In this chapter we consider the circumstance of karyotypically normal parents who have had a child with a structural chromosome rearrangement. Under this heading, we distinguish in particular deletions (partial monosomy) and duplications (partial trisomy). If the rearrangement occurs during meiosis or at a postzygotic mitosis, we generally assume a recurrence risk no different from that of the general population. These cases arise anew—de novo—with the affected child. If, however, the rearrangement arises at a premeiotic mitosis, the parent would be a gonadal mosaic, and an increased risk for recurrence would apply. Usually, no prior distinction between these two possibilities can be made. Here we consider those deletions and duplications in which cytogenetic or molecular cytogenetic techniques are important in demonstrating the defect, and which are generally thought of as being chromosomal conditions. In some, there is a possibility for parental heterozygosity, and so our focus is not exclusively on de novo defects.
BIOLOGY
Mechanisms of Formation of Structural Rearrangement
Human chromosomes are disconcertingly dynamic structures (Piccini et al., 2001)! Megabase lengths of genomic DNA can be deleted, duplicated, and moved around, with unfortunate results. Conditions that are the consequence of this chromosomal unruliness have gained a new title: these are “genomic disorders” (Lupski, 1998, 2003). The point of distinction is that these genetic diseases are due to illegitimate recombination. Older expressions, not to be discarded, include “partial aneuploidies,” “segmental aneusomies,” and “contiguous gene disorders.” Many, although not all, are cytogenetic conditions (an example of a non-cytogenetic genomic disorder is facioscapulohumeral muscular dystrophy, in which kilobase-sized lengths of a subtelomeric repeat sequence at 4qter are deleted). The common basis for many of these rearrangements, the source of this vulnerability, lies in the existence of multiple DNA sequences throughout the genome, generally of some thousands of base pairs, which are sufficiently similar (“paralogous”) that they enable the erroneous coming together of different chromosome regions. Within the two sequences involved in a particular exchange, there is a length of perfect or near-perfect homology, and this is the site of the actual strand exchange (“non-allelic homologous recombination”). These sequences, whose misalignment during meiosis sets the stage for the illegitimate recombination between the two regions, have been dubbed “duplicons” (Ji et al., 2000; Peoples et al., 2000). Considering the case of recombination involving homologs (homologous recombination), this process can take place between homologs (interchromosomal recombination), between sister chromatids, or within the same chromosome arm (intrachromosomal recombination). The distinction may to some extent be made on an assessment of the identity or difference of alleles within the region (Roberts et al., 2002).
Numerous mechanisms can be envisaged. Asymmetric pairing of homologous chromosomes at meiosis could cause nonmatching segments to be adjacent (Chandley, 1989), as an interchromosomal event. If the mismatch remains, and the homologs recombine illegitimately, the recombinant products will be reciprocally imbalanced: one with a deficiency, the other with a duplication. Kozma et al. (1991) propose such a scenario for the 17p11.2 deletion of Smith-Magenis syndrome and the dup(17)(p11.2) syndrome (Fig. 17-1), and this mechanism has been elegantly substantiated at the molecular cytogenetic level (Potocki et al., 2000a). Complementary deletion and duplication products may also arise from sister chromatid exchange. Or, a segment may simply be looped out and excised on one arm of a chromosome, as an intrachromosomal event, producing an interstitial deletion (and thus with no countertype duplication). An intrinsic predisposition to the generation of deletion is illustrated by the male heterozygote for Bloom syndrome (p. 303): sperm of these men have a high frequency of chromosome breaks (Martin et al., 1994). There may be a predisposition to submicroscopic intrachromosomal rearrangement due to the “bouquet” association of telomeres at meiosis, whereby the short- and long-arm telomeric regions of a chromosome come to be in contact, allowing subtelomeric exchange (Daniel et al., 2003b).
|
|
|
Figure 17-1. One theoretical mechanism to produce a duplication and a deletion. Similar sequences (cross-hatched segments) exist at numerous places along the chromosome. Consider a segment between two such segments, indicated here by the black and white dots. Misalignment of the two flanking nonhomologous sequences, followed by illegitimate recombination within these sequences (×), produces recombinant products that are reciprocally imbalanced: one with a deficiency of the chromatin between the two sequences, and the other with a duplication. The deletion chromosome is shown with no black and white dot, while the duplication chromosome has a double set of dots. The general case is drawn after Chandley (1989). If this were chromosome 17, for example, Smith-Magenis syndrome and the dup(17)(p11.2p11.2) syndrome, respectively, could result (Potocki et al., 2000a; Schneider et al., 2000). Or, on a smaller scale, if the segment were the 1.4 Mb of DNA within 17p12 including the PMP22 gene, Charcot-Marie-Tooth disease and hereditary pressure-sensitive neuropathy would be the two corresponding phenotypes (see p. 286). |
Some regions have a propensity to the formation of deletions (hot spot) (Christ et al., 1999), chromosomes 15 and 22 providing particular examples. Low-copy repeat (LCR) duplicons at 15q11–q13 and 22q11 permit unequal exchange and recombination in these regions, generating deletions, duplications, triplications and extra structurally abnormal chromosomes (ESACs) in chromosome 15, and deletions, duplications, and translocations in chromosome 22 (McDermid and Morrow, 2002; Roberts et al., 2002). A hot spot within 16p13.3 causes rearrangements that delete the CBP gene, resulting in Rubinstein-Taybi syndrome (interestingly, this same hot spot may underlie the t(8;16)(p11;p13.3) translocation that, as a somatic event, leads to acute myeloid leukemia; Petrij et al., 2000a,b). A series of hot spots in the same region may underlie deletions that are therefore different at the molecular level, but indistinguishable cytogenetically and, probably, clinically. The 9p23 deletion may be an example of this (Christ et al., 1999).
A particular category of complementary dele-tion/duplication may be due to aberrant behavior at the replication fork, and here the chromosome needs to be envisaged at its basic level, comprising double-stranded DNA. During replication, the strands separate and act as templates. Replication proceeding along in the correct direction on one strand could skip across to the other, and continue in the opposite direction for a stretch, and then skip back to where it should have been. This leads to a chromosome with different-sized chromatids, one with a duplicated segment, and the other with the same segment deleted. If this should happen at the very first somatic replication following conception, two countertype cell lines arise at the two-cell stage. Tharapel et al. (1999) discuss this scenario, and describe their case of an abnormal infant who had mosaicism with the karyotype 46,XX,inv dup(11)(q23q13)/46,XX, del(11)(q13q23), the two lines being in approximately equal proportions. If this sort of misreplication should happen at the second mitosis, there will be a normal cell line as well, and Tharapel et al. illustrate this circumstance in a child initially identified at prenatal diagnosis because of a choroid plexus cyst and echogenic bowel. In this child, the normal cell line was present in about half of the cells, with the remaining cells containing either a deletion for 7p11.2–p13 or a duplication for this segment.
Respecting the requirement that integrity of the telomere be maintained, some mechanisms of terminal deletion need to include a process to restore the telomere (Ballif et al., 2000a). If the terminal deletion is interstitial, then the original telomere simply remains intact. If, however, the telomere is lost in the deletional process, a new telomere can be regenerated (“telomere healing”). If another chromosome is involved in the process, its telomere can be “captured” to fulfill the requirement. Ballif et al. studied the 1p36 deletion in particular (see below) and showed that all three mechanisms were involved, although in most of them the telomere was restored by healing (see also 18p Deletions, below).
Influence of Sex of Parent
Certain duplication/deletion rearrangements may have a predilection for happening in one or the other sex of parent. Chromosome 17p11.2 rearrangements are more often of paternal origin, and may be intrachromosomal or interchromosomal in their generation. The nearby 17p12 region is also more susceptible to rearrangement in the paternal gonad, although, in contrast to 17p11.2, paternal dupli-cation/deletions are always interchromosomal, and the uncommon maternal cases are all intrachromosomal (Potocki et al., 2000a). The X chromosome has a particular vulnerability in the male, perhaps because it is largely unpaired at meiosis, and it can refold up and down its length (Giglio et al., 2000).
