Introduction

The first clinical description of a child with a de novo supernumerary isochromosome 18p appeared in 1963 [1]; more than 50 cases have since been published. The tetrasomy 18p syndrome is characterized by psychomotor retardation, microcephaly, long facies with oval shape, high-arched eyebrows, short palpebral fissures and small pinched nose. Facial asymmetry is not unusual. Scoliosis and/or kyphosis, long fingers with contractures and feet malformations are frequent [1–3].

The chromosome 18 specificity of these marker chromosomes can now be confirmed by in situ hybridization with chromosome-18-specific probes [2, 4, 5]. Although previous reports determined only the chromosome 18 derivation of the marker, it can now be unambiguously shown using an 18p-specific library, that the marker chromosome is composed of only short-arm material [6].

Determination of the parental origin of the i(18p)s has previously been hampered by the lack of cytogenetic heteromorphisms of chromosome 18. Several models have been proposed to explain the formation of isochromosomes, but to our knowledge the mechanisms leading to tetrasomy 18p have not been clarified.

We have investigated 9 individuals with de novo tetrasomy 18p and their parents at polymorphic loci to determine the parental origin and mechanism of formation of the additional i(18p) chromosome. The cases were diagnosed by conventional chromosome banding and verified as i(18p)s by fluorescence in situ hybridization (FISH) using a flow-sorted 18p-specific library.

Material and Methods

Patient Population

The 9 patients were 4 girls and 5 boys, referred to cytogenetic analysis because of dysmorphic features and mental retardation. An overview of the patients is given in table 1.

Table 1 Parental ages and clinical findings in 9 cases of de novo tetrasomy 18p
Full size table

Cytogenetic Studies

Chromosome analysis was performed on cultured lymphocytes using one or more of the following standard methods: QFQ-, CBG-, GTG-, and prometaphase RBA-banding as described [5–8].

Fluorescence in situ Hybridization

Fluorescence in situ hybridization (FISH) analysis was done using a biotin-dUTP-labeled chromosome 18 centromere-specific probe (ONCOR) according to the manufacturer’s instructions. In case 2, in situ hybridization had been performed using a radioactively labeled probe [5].

An 18p-specific library was constructed by flow sorting of an i(18p), and the flow-sorted material was amplified and biotinylated by degenerate oligonucleotide primed-polymerase chain reaction (DOP-PCR) [6]. Probe preparation, hybridization, detection and visualization were performed in all cases by one of us (E.B.) as previously described [6, 9]. An 18q-specific library was also used as control. The construction of this library, labeling, and FISH analysis were performed as above.

DNA Analysis

DNA was extracted from EDTA-anticoagulated blood by a salting-out procedure [10]. Eight DNA polymorphisms were detected after PCR amplification of genomic DNA [11]. End-labeling of primer, PCR amplification conditions, Polyacrylamide gel electrophoresis of the amplification product, and autoradiography were performed as described elsewhere [12]. The microsatellites used were (CA)n repeats at the following loci: D18S170 and D18S40 [13], D18S59, D18S63, D18S54, D18S62, D18S53, and D18S71 [14]. The scoring of polymorphic alleles was performed based on the knowledge of the karyotype showing tetrasomy 18p, i.e. an 18p microsatellite showing homozygosity for one allele in the proband would be interpreted as four copies of the same allele (1111), two alleles of almost equal intensity in the proband would be scored as 1122, two alleles with the smallest allele of much stronger intensity would be scored as 1112, and three alleles with the smallest allele of significantly stronger intensity would be scored as 1123, etc. The scoring was analogous to that of a trisomic condition, where dosage analysis by visual inspection was shown to correlate with densitometric reading of allele intensities [15].

Results

Parental Age Distribution

Range for mothers 24–39 years, mean maternal age 29 years; range for fathers 25–45 years, mean paternal age 32.1 years (table 1).

Chromosome analysis revealed 47 chromosomes in all 9 cases; the size and banding pattern of the extra small metacentric chromosome were compatible with an isochromosome of the short arm of chromosome 18. All cases were nonmosaic. All parental karyotypes were normal.

In situ hybridization with a chromosome 18 centromere-specific probe showed a monocentromeric signal on the marker chromosome in all cases. FISH using an 18p-specific library showed labeling of the short arms of chromosomes 18 and the whole of i(18p). No labeling of the marker chromosome was observed after FISH with an 18q-specific library (fig. 1).

Fig. 1
figure 1

FISH analysis using an 18p-specific library shows labeling of the p-arms of the two chromosomes 18 and the entire marker chromosome (arrow). Hybridization using an 18q-specific library shows labeling of the q-arms of the two chromosomes 18. The marker chromosome is not labeled (arrow).

Full size image

DNA analysis indicated that all nine i(18p) chromosomes were of maternal origin (table 2, fig. 2). An uniparental origin of the normal chromosomes 18 could be excluded. In 6 cases (cases 1–6), the 18p markers showed that the maternal heterozygous alleles were reduced to homozygosity in the proband in the pericentromeric re-

Fig. 2
figure 2

Determination of parental origin of additional chromosome i(18p) with the dinucleotide repeat D18S53 and D18S59 in families where the proband has tetrasomy 18p. The genotypic scoring of alleles is indicated in each individual. Fa = father; +i(18p) = tetrasomy 18p proband; Mo = mother.

