linkage mapping of the highly informative DNA marker D21S156 to human chromosome 21 using a polymorphic GT dinucleotide repeat. Genomics 8:400-402

McInnis MG, Lutfalla G, Slaugenhaupt SA, Petersen MB, Uze G, Chakravarti A, Antonarakis SE (1991) Linkage mapping of highly informative polymorphisms of the human interferon alpha receptor gene on chromosome 21. Genomics 11:573-576

Mikkelsen M, Stene J (1979) Previous child with Down syndrome and other chromosome aberration: group report. In: Murken JD, Stengel-Rutkowski S, Schwinger E (eds) Prenatal diagnosis: proceedings of the 3rd European Conference on Prenatal Diagnosis of Genetic Disorders. Enke, Stuttgart, pp 22-29

Nielsen KG, Poulsen H, Mikkelsen M, Steuber E (1988) Multiple recurrence of trisomy 21 Down syndrome. Hum Genet 78:103-105

Pangalos C, Serre JL, Ghica M, Abazis D, Sinet PM, Rethore MO, Lejeune J (1988) Molecular analysis of the parental origin of trisomy in two families with two children having regular trisomy 21. Ann Genet 31:151-154

Petersen MB, Economou EP, Slaugenhaupt SA, Chakravarti A, Antonarakis SE (1990) Linkage analysis of the human HMG14 gene on chromosome 21 using a GT dinucleotide repeat as polymorphic marker. Genomics 7:136-138

Petersen MB, Schinzel AA, Binkert F, Tranebjaerg L, Mikkelsen M, Collins FA, Economou EP, et al (1991a) Use of short sequence repeat DNA polymorphisms, after PCR amplification to detect the parental origin of the additional chromosome 21 in Down syndrome. Am J Hum Genet 48:65-71

Petersen MB, Slaugenhaupt SA, Lewis JG, Warren AC, Chakravarti A, Antonarakis SE (1991b) A genetic linkage map of 27 markers on human chromosome 21. Genomics 9:407-419

Saiki RK, Scharf S, Faloona F, Mullis KB, Horn GT, Erlich HA, Arnheim N (1985) Enzymatic amplification of betaglobin genomic sequences and restriction site analysis for diagnosis of sickle cell anemia. Science 230:1350-1354

Schwartz S, Roulston D, Cohen MM (1989) dNORs and meiotic nondisjunction. Am J Hum Genet 44:627-430

Seabright M (1971) A rapid banding technique for human chromosomes. Lancet 2:971-972

Sherman SL, Takeasu N, Freeman SB, Grantham M, Phillips C, Blackston RD, Jacobs PA, et al (1991) Trisomy 21: association between reduced recombination and nondisjunction. Am J Hum Genet 49:608-620

Southern EM (1975) Detection of specific sequences among DNA fragments separated by gel electrophoresis. J Mol Biol 98:503-517

Tamaren J, Spuhler K, Sujansky E (1983) Risk of Down syndrome among second and third degree relatives of a proband with trisomy 21. Am J Med Genet 15:393-404

Uchida IA, Freeman VCP (1985) Trisomy 21 Down syndrome: parental mosaicism. Hum Genet 70:246-248

Van Camp G, Backhovens H, Cruts M, Wehnert A, Van Hul W, Stinissen P, Van Broeckhoven C (1991) Identification of chromosome 21 DNA polymorphisms for genetic studies in Alzheimer disease and Down syndrome. Hum Genet 87:49-653

Warren AC, Chakravarti A, Wang C, Slaugenhaupt SA, Halloran SL, Watkins PC, Metaxotou C, et al (1987) Evidence for reduced recombination an the nondisjoined chromosome 21 in Down syndrome. Science 237:652-654

Warren AC, Petersen MB, van Hul W, McInnis M.G, van Broeckhoven C, Cox TK, Chakravarti A, Antonarakis SE (1992) D21S215 is a (GT)n polymorphic marker close to the centromeric alphoid sequences on chromosome 21. Genomics 13:1365-1367

Warren AC, Slaugenhaupt SA, Lewis JG, Chakravarti A, Antonarakis SE (1989) A genetic linkage map of 17 markers on human chromosome 21. Genomics 4:579-591

Yokohama S, Reich T, Morgan K (1981) Inbreeding and the genetic control of nondisjunction. Hum Genet 59:125-128