Origin Pre-, Intra-, or Postmeiosis
Most de novo deletions are considered to originate at meiosis, and the child is nonmosaic. There are two other possibilities. First, the abnormality may have arisen at a premeiotic mitosis, such that the parent is a gonadal mosaic (note that the child would be nonmosaic). Thus it is usually appropriate to check the parental karyotypes, to test the possibilities that one may either be a carrier of a balanced rearrangement or a mosaic for the abnormal chromosome (the normal cell line in a phenotypically normal parent being, presumably, predominant). An example is illustrated in Figure 2-12 of an interstitial deletion del(1)(q25q31.2) that was identified at amniocentesis, which led to the discovery of 46,XY,del(1)[80%]/46,XY[20%] mosaicism on blood karyotyping of the father, thus revealing him to be a somatic–gonadal mosaic. Normal parental karyotypes do not, of course, exclude the possibility of gonadal mosaicism, as exemplified in two sisters with a chromosome 16 deletion whose parents' karyotypes were normal (Hoo et al., 1985). A fuller discussion of gonadal mosaicism is given on p. 44. The second possibility is that a re-arrangement can have arisen at a postzygotic mitosis, in which case the child may be a mosaic, generally for a normal and for the abnormal cell line. This category will have no increased risk for recurrence.
Mosaicism
Mosaicism for a structural rearrangement is rarely recognized (Gardner et al., 1994; Leegte et al., 1998; Zaslav et al., 1999). It requires the agency of a postzygotic event, and two major scenarios warrant consideration. In the first, a conceptus is chromosomally normal, and at a subsequent mitosis an abnormality is generated which gives rise to a karyotypically abnormal cell line, along with the 46,N line. The karyotype becomes 46,(abn)/46,N. In the second scenario, from an initially 47,+(abn) conceptus a postzygotic “correction” with loss of the abnormal chromosome generates a normal cell line, and so the karyotype becomes 47,+(abn)/46,N.
DELETION
Most of the “new” syndromes joining the ranks of the partial aneuploidies are due to deletions. The general karyotypic form of an interstitial deletion, in any chromosome A, is 46,del(A) (p00p00) or 46,del(A)(q00q00). We have traveled a distance from the earliest days of cytogenetics when the first deletion was published, which was large enough to be seen on a solid-stained B group chromosome, and was associated with cri du chat syndrome (Lejeune et al., 1963). Now we have a spectrum ranging from large deletions (classical cytogenetic deletion syndromes), to microdeletions detectable only since the use of high-resolution banding, to deletions beyond the range of banding but detected on combined molecular/cytogenetic (FISH) or purely molecular methodology, to deletions that are so small that only a single locus is removed. A very subtle deletion in the subtelomeric region may submit only to FISH and molecular methodology, and somewhere in the vicinity of 2%–10% of patients with unclassified multiple congenital anomalies/men-tal retardation (MCA/MR) may prove to have a subtelomeric deletion, as discussed in Chapter 2 (p. 41). A less affected parent might have the same deletion, and thus parental studies are warranted. Some apparent subtelomeric “deletions” may actually be due to normal polymorphism, and this is another reason for parental chromosomes to be studied, along with careful clinical assessments, carefully interpreted (Ballif et al., 2000a). A methodology that covers the whole chromosome is comparative genomic hybridization (CGH), and in one exploratory study the yield of abnormality (including both deletions and duplications) was greater than that achieved by subtelomeric FISH (Ness et al., 2002).
Contiguous Gene Syndrome
Recollect that loci are arranged in linear order along a chromosome. Often there is no apparent reason for the order: the nonsignificance of the contiguity of two loci has been likened to the unimportance one would attach to Appalachian mountains being next to Apple in an encyclopedia. Our genome differs from an encyclopedia in that about a third of all the entries relate to one topic: development of the brain. Many of the other entries (loci) relate to the control of morphogenesis during embryonic life. If a length of chromosome is deleted, a sequence of adjacent (contiguous) genes will be lost. The phenotype resulting from this can be described as a contiguous gene syndrome (Tommerup, 1993), and other descriptive terms have been noted above. In almost any deletion detectable cytogenetically, some of the deleted loci will be brain loci, while others could be for anything, but probably including some morphogenesis loci. Thus, we have the classic clinical picture in deletion syndromes of intellectual deficit of some degree, dysmorphism, and organ malformation. The deletion produces a monosomy, or haploinsufficiency,1 for the region of the chromosome that has been removed, and loci in this segment are under-expressed. Proof that genetic expression is reduced by 50%, in the case of the 18q–syndrome at least, was adduced by Wang et al. (1999c) in measuring mRNA from a number of 18q loci.
Some of the loci whose haploinsufficiency contributes to the phenotype in the various deletion syndromes are beginning to be defined, as noted in individual entries in the Genetic Counseling section. It seems likely that many such genes will have their untoward outcome not in a simple one-to-one relationship with a single gene product, but rather in a complex layering and interlacing of consequential effects. As yet, however, it is only the simple case that we can begin to understand—such as, for example, the brain white matter abnormality of the 18q–syndrome that is presumably a direct consequence of the loss of a structural myelin gene on 18q.
An infrequent mechanism for abnormality is that the deletion may “unmask heterozygosity.” Ikegawa et al. (1998) suggest this in a man they studied with pseudoachondroplasia and a del(11)(q21q22.2). Pseudoachondroplasia is a form of dwarfism, due to a recessive gene. Although one known gene for this condition is on chromosome 19, there remains the possibility of genetic heterogeneity, and there may be another similar locus on chromosome 11. If that were so, one chromosome 11 might have a mutated gene, and on the other chromosome 11 the normal gene might have been lost because of the deletion. Equally, the association may have been coincidental. A firmer case can be made for the child with Wilson syndrome and concomitant 13q–, noted on p. 283.
Simple and Complex Deletions
The simple scenario of a “clean-cut” deletion may in some instances be an oversimplification. Davies et al. (2003) restudied a group of 16 deletion patients, and in three the supposed deletion proved to be a rearrangement, involving subtelomeric regions. Gunn et al. (2003) studied a child initially karyotyped as 46,XY but whose clinical features suggested an 18q deletion. This was indeed proved, but the deletion seemed rather small given the severity of the phenotype. Using FISH and microarray analysis, they could show that a segment from distal 4q had been inserted into the site of the 18q deletion, giving a partial 4q trisomy along with the partial 18q monosomy.
Gene Discovery
Deletions can point the way to discovery of genes coding for particular organs and tissues. A deletion that, from clinical observations, has a particular clinical association with clefting, may remove, say, contiguous genes w, x, y, and z. It could reasonably be imagined that hap-loinsufficiency of one of these genes, e.g., gene x, could contribute to the cause of cleft lip. It could then further be assumed that the normal role of gene x is to contribute to the process of lip formation during early embryogenesis. Brewer et al. (1998) have reviewed some hundreds of small deletions listed in the Oxford Cytogenetic Database, correlating the malformations with which these deletions have been associated. Some patterns have emerged: some deletions seem particularly likely to lead to a heart defect, while others may be prone to cause clefting. Researchers hoping to find genes directing development of the heart, or genes controlling lip formation, could focus their searches in these chromosomal regions. Similarly, nephrogenesis genes and neurogenesis genes may come to light in the analysis of deletions associated with renal maldevelopment and epilepsy, respectively (Amor et al., 2003; Singh et al., 2002a). Taking a closer focus, a child with a chromosomal deletion and having a known multiple malformation syndrome that has otherwise been associated with a normal karyotype may point toward the location of a putative gene or genes. For example, Rauen et al. (2000) make such a proposition in the case of their patient with cardiofaciocutaneous syndrome and a del(12)(q21.2q22), subsequently supported by their report of a similarly affected patient with a deletion at the same site (Rauen et al., 2002).
DUPLICATION
Duplicated segments may arise from within the same chromatid, from the sister chromatid, from the same arm, from the other arm, or from a different chromosome. As with the deletion, the association of different duplicated segments with particular phenotypes offers an insight into which regions of the genome may harbor specific critical genes (Brewer et al., 1999).