Full size image
Table 2 DNA polymorphisms on the short arm of chromosome 18 in nine families where the proband has de novo tetrasomy 18p
Full size table

Discussion

In their original report, published in 1963, before the banding era, Frøland et al. [1] suggested that the extra chromosome was an isochromosome of the short arms of chromosome 17 or 18. Later, with banding an isochromosome for 18p was suspected. However, by classical chromosome analysis it is sometimes difficult to distinguish an i(18p) from a deleted chromosome 18 [4, 6–8].

In the present cases, FISH analysis with a centromere-specific probe as well as with 18p- and 18q-specific libraries, showed that the marker chromosome was monocentric, and that it was unequivocally an i(18p), since the whole of the marker was painted with the 18p-specific library, with no labeling using the 18q library (fig. 1).

Formation of a nonmosaic de novo tetrasomy 18p due to an additional monocentric isochromosome 18p requires two events: nondisjunction and centromeric misdivision (= mitotic centromeric fission with fusion of sister p-arms).

The segregation of polymorphic DNA markers in different categories of premeiotic, meiotic or mitotic errors is presented in figure 3. The DNA markers used in this study cover the entire length of the short arm of chromosome 18 [16]. In 6 cases (cases 1–6), we found that the maternal heterozygous alleles were reduced to homozygosity in the pericentromeric region in the proband but retained heterozygosity in the telomeric region (table 2). To explain those 6 cases, we propose a model in which the tetrasomy is a result of meiosis II nondisjunction followed by postzygotic misdivision of one of the chromosomes with subsequent loss of all long-arm material (fig. 3, II).

Fig. 3
figure 3

Schematic representation of the segregation of a locus A near the telomere and a locus B near the centromere, each with two alleles (A, a and B, b) in normal meiosis and following premeiotic, meiotic, or postzygotic errors. Cross-over in meiosis I is indicated with ×. Not all the products of meiosis are shown. R = Reduction to homozygosity in the individual with tetrasomy 18p, given that the polymorphism was heterozygous in the parent who contributed the supernumerary isochromosome 18p; NR = nonreduction to homozygosity in the individual with tetrasomy 18p, given that the polymorphism was heterozygous in the parent who contributed the supernumerary isochromosome 18p; MI = meiosis I; MII = meiosis II; PZG = postzygotic division; NDJ = nondisjunction; MD = centromeric misdivision.

Full size image

An alternative series of events compatible with the data is a premeiotic (mitotic) nondisjunction followed by a centromeric misdivision in MII [17].

In case 9, 6 of the 8 DNA markers were informative and showed reduction to homozygosity of the maternal alleles in the proband (table 2). Only postzygotic nondisjunction of chromosome 18 followed by misdivision of one of the chromosomes can explain these results (fig. 3, III). This assumes that no marker between the centromere and D18S40 shows nonreduction to homozygosity of maternal alleles. A postzygotic misdivision of one of the chromosomes in a maternally derived trisomy 18 may theoretically occur on the paternal chromosome in one third of the cases. Such cases would have maternal uniparental disomy for the long arm of chromosome 18. We tested this possibility with DNA markers mapping to the long arm of chromosome 18. At three informative 18q loci: D18S36, D18S35, and D18S38 biparental inheritance was demonstrated (data not shown). Usually mosaicism would be expected in cases of postzygotic nondisjunction. Analysis of 20 mitoses in blood and 59 mitoses in fibroblast cultures did not disclose any mitosis lacking the i(18p).

In 2 cases (cases 7 and 8), the probands were heterozygous for all informative maternal alleles. Unfortunately, the one or two markers, respectively, located closest to the centromere were not informative as the mothers were homozygous (table 2). If the maternal alleles in these probands were all nonreduced, the i(18p) formation could be explained by two previously postulated mechanisms: a centromeric misdivision occurring during a germinal premeiotic mitosis followed by MI nondisjunction [18] (fig. 3, IV), or an MI nondisjunction followed by a centrometric misdivision of one of the univalents during meiosis II [19] (fig. 3,I).

The proposed models for formation of the tetrasomy 18p are based on deduced haplotypes for the normal chromosomes 18 and the isochromosome 18p. However, in cases 7 and 8 alternative haplotypes could be determined from the pattern of the polymorphisms in the families. Determination of the i(l 8p) haplotypes, e.g. by PCR analysis of flow-sorted or microdissected i(18p) chromosomes would allow us to verify whether the deductions are correct or not.

Nondisjunction of chromosome 18 must be involved in the formation of tetrasomy 18p. DNA studies of trisomy 18 have shown that the majority of cases result from a maternal MII error [16]. This is in agreement with the most common model we propose, invoking MII nondisjunction as the first step in the formation of tetrasomy 18p.

Molecular analyses have also demonstrated that paternally derived trisomy 18 occurs in less than 5% of cases [16, 20]. In the present study, the isochromosome 18p was of maternal origin in all 9 cases.

Trisomy 18 is associated with advanced maternal age. Fisher et al. [16] found that the maternal age effect is significant in the MII error group. So far, an association of advanced parental age with tetrasomy 18p has not been demonstrated. In 1978, Nielsen et al. [7] found a mean maternal age of 30 years for the 25 cases of tetrasomy 18p reported previously. However, none of these papers explained the mechanisms involved in the i(18p) formation. Since we invoke that maternal MII nondisjunction is involved in several of the cases, increased maternal age would be expected. However, in the present limited number of cases, maternal age was increased in only 2 cases (cases 4 and 6, table 1).

The results of the present study serve to demonstrate that no single mechanism can account for all i(18p) cases. Based on genotyping, both premeiotic, meiotic and postzygotic division errors can be involved alone or in combination. Further studies are needed to determine the parental origin and to clarify the mechanisms of formation.