Direct and Inverted Intrachromosomal (Tandem) Duplication
The duplication comprises chromatin of the same chromosome, the original and the duplicated segments being ordered in tandem fashion. If the linear orientation of a chromosome A is maintained, the rearrangement is a direct duplication, 46, dir dup(A); if it is reversed, it is an inverted duplication, 46, inv dup(A). Possible mechanisms include mismatched pairing of homologs or of the chromatids of one homolog, which, after recombination or sister chromatid exchange, can produce countertype duplications and deletions, respectively (Van Dyke, 1988). Hoo et al. (1995) discuss the particular case of the inverted terminal duplication due to U-loop reunion following a two-strand break in a telomeric segment. These inverted duplications may have an accompanying deletion: thus, an “inv dup/del” rearrangement, and chromosome 8p is disproportionately represented in this category (Minelli et al., 1993; Guo et al., 1995; Kondoh et al., 2003). In an analysis of 20 cases of de novo tandem duplications, Kotzot et al. (2000b) found two-thirds to be due to a duplication of sister chromatids, with the remainder involving non-sis-ter chromatids. The parental origins were equally maternal and paternal. Compared with other rearrangements, a possibly higher frequency of mosaicism in the dir dup suggests that postzygotic mitosis is the setting in which some arise (Gardner et al., 1994; de Silva et al., 1998), although Kotzot et al. do consider the majority to be due to meiotic error. The particular case of the dup(15)(q11q13) is dealt with on p. 290.
One example most likely due to postzygotic sister chromatid exchange is seen in the patient described in Faivre et al. (2000b), a child with a condition resembling Sotos syndrome and having mosaicism for a dup(20)(p11.2p12.1) in 23% of cells. On checking DNA markers that mapped to the region of the duplication in 20p, each parent had contributed just one allele. The parsimonious explanation is that a single event of unequal sister chromatid exchange happened at a mitosis, in an initially normal conceptus, generating the dup(20p) line, with the complementary deletion being lost (Fig. 17-2).
Molecular cytogenetic techniques bring a powerful focus on the mechanisms underlying these types of rearrangement. A good example is given by Bonaglia et al. (2000), who report a de novo inv dup (2q) in a retarded and dysmorphic child. The use of multicolored FISH enables an appreciation, at first glance, that a particular 2q region is duplicated and that it is inverted, and that another part of 2q is deleted. The parental origin of this abnormal chromosome was determined by analyzing DNA markers within the regions of interest. A double amount of the maternal alleles being tested within the region 2q33–q36 indicated duplication. More distally, two markers from 2q37 showed hemizygosity, with no maternal allele present, thus revealing a deletion here (too small to be seen cytogenetically). From this, an interpretation could be offered about the sequence of events leading to the dup/del defect. “Inverted duplicons” exist on the chromosome, regions with the same or very similar DNA sequences but running in opposite directions. Refolding of the chromatids in one chromosome at meiosis I enables a coherent apposition of two duplicons to take place, and then recombination between them. This complexity is “resolved” by a rupture of the recombinant, to produce the dup/del product, as well as a del chromosome and a small fragment. It is likely that many “inv dup” chromosomes arise by this mechanism and, as noted above, are actually inv dup/del chromosomes.
|
|
|
Figure 17-2. A mechanism whereby a mosaic duplication and/or deletion could be produced. An unequal sister chromatid exchange at a postzygotic cell cycle generates a chromosome with a dup in one chromatid and a del in the other. At the next mitosis, the two chromatids segregate to the daughter cells, giving rise to a dup cell line and a del cell line. The del line may be lost, since partial monosomies are generally less survivable than partial trisomies, in which case only the cells of the dup lineage would exist alongside the normal cells. (After Faivre et al., 2000b.) |
Additional Material from Another Chromosome
In other rearrangements, the duplicated material has come from another chromosome. Illegitimate pairing between nonhomologs, followed by crossing-over (presumably within a region of homology or near-homology), produces reciprocal products that are two derivative chromosomes. In a single-segment exchange, one of these will have a duplication, and the other a deletion. If this occurred during meiosis, and if segregation were then asymmetric, gametes with a duplication or with a deficiency would be produced. Other scenarios with more complex mechanisms may be imagined. Coles et al. (1992), for example, studied a child with Wolf-Hirschhorn syndrome who had two separate de novo rearrangements of the X chromosome with a no. 4 and the Y, respectively, and they propose that simultaneous or sequential crossovers happened in a meiotic “octad” of four synapsing chromosomes.
If a de novo unbalanced rearrangement could be shown to have its component parts originating from a maternal and a paternal chromosome, the fact of its postzygotic origin would be thereby demonstrated. Sarri et al. (1997) offer an example of this scenario in a malformed child with 46,X,der(X),t(X;17)(q27;q22) whose der(X) originated from the paternal X and the maternal 17 chromosomes. Eggermann et al. (1997a) report a similar case, an abnormal child with a de novo der(18)t(13;18)(q14.3;q23). The chromosome 18 component of the translocation came from the paternal no. 18, and the chromosome 13 component from the maternal no. 13. In this type of biparental rearrangement,even the very small risk otherwise associated with parental gonadal mosaicism could confidently be excluded.
Similarly, mosaicism in the presence of a normal cell line would typically allow the presumption of a mitotic origin. Zaslav et al. (1999) report a child with a severe brain malformation who had the karyotype 46,XX,der(4)t(4;15) (q35;q22)/46,XX. They propose that the chromosome constitution at conception was 46,XX. At an early cell division, a reciprocal exchange occurred between chromatids of chromosomes 4 and 15. Then, at anaphase, there was an unfortunate segregation. The newly generated der(4) passed to one daughter cell, along with the normal chromosome 15; and, vice versa, the der(15) and the chromosome 4 passed to the other. The former produced a cell line with a del(4)/dup(15) imbalance, and the presence of this cell line in the developing nervous system presumably caused the brain maldevelopment. The other cell line was not seen (on a peripheral blood karyotype), and it may have been selected against. (If the segregation of the chromosomes at that crucial mitosis had been balanced, then the child would likely have been a phenotypically normal mosaic balanced translocation carrier.) Reddy and Mak (2001) could demonstrate mosaicism in both blood (conventional karyotyping) and on buccal mucosal cells (FISH) in two patients with additional material from another chromosome. For example, one patient had mosaicism for an add(5), the additional material coming from 3p26–pter, in 32% of lymphocytes. Using a 3p-subtelomere probe, a very similar level of mosaicism (40%) was shown in buccal epithelial cells.
Rare Complexities
Triplication. A very few cases are known of a segment of chromosome replicating twice over, being in threefold amount on that homolog. The segment is thus present, in total, in fourfold dose. Triplications are reported for chromosomes 2, 5, 7, 9, and 15 (Wang et al., 1999a; Roberts et al., 2002; Vialard et al., 2003).
Jumping Translocation (“Translocation Sauteuse”). This evocative expression describes a mitotic rearrangement whereby the same piece of one chromosome breaks off, on more than one occasion, and attaches to the tips of other chromosomes. The site of breakage in the donor chromosome is characterized by the presence of an interstitial (internal) telomere, and this region offers the possibility of fusion with the recipient chromosomes. Only 22 constitutional cases are listed in the review of Reddy and Murphy (2000). Levy et al. (2000) identified the phenomenon in two couples, themselves karyotypically normal who presented with recurrent miscarriage and had evolving “jumping” cell lines in the cultured products of conception. Lefort et al. (2001) describe in some detail their own case, an otherwise normal boy with a (possibly coincidental) structural cerebellar defect. He had four separate cell lines on blood and skin biopsy samples, with the segment 2p12–pter attached to 1pter, 5qter, 6qter, and 12qter, respectively. In each, the rearrangement appeared to be balanced. These authors proposed that these translocations were truly one-way, that is to say, they had no reciprocal exchange, with healing of the 2p12 stump occurring by the formation of new telomeric sequences. These cases are typically de novo, and the reason for the chromosome suddenly becoming susceptible in the individual is unknown. The genetic implications for the next generation, if any, remain to be understood.
Extra Structurally Abnormal Chromosome
Many different ESACs exist in the general karyotype 47,+ESAC; they are also known as marker, supernumerary, accessory, and B-chromosomes. The ICSN nomenclature is “mar,” deriving from the word marker. The 47,+ESAC individual has a duplication (partial trisomy) or, in some cases, a triplication (partial tetrasomy) of the material comprising the ESAC. The birth prevalence is in the range 2–7 per 10,000. Blennow et al. (1995) record a large Scandinavian experience: of 50 ESACs, almost half were idic(15), six were small rings deriving from various autosomes, six were isochromosomes of 18p or 12p, and most of the remainder were harmless ESACs derived from acrocentric chromosomes. A particular category is the ESAC that lacks α-satellite DNA (a component of the normal centromere) but possesses a “neocentromere,” a point of considerable theoretical interest and possibly practical significance (see below).
With increasing sophistication of staining techniques, and as these become more widely available, the expression “ESAC,” or “marker” could be seen as a temporary designation, awaiting the full delineation of whatever partial trisomy it may be. Readers who delight in bold colors should refer to Reichenbach et al. (1999). These authors describe a child with 47,+mar who, upon multicolor banding, could be seen as having the karyotype 47,+del(5)(q11). By contrast, an absence of color was, in a sense, even more dramatic in a mildly abnormal child studied by Mackie Ogilvie et al. (2001): he had a C-band–nega-tive ESAC that failed to show hybridization with any whole chromosome paint and could not be identified using a range of other cytogenetic methods (it also lacked a conventional centromere). The origin of this ESAC is thus quite baffling.
Isochromosomes
An isochromosome is a mirror-image chromosome, with two identical arms on either side of the centromere. When present as a supernumerary chromosome, it imposes a tetrasomic state for the chromosomal arm concerned. Recorded isochromosomes include i(5p), i(8p), i(9p), i(10p), i(12p) (Pallister-Killian syndrome2), i(18p), i(18q), i(20p), and i(Xq) (not including the Robertsonian isochromosomes, which are dealt with in Chapter 6, and the i(21q) in Chapter 16). Isochromosomes can arise by a variety of mechanisms and at diverse times and places. The simplest and classical mode of isochromosome formation is a misdivision at the centromere, also known as “centric fission” (Rivera and Cantú, 1986). It can be thought of as a horizontal rather than a vertical division (Fig. 17-3). This gives rise, in a chromosome A, to short arm and long arm isochromosomes, an i(Ap) and an i(Aq). More complex scenarios can be devised, such as the U-type exchange, with the chromatid of one arm of a chromosome “looping around” to join with its fellow. Whatever the mechanism, the process could occur in a premeiotic gametocyte (which would give gonadal mosaicism); during one or another meiotic division; in the zygote; or at an early or a later postzygotic division in an initially normal or an initially trisomic conceptus. Subsequent mitotic loss of an isochromosome in one cell lineage can bring about a mosaic state, potentially having a less severe functional imbalance. This is not necessarily a good thing, since an otherwise early in utero lethal imbalance could convert to a survivable but profoundly abnormal phenotypic state. The patterns of recombination within the isochromosome enable insight into the way the rearrangement has arisen. Of the various possibilities, the most frequent chain of events may be the following: a maternal meiotic nondisjunction, misdivision of the centromere to produce the isochromosome, generation of a 24,i gamete, and postzygotic loss of the isochromosome in one cell which then gives rise to the normal cell line of the embryo (Dutly et al., 1998a; Eggermann et al., 1998).
|
|
|
Figure 17-3. Outline of the classical theoretical mechanism to produce an isochromosome by horizontal misdivision at the centromere. In this example, the site of its generation is at the first division of the zygote, considering the case in point of the i(21q) form of Down syndrome (DS). The normal zygote has two no. 21 homologs (maternal is cross-hatched, paternal is open). At the first mitosis, the maternal chromosome 21 divides appropriately at the centromere to give two normal daughter chromatids, but the paternal chromosome 21 misdivides. One product of the misdivision is an i(21q), and the cell resulting is trisomic for 21q (lower left). The other product is an i(21p), and this cell essentially has a 21 monosomy; its lineage does not survive. Thus the child has a nonmosaic i(21q), and a typical DS phenotype. If the abnormal cell division occurs at a later mitosis (the second or subsequent), a mosaic 46,i(21q)/46,N karyotype could result. |
However, the very rare circumstance of recurrence of an isochromosome in siblings presumably reflects a premeiotic generation of the abnormality in a parental gonad (Krüger et al., 1987). The error in this circumstance could have occurred at mitosis in the parent, some distance along in their embryonic development, with only a small fraction of the body (minimally, a part of one gonad) being involved. Or, the error could have been present at the parent's conception, with a postzygotic loss of the supernumerary isochromosome in one cell line which becomes 46,N, and favorable distribution of this line in the soma, but not gonad, thereafter. Boyle et al. (2001) propose such an evolution in a family in which a mother, with 46,XX on blood karyotype, had two children (half-sisters) with nonmosaic 47,XX,i(18p). She herself, therefore, must be a gonadal mosaic. As these workers show, it is likely that she had been 47,XX,i(18p) at her own conception, due either to a meiosis II error or a premeiotic mitotic error in her mother's oogenesis. In her postzygotic development, the i(18p) line was lost in most tissue, but not in gonad. And surely, one would have thought, gonadal mosaicism was the reason for the father in Williams et al.'s (2001) study having had one child with i(18p) and another with i(18q). A centromere misdivision in an early gametic stem cell mitosis was the obvious explanation. But on sperm analysis, over 1000 cells counted each had just a single 18p and a single 18q signal. Examples like these oblige some caution in counseling, although it is true that such cases are very rare.
Isodicentric 15. An ESAC that warrants special attention is the bisatellited dicentric marker, idic(15), also known as inverted duplication 15, inv dup(15), or pseudodicentric 15. These may arise from a U-loop mechanism, as discussed above, a vulnerability that is set up by the existence of hot spots for rearrangement in this part of chromosome 15 (Huang et al., 1997). Webb (1994) divides the idic(15) into three groups: (1) very small chromosomes with so little chromatin between the centromeres that the appearance is monocentric; (2) medium-sized chromosomes with two distinct centromeres with visible intervening chromatin; and (3) larger chromosomes, greater in size than a G-group chromosome. Probes for D15S10 and SNRPN in 15q12–q13 are very useful in assessing the amount of material (Eggermann et al., 2002). The very small idic(15), in which the additional 15q material is confined to q11 and which is FISH-negative for D15S10/SNRPN, is usually, perhaps always, of no harm per se (other than a possible association with infertility in the male). With the medium-sized and larger chromosomes, the presence in trisomic or tetrasomic dosage3 of the segment recognized by D15S10/SNRPN— particularly when of maternal origin—corre-lates with abnormality, and the phenotype encompasses pervasive developmental disorder with autism, epilepsy, and minor physical defects (p. 290) (Rineer et al., 1998; Wolpert et al., 2000; Torrisi et al., 2001).
Isodicentric 22: Cat-Eye Syndrome. One of the better known ESACs is the inv dup(22)(pter–q11.2) of the cat-eye syndrome, and Mears et al. (1994) provide a review. Two loci, CECR1 and CECR2, may be critical dos-age-sensitive genes (McDermid and Morrow, 2002). The region that is duplicated can vary, and the chromosome is not necessarily symmetrical (not truly “iso”), depending upon the actual sites of homologous recombination within 22q11 that led to the rearrangement. The euchromatic region may thus be present in trisomic or tetrasomic state. Most cases arise de novo, but familial transmission is recorded including, remarkably, familial mosaicism (Urioste et al., 1994b). The phenotype appears not to correlate well with the size of the chromosome and indeed the person may show no signs of the syndrome (Crolla et al., 1997). Bergman and Blennow (2000) describe the unique case of a phenotypically normal man with inv dup(22) mosaicism, who also had a 22q11 deletion and a ring 22; it is plausible that the three abnormal chromosomes arose from related recombinations.
Rare Complexities
Extra Structurally Abnormal Chromsome with Neocentromere. An acentric fragment, which would usually be lost, can rescue itself by generating a neocentromere. These neocentromeres lack α-satellite DNA and its centromere binding protein (CENPB) (hence the alternative name “analphoid centromere”). In other respects, neocentromeres function similarly to normal centromeres, as evidenced by mitotic stability (in many, but not all cases) and by the binding of other known centromere proteins, such as CENPA, CENPC, and CENPE (Saffery et al., 2000; Voullaire et al., 2001). Most chromosomes are now represented in the list of those in which a neocentromere has been identified. Neocentromerization can be considered as a process based upon an epigenetic mechanism: the DNA in this region is the same as that in the normal chromosome, but it has now been influenced to take on a new identity and function (Amor and Choo, 2002).
Neocentromere–ESACs can result from a number of different chromosome rearrangements, the common feature of which is the generation of a chromosome fragment that does not contain a conventional centromere. Most neocentromere–ESACs exist as an inv dup, with the neocentromere forming on one of the two otherwise identical “arms” (rather than at the inv dup breakpoint). If the neocentromere forms some distance down one of the arms, a submetacentric chromosome results. In the more common category (class I), the neocentromeric ESAC exists supernumerary to an otherwise normal karyotype, conferring a tetrasomy for the segment concerned. The class II neocentromere–ESAC results following a simple chromosome break. The proximal fragment, which contains the normal centromere, now exists as a deletion chromosome. The distal acentric fragment is able to survive by generating a neocentromere and duplicating itself, forming a mirror-image iso-fragment. The net result of this rearrangement is trisomy for the affected chromosome segment. If the iso-fragment replaces one normal homolog, to give a 46 chromosome count, it is not, to be precise, an ESAC (rather, a “SAC”), but we may reasonably make use of this term here. Barbi et al. (2000)report a mirror-image chromosome 21 replacing a normal chromosome 21. The isoacentric (or iso-neocentric) 21 in this child carried duplicated copies of 21q21.1–qter, almost all of the long arm, although in fact the phenotype was not typical for Down syndrome. Two examples are known of a phenotypically normal person with an essentially balanced karyotype, in which the ESAC with a neocentromere (class II) was derived from an interstitial deletion, of 1p and 13q, respectively. The first was a man presenting with infertility due to oligospermia, and the second a woman studied because of recurrent miscarriage (see also p. 184) (Slater et al., 1999; Knegt et al., 2003).
Voullaire et al. (2001) analyze possible mechanisms for the generation of neocentric ESACs. The acentric fragment could arise either at meiosis or at a very early mitosis. Plausibly, in some cells of the early dividing zygote, the acentric fragment will be lost, yielding normal cell lines. In others, neocentromerization of the fragment could occur a few mitoses thereafter, at the blastocyst developmental stage, allowing its continuing survival in those lineages in which the process had taken place. The karyotype in one of their cases was written 47,XY,+inv dup(8)(pter → p23.2[neocen]p23.2 → p23.1::p23.1 → pter)/46,XY; the reader will be able to determine the precise structure of the ESAC from this description, and appreciate that this child had a mosaic tetrasomy for 8p23.1–pter.
Chromosomes 13 and 15 seem most prone to neocentromerization, and certain spots in chromosome 13, for example, appear to possess this propensity to neocentromere generation, including 13q21 and 13q32 (Li et al., 2002b). Considering the example of chromosome 13, the proximal fragment becomes a 13q chromosome, while the iso-fragment exists as a distal 13q duplication. The net result is a distal 13q trisomy.
Centromere–Telomere Fusion. Centromere–telomere fusion is a rare mechanism to form an isochromosome. So far, this has been observed with an isochromosome for the short arm, and the long arm of that chromosome being attached to the telomere of another. Rivera et al. (1999b), for example, describe an iso(12p), in which the 12q element was translocated to 8pter. They were able to show FISH signals for both centromere and telomere probes at the 8pter/12cen union point in the translocation chromosome. The probable mechanism is a postzygotic centric fission of the 12, with the 12q element combining through a centromere–telomere fusion, and with the 12p element doubling up to produce the isochromosome. The end result was trisomy for 12p. Besides 12p, this scenario has been observed in isochromosomes for 4p, 5p, 7p, 9p, and 10p. All cases have been de novo.
Isodicentric Chromosome with Deletion. An isodicentric chromosome may have a deletion at the center of symmetry. Piantanida et al. (1997) describe a unique idic(8)(p23.3) in which there was a nullisomy for a very small subtelomeric segment. Both elements of the idic(8) were of maternal origin, but it seemed more probable that the abnormal phenotype in the child was due to the deletion rather than a uniparental disomy.
Complementary Isochromosomes. This very rare circumstance is dealt with in Chapter 7.
APPARENTLY BALANCED BUT ACTUALLY UNBALANCED REARRANGEMENT
De novo translocations, inversions, and insertions can appear to be balanced at the cytogenetic level, but in fact a locus or loci at the breakpoint site(s) is disrupted or deleted, and is the cause of an abnormal phenotype. An imbalance may be suspected at the level of routine banding and then be confirmed with fluorescence studies. For example, Mohrschladt et al. (2000) describe a child with mild retardation and “soft” dysmorphic signs, with a de novo t(6;9), the exact breakpoints being uncertain. Analysis with 500-band karyotyping suggested that the translocation might be unbalanced, and this was confirmed on FISH, showing that, in addition to a translocation involving breakpoints at 6q27 and 9q22.1, a small segment of chromosome 9 (9q21.2q22.1) was duplicated. The clinical picture reflected, presumably, a 9q21.2–q22.1 trisomy. The duplication was large enough to be appreciated (if not precisely delineated) on standard karyotyping. We may suppose that similar processes, involving duplicated or deleted segments of lesser size, not visible on standard karyotyping and perhaps not even on FISH, may underlie some other cases of apparently balanced de novo translocations. In one marvellously complex illustration, what was initially thought to be a simple de novo balanced insertion in an abnormal child, 46,XX,inv ins(5;7) (p13;qterq31), turned out to be a 6-breakpoint rearrangement which included a 6 Mb deletion: 46,XX,der(5)(5pter→5p13::7q31.32→7q34::7q35→7q36::7q31.32→7q31.32::5p13→5qter), der(7)(7pter→7q31.32::7q36→7qter) (Kraus et al., 2003a).
The coincidence of abnormal karyotype and abnormal phenotype may be fortuitous. Hersh et al. (1995) very reasonably suggested a de novo t(1;10) in an infant with thanatophoric dysplasia (TD) might indicate the location of the TD gene, but in fact it proved to be the fibroblast growth factor receptor-3 locus, which is on chromosome 4p16.3 (Tavormina et al., 1995). Where there is supporting evidence from other examples, the case is stronger. Given that there are reports in the literature associating a distal 3q–deletion with microophthalmia or anophthalmia, Driggers et al. (1999) were well placed to propose that their patient with isolated anophthalmia and a de novo t(3;11)(q27;p11.2) may point to a locus for eye formation in the region of 3q27. Two translocations with a common breakpoint actually led to the discovery of the gene for Sotos syndrome (p. 280).
|
|
|
Figure 17-4. An apparently balanced translocation causing the syndrome of campomelic dysplasia (which includes skeletal, genital, and brain defects). One breakpoint is at 17q25.1, on or close to the SOX9 locus (shown as a dot on the cartoon karyotype), where the basis of the syndrome lies. One possibility is that the gene is disrupted. Or, an influence of adjacent no. 5 chromatin leads to inactivation of the SOX9 gene on the der(17), the functional SOX9 haploinsufficiency then being responsible for the phenotype. (Case of R. Savarirayan; Savarirayan and Bankier, 1998.) |
Position effect is another mechanism whereby a “balanced” rearrangement can lead to phenotypic abnormality. The SOX9 gene on chromosome 17 at band q25.1, the basis of campomelic syndrome, provides an example with respect to both a translocation and an inversion. The de novo translocation t(5;17) (q15;q25.1) in Figure 17-4 was seen in a child with this syndrome, as was the de novo para-centric inversion inv(17)(q24.3q25.1) reported in Maraia et al. (1991). Here, if there is an actual structural or functional imbalance, it is at the finest molecular level, and may involve only the SOX9 locus.
GENETIC COUNSELING
DELETION
In most children with deletions, the parents type as 46,XX and 46,XY, and the defect is de novo. The risk for recurrence is very small, but it is not nonexistent. We assume that these rare recurrences are due to an occult parental mosaicism, which the routine blood chromosome study could not detect. The abnormal line may be gonadal (confined to gametic tissue) or so-matic–gonadal (some somatic tissues are involved as well, but not, apparently, blood). The observation of rarity of recurrence allows us to propose the empiric advice that, in the individual case, recurrence is most unlikely. A figure that is appropriate in this setting is less than 0.5%; the counselor should note the converse, greater than 99.5%, for a child without the chromosome defect. Nevertheless, in practice, many couples advised of a very low risk still request prenatal diagnosis for reassurance, and one can sympathize with this request. If testing is to be based on classical cytogenetics alone, care must be exercised in offering prenatal diagnosis of small deletions seen only on high-resolution lymphocyte chromosomes. The technical ability to demonstrate such small changes in amniotic fluid or chorionic villus cells may be limited.
|
|
|
Figure 17-5. Composite karyotype showing the site of the cytogenetic defect in some of the deletion syndromes. AS, Angelman syndrome; CMT, Charcot-Marie-Tooth neuropathy (a duplication); HPSN, hereditary pressure-sensitive neuropathy; PWS, Prader-Willi syndrome; Rb, retinoblastoma plus other features; WAGR, Wilms tumor, aniridia, genital defects, retardation syndrome. |
In those deletions where a parent is shown to carry a balanced rearrangement, a substantial recurrence risk is probable, and the appropriate chapter should be consulted.
Brief sketches of the major deletion syndromes follow, as well as some less well-known ones, in numerical order of chromosomes. Some gain inclusion because of one specific and striking feature, such as the del(5) syndrome with polyposis. Some that are speculative (Cornelia de Lange syndrome, for example) warrant listing because of their particular interest to the dysmorphologist. We comment in greater or lesser length on the genetics of each. In some, we make mention of familial transmission; but primarily we are dealing with de novo defects. The better known ones are depicted in the composite karyotype in Figure 17-5. (In the limit, every different deletion, even if only one case is known, could be regarded as a new syndrome.) Chromosomal atlases and catalogs provide further clinical information, and the essays in Cassidy and Allanson's Management of Genetic Syndromes (2001) offer detailed commentaries for some of the more common of these syndromes (Angelman, Cornelia de Lange, Prader-Willi, Russell-Silver, Smith-Magenis, velocardiofacial, and Williams syndromes).
Chromosome 1
del 1p36.3
The del(1)(p36.3) syndrome may be of similar frequency to that of the del(22)(q11) syndrome, with these two conditions being the most common syndromes of terminal and interstitial deletion, respectively. The facies is variably dysmorphic, and several minor physical anomalies may be observed (Zenker et al., 2002). The mental defect is usually severe, although less marked when certain 1p36.3 sequences at the molecular level are retained (Wu et al., 1999a); perisylvian polymicrogyria is a characteristic neuroradiological correlate. The deletion is not always discernible even on the best-quality preparations. This reflects the facts that 1p36.3 is a light-staining region and the deletions are of variable size, some being simply too small to be seen. In three subtelomeric FISH studies of “karyotypically normal” retarded and dysmorphic populations, two del(1)(p36.3) cases were identified among a total of 411 individuals tested (Anderlid et al., 2002a; Baker et al., 2002a; Clarkson et al., 2002). Using an initial molecular approach (a panel of microsatellite markers to pick up 1q36.3 hemizygosity), Giraudeau et al. (2001) screened 567 patients with mental retardation and found three with del(1p), in whom FISH confirmed the deletion. These combined data suggest that about 0.5% of this category of population, in whom a standard karyotype has been interpreted as normal, may have this subtelomeric deletion. Almost all cases of del(1)(p36.3) syndrome arise de novo, but transmission from a parental translocation is recorded (Shapira et al., 1997b).
del 1q42, q43, q44. Terminal deletions of 1q may come to be recognized more often since a characteristic facies has been proposed and now that subtelomeric FISH analysis is beginning to enter the cytogenetic armamentarium (de Vries et al., 2001a,b). Breakpoints have been listed in 1q42, q43, and q44, the last of these needing FISH for recognition. Familial translocations have been seen, and this possibility must be pursued. The mental phenotype is severe. An interstitial deletion has less phenotypic effect, attested to by the fact of mother-to-son transmission, as documented in Sanford Hanna et al. (2001).
Chromosome 2
del 2q37: Albright-like Syndrome
This cytogenetic defect should specifically be sought in patients with a morphological phenotype reminiscent of Albright hereditary osteodystrophy (short stature, round facies, short metacarpals) and intellectual deficit. It may be among the more frequent of the deletion syndromes, the subtlety of the cytogenetic defect having previously obscured its role (Phelan et al., 1995; Wilson et al., 1995). Familial translocations have been reported (Bijlsma et al., 1999; Batstone et al., 2003). The 2q37 duplicated state has been reported associated with intellect within the normal range, and little or no dysmorphism (Batstone et al., 2003).
Chromosome 3
rea 3q26.3: Cornelia de Lange syndrome
The karyotype in Cornelia de Lange syndrome (CDLS) is almost always normal. The syndrome bears a phenotypic resemblance to the distal 3q duplication syndrome (Rizzu et al., 1995), and this had raised a question of UPD 3; in fact, there is biparental inheritance of chromosome 3 (Shaffer et al., 1993). Ireland et al. (1991, 1995) report a very typical CDLS patient having a de novo translocation with one breakpoint at 3q26.3, and restudied two “mild” CDLS sibs with a 3q26.3 duplication from a paternal intrachromosomal insertion. Rizzu et al. (1995) have studied a patient with a maternally inherited paracentric inversion with a 3q27 breakpoint. These observations were suggestive of a “CDLS locus” in this region, but formal studies of familial cases have failed to adduce further supportive evidence (Smith et al., 1999c; Krantz et al., 2001).
Chromosome 4
del 4p: Wolf-Hirschhorn Syndrome
This well-known deletion syndrome identified in the prebanding era is one of the few that can, in its classic form, be confidently recognized clinically. The natural history is discussed in Battaglia and Carey (1999). Classical deletions are detected on routine cytogenetics, whereas molecular approaches are needed to identify subtler deletions. Zollino et al. (2000, 2003b) propose that the condition be considered in two forms, the severe classical form and a relatively mild form, which correlate with the extent of the deletion. The core phenotype, which includes the distinctive facies, resides in haplo-insufficiency for an ~500 kb segment within 4p16.3. (A more proximal deletion, in 4p15, produces a separate phenotype [Fryns, 1995]). The majority of Wolf-Hirschhorn syndrome occurs de novo, although a few cases are the consequence of parental balanced rearrangement.
The improvement in precision of cytogenetic methodology has enabled the elucidation of some malformation syndromes previously supposed to have been due to autosomal recessive inheritance. The Pitt-Rogers-Danks syndrome turned out to be a less severe phenotypic version of Wolf-Hirschhorn syndrome, with the same deletion capable of causing either syndrome (Kant et al., 1997; Battaglia and Carey, 1998). A fine argument turns on whether Wolf-Hirschhorn syndrome should be considered as having a continuum of severity and subsuming Pitt-Rogers-Danks syndrome, or whether a true distinction exists (Partington and Turner, 1999; Wright et al., 1999). A family with the supposed recessive Lambotte syndrome of multiple malformation, having surprisingly produced a further affected child in a succeeding generation, was subjected to detailed FISH analysis, and a familial t(2;4)(q37.1;p16.2) was identified (Herens et al., 1997). Here, the phenotype in fact reflected the combined effects of partial 4p monosomy and 2q trisomy. A syndrome of mental retardation with polymicrogyria originally reported as being X-linked turned out to be due to an autosomal translocation, t(1;12)(q44;p13.3), with both adjacent-1 segregations being represented (Zollino et al., 2003a). Other examples exist, and Verloes et al. (2000) speak of these as “pseudo-recessive disorders” being revealed in their true colors as due to cryptic translocations.
del 4q34
It may be that among all the chromosomal disorders, no single discrete physical sign is pathognomonic for a particular aneuploidy (that is, seen in this specific condition, and in no other). But there is one sign that might be: a duplicated nail of the fifth finger, with one nail in the normal position, and the other where the fingertip pad should be (thus, dorsal and volar surfaces). This curious observation, it is proposed, can enable the clinician to predict a distal 4q deletion, to be precise, of 4q34.2 (Zackai, 1999). A deletion just a little above, into 4q33, may be associated notably (although not pathognomonically) with abnormality of the ulnar ray of the upper limb (Keeling et al., 2001).
Chromosome 5
del 5p: Cri du Chat Syndrome
The breakpoints in this famous syndrome are very variable: in a molecular study of 62 Italian cases, for example, at least 28 different sites were identified, from p13 through p15.2. The “cri” region is pinpointed to proximal p15.3, and certain other components of the phenotype can be attributed to certain segments within p14, p15.1, p15.2, and p15.3 (Fig. 17-6) (Kjær and Niebuhr, 1999; Cerruti Mainardi et al., 2001). Kjær and Niebuhr suggest that anomalous formation of the notochord in the early embryo may then compromise the development of certain cranial nerve nuclei in the subjacent brain stem, affecting the innervation of the larynx, which is the actual anatomic structure that produces the cat-like cry. Deletion of the segment 5p15.3 alone can produce the typical cry in an otherwise normal child. Van Buggenhout et al. (2000) document in quite some detail, with several photographs, the phenotypes in seven older individuals, teenagers, and adults. All but one had severe or profound mental defect. Perhaps surprisingly, the neurodevelopmental compromise does not correlate with the size of the deletion (Marinescu et al., 1999a). Growth charts have been compiled from international data from 374 cases, from birth to age 18 years, and these document a substantial downward shift in the graphs for the three major indices (weight, height, and head circumference) (Marinescu et al., 2000). While most cases are sporadic, familial transmission from a parental translocation is recorded (Cotter and Musci, 2001), and this possibility should be checked for in each case. One case of recurrence of del(5)(p15.2), identified at prenatal diagnosis, attests to the reality of gonadal mosaicism (Hajianpour et al., 1991).
|
|
|
Figure 17-6. An example of the archetypal del(5p) chromosome (above), deletion of which is the basis of cri du chat syndrome, first described in 1963. A display of a detailed karyotype–phenotype correlation is shown below. Regions which may be implicated in speech, the characteristic cry, and facial anomalies are indicated at left. The arrows indicate five different breakpoints defining five different extents of terminal deletion, the lengths of which are reflected in the vertical bars to the right (numbered 1–5), and the associated phenotypes are noted alongside. The vertical bar to the left (numbered 6) identifies an interstitial segment deletion of which does not cause phenotypic abnormality. (From Kjær, I. and Niebuhr, J. Studies of the cranial base in 23 patients with cri-du-chat syndrome suggest a cranial developmental field involved in the condition. Am. J. Med. Genet. 82, 6–14. © 1999 Am. J. Med. Genet., courtesy I. Kjær and J. Niebuhr. Reprinted with the permission of Wiley-Liss, Inc., a subsidiary of John Wiley & Sons, Inc.) |
del 5q22–q23: Polyposis Plus Syndrome
A minor degree of facial dysmorphism and mild to moderate mental retardation are nonspecific features seen in deletions in the region of 5q22–q23; the unique feature is adenomatous polyposis of the bowel, and indeed it was such a deletion that led to discovery of the APC (adenomatous polyposis coli) gene (Hockey et al., 1989; Kobayashi et al., 1991). Absence of one APC allele of itself allows polyps to develop (Hodgson et al., 1994), and any subsequent mutation/loss of the allele on the intact chromosome 5 causes loss, in turn, of the tumor suppressor function of this gene.
Other similar examples exist of constitutional deletions that convey, in addition to congenital abnormality, a cancer predisposition. The loss of one allele of a tumor suppressor gene on the deleted chromosome comprises the “first hit” in the process of tumorigenesis. Jacoby et al. (1997) describe a deletion of 10q22.3–q24.1 in a patient with multiple congenital malformations and juvenile polyposis, the latter presumably reflecting the loss of one copy of the PTEN gene. A 9q22 deletion has been associated with a “Gorlin syndrome plus” phenotype, and an increased risk for cancer is to be expected (Shimkets et al., 1996). Carcinogenesis may also be due to gene amplification, and constitutional gain of a particular gene may thus contribute a risk. Willatt et al. (2001b) and Seven et al. (2002) draw attention to the link between constitutional 2p23 duplication and consequent duplication of the oncogene N-myc and neuroblastoma.
del 5q35: Sotos Syndrome
Most Sotos syndrome (cerebral gigantism) is seen with a normal karyotype. A few patients have had a chromosomal abnormality, and since two of these involved translocations with one of the breakpoints at 5q35, and since also some del(5)(q35) cases have had a Sotos-like phenotype, this suggested there might be a Sotos locus in the region. This hypothesis was subsequently vindicated, the gene in fact being NSD1 (Kurotaki et al., 2002). Most patients have a 2.2 Mb deletion, which encompasses the NSD1 locus.4
Chromosome 6
del 6q23.3–q24.2
This deletion is recorded in only a single case, but is worth noting for the observation of intellectual normality. Kumar et al. (1999) describe a 3-year-old girl whose development was “completely normal to advanced,” who had been karyotyped as a newborn because of low birth weight. While the facies was distinctive, she was said to resemble her family. One might imagine that this particular segment contains no brain loci; or, at least, none that might cause abnormality if haploinsufficient. Arguably, this example could have been listed in the chapter on variant chromosomes, but the fact of the deletion being de novo weighed in favor of its placement here.
Chromosome 7
del 7p13: Greig Cephalosyndactyly Syndrome
This acrocephalopolysyndactyly syndrome, classically inherited as an autosomal dominant, is due to mutation at the GLI3 locus. Rare microdeletion cases have haploinsufficiency of GLI3, as well as loss of some adjacent loci, and the phenotype is combined Greig syndrome with neurodevelopmental defect, seizures, and other abnormalities (Kroisel et al., 2001).
del 7p21.1: Saethre-Chotzen Syndrome
Most cases of Saethre-Chotzen syndrome (a type of acrocephalosyndactyly) are due to point mutation in the TWIST gene at 7p21.1. Cytogenetic forms include microdeletion within this region, the larger of which add in learning disability as a clinical trait (Johnson et al., 1998). An apparently balanced translocation, which might disrupt the TWIST gene or perturb its function, is another mechanism. It appears that either mechanism, haploinsufficiency or point mutation, can lead to the similar phenotype. The skull defect in Saethre-Chotzen syndrome is premature fusion of cranial bones; it is interesting that a duplication (“triplo-excess”) at this locus can produce the opposite effect, an underdevelopment of the cranial bones.
del 7q11.2: Williams Syndrome
Williams syndrome (WS) is due to the deletion of a consistent segment of about 1.5 Mb of DNA within chromosome 7q, usually the consequence of an unequal crossing-over in meiosis in one parent (Meng et al., 1998; Wang et al., 1999b; Peoples et al., 2000); no instance is known of the concomitant duplication. The deleted loci include the elastin gene, which is responsible for the cardiovascular component of the phenotype, and a number of contiguous brain and morphogenesis loci. Meng et al. (1998) and Peoples et al. (2000) have established physical maps of the region, which will provide the basis for the delineation of these genes. Some have been identified, such as the LIMK1 gene, whose haploinsufficient state may underlie the defect of visuospatial constructive cognition of WS. Children with WS find it confusing, for example, to step from carpet to wooden planking even though the floor is flat (Frangiskakis et al., 1996). The characteristic psychological phenotype is that of a mild intellectual disability, with overfriendliness to strangers, and a lacking in social judgment. Earlier impressions that aspects of language might be intact are now refuted (Donnai and Karmiloff-Smith, 2000). Affected monozygous twins generally have a rather similar phenotype (Castorina et al., 1997). Deletion of the elastin locus alone produces the characteristic heart defect in isolation. The WS deletion has similar frequencies of paternal and maternal origin, and the phenotypes do not differ accordingly (Wang et al., 1999b). We know of no record of recurrence in siblings of undoubted WS to normal parents. Rare instances of parent-to-child transmission are recorded (Pankau et al., 2001).
One case exists of a child with a large deletion at 7q11.2, whose clinical picture included components of WS along with other features, including a severe mental defect (Wu et al., 1999b). Presumably, other loci, as well as the “WS loci,” were removed in this deletion.
del 7q21.3: Ectrodactyly Plus Syndrome
One type of split-hand and split-foot malformation is associated with deletions in 7q21, and loss of one allele at a “digit-formation locus” in this region may be the basis. Loss of contiguous genes encompassed by the deletion may contribute to other less specific dysmorphology and to a diminution of intellectual function (Roberts et al., 1991).
del 7q32–qter: Holoprosencephaly Plus Syndrome
Holoprosencephaly is a developmental brain defect that can vary from devastatingly severe to rather mild, and there are several different genetic causes. Chromosomes 7 and 13 are important contributors. Distal 7q deletions were instrumental in the mapping of one locus, HPE3 (Frints et al., 1998). De novo deletion is the rule, but familial holoprosencephaly has been recorded in the setting of a familial 7q36 translocation (Hatziioannou et al., 1991) Deletion 7q holoprosencephaly is due to HPE3 haploinsufficiency. Another deletional cause of holoprosencephaly resides in chromosome 13, del(13)(q32), and reflects haploinsufficiency for the ZIC2 brain morphogenesis gene (Brown et al., 1998b), in contrast to the duplication mechanism that applies in typical trisomy 13.
Chromosome 8
del 8p23.1–pter, or del 8p23.1
Small terminal deletions of 8p are coming to be more frequently recognized. A notable aspect of the phenotype is a severe behavioral disturbance in childhood, on the background of a mild mental defect. Sudden and extreme changes in behavior are observed, with outbursts of aggressiveness and destructiveness. Frustration tolerance is very low. Behavior seems to improve in later adolescence. Possibly, this picture may reflect the location of “neurobehavioral genes” on distal 8p (Claeys et al., 1997). A deletion of 8p23.1 (terminal or interstitial) may remove the gene for a cardiac transcription factor, GATA4, and this may be the basis of the observation that heart defects are frequent (Pehlivan et al., 1999).
del 8q24.11–q24.13: Langer-Giedion Syndrome (Tricho-Rhino-Phalangeal Syndrome Type II)
The facies is distinctive, and diagnosis can be made with some confidence on clinical grounds. The condition is due to a deletion that removes the gene for tricho-rhino-phalangeal (TRP) syndrome type I, a bone growth control gene (EXT1, which causes exostoses5), and several other genes, to give the broader picture of Langer-Giedion syndrome (Lüdecke et al., 1999; Momeni et al., 2000). The deletion may arise on the chromosome 8 of either parent (Nardmann et al., 1997).
Chromosome 9
del 9p22–p23
Quite a number of 9p–cases are recorded, over 100, and they present a characteristic phenotype. Many are due to deletions occurring in a region of about 5 Mb in 9p23, which is not so much a hot spot, but a series of hot spots. Some cases which may at first glance seem to be simple deletions will turn out on FISH studies to be due to other more complicated rearrangements (Christ et al., 1999).
del 9p24.3: Sex Reversal Plus
This deletion syndrome is notable in having pointed the way to discovery of the DMRT1 gene (Drosophila Doublesex C. elegans Mab3 related transcription factor 1) (see also p. 294). This is the most conserved of anyknown sex-determining gene, and is actually on the Z chromosome (the homogametic chromosome) of birds. Its expression is normally greater in the male than in the female embryo, and this dosage may be the basis of its testis-inducing action. It is proposed that the loss (or perturbation) of one DMRT1 allele in (or adjacent to) the deleted segment brings the amount of product down below this threshold, and thus the 46,XY,del (9)(p24.3) person develops as a female (Calvari et al., 2000). An incomplete loss of function may lead to genital ambiguity. Loss of adjacent genes presumably contributes to the wider phenotype.
Chromosome 10
del 10p13: DiGeorge Syndrome II
This deletion is the basis of a phenocopy of DiGeorge syndrome (DGS; see Deletion, Chromosome 22, below), and indeed the condition has been labeled DGS II (Dasouki et al., 1997; Gottlieb et al., 1998; Schuffenhauer et al., 1998). Some features of DGS II, such as ptosis and hearing loss, are not present in the 22q deletion form. It is clearly a rare cause of “non-22q DGS,” as no cases were found in a European study specifically addressing this question, and indeed the authors propose that searching for 10p microdeletions is not warranted in the service laboratory (Bartsch et al., 1999, 2003). In an American study, in which a dual-probe FISH for DGS I and II was used, 1 case was found in 412 patients presenting with a possible diagnosis of DGS (54 had DGS I) (Berend et al., 2000b). Hypoparathyroidism is a frequent observation in del(10)(p13), but this may in fact be due to the HDR (hypoparathyroidism, sensorineural deafness, renal dysplasia) syndrome, whose locus may lie in this region (Fujimoto et al., 1999).
del 10q11.2: Hirschsprung Disease Plus Syndrome
The locus for the receptor kinase gene RET is at 10q11.2 (the Hirschsprung chromosome region 1, HSCR1). In the haploinsufficient state, certain neurons/neural crest cells may fail to migrate to their proper place and/or fail to undergo proper neuronal maturation in the intestinal wall. Without this nervous control, the segment of bowel is chronically contracted, and this causes a partial or complete obstruction (Hirschsprung disease). The loss of adjacent loci contributes to a wider phenotype (Fewtrell et al., 1994). The similar circumstance applies with respect to HSCR2, at 13q22. Loss of the EDNRB locus causes Hirschsprung disease, and contiguous loci are responsible for other features that may add up to a specific syndrome (Shanske et al., 2001).
Chromosome 11
del 11p13: WAGR Syndrome (Wilms Tumor, Aniridia, Genital Defects, Mental Retardation)
Haploinsufficiency of the PAX6 morphogenesis gene causes aniridia (absence of the iris). Loss of one WT1 allele can comprise the first hit in the sequence of events to cause Wilms tumor, and may also be responsible for the impairment of genital development. These two genes, and presumably some brain genes, are removed in the 11p13 deletion, and the tout ensemble adds up to the WAGR syndrome. Severe obesity may also be an occasional part of the picture (Gül et al., 2002). 11p13 deletions and translocations with a presumed position effect are the cause of a substantial fraction, about 40%, of all cases of aniridia (Crolla and van Heyningen, 2002). Interestingly, a duplication for the segment 11p12–p13 also produces an eye defect, suggesting that a PAX6 dosage effect, whether an insufficiency or an excess, influences the morphogenesis of the eye (Aalfs et al., 1997).
del 11q23: Jacobsen Syndrome
This syndrome is of some interest because very rarely the fragile site (FRA11B) at 11q23.3 may predispose to the generation of the deletion (Jones et al., 1994). The great majority of patients do not have their deletion breakpoint at or immediately adjacent to the fragile site.
Chromosome 13
del 13q14: Retinoblastoma Plus
The association of retinoblastoma with constitutional 13q–was recognized in the early days of cytogenetics, and this observation gave the clue to the position of the retinoblastoma gene (RB) on this chromosome, which was since localized more precisely to 13q14. A wider syndrome can accompany the deletion, including mental retardation and facial dysmorphism (Fig. 17-7). The severity is according to the extent of the deletion, those extending distally to q14.1 being of more severe effect than those extending proximally (Baud et al., 1999). One 13q–child is recorded with retinoblastoma detected at age 4, and at age 11 Wilson disease, an autosomal recessive disorder of copper transport, was diagnosed. Since the Wilson locus is also in this region (13q14.3–q21.1), the assumption is that the child's other chromosome carried a Wilson mutation, and the deletion “exposed” this mutation in the hemizygous state (Riley et al., 2001).
|
|
|
Figure 17-7. A deletion of chromosome 13, from a woman who had bilateral retinoblastomas as a child, and who has a mild intellectual deficiency and minor facial dysmorphism. The karyotype is 46,XX,del(13)(q12q14.2). The deleted chromosome is on the right; the segment q12–q14.2 is indicated on the intact chromosome, left. The RB (retinoblastoma) locus is in band p14.2. Her parents and brother karyotyped normal. (Case of L. V. Hills.) |
Chromosome 15
del 15q11–q13: Prader-Willi Syndrome, Angelman Syndrome
See Chapter 20.
Chromosome 16
del 16p13.3: α-Thalassemia and Mental Retardation
This is one of two α-thalassemia and mental retardation (ATR) syndromes (the other being an X-linked Mendelian condition) (Lindor et al., 1997). In the del(16p) ATR syndrome there is monosomy for a 1 Mb segment including the α-chain globin loci and some brain loci. This can be a true terminal deletion, with “healing” of the single breakpoint by the addition of telomeric sequences (Lamb et al., 1993). A larger deletion determines a broader phenotype, with tuberous sclerosis and polycystic kidney disease as well as the ATR (Eussen et al., 2000). Smaller deletions, in the range 2.7 to 268 kb, rather surprisingly produce no phenotype other than thalassemia, in spite of the deletion of from 1 to 15 other genes (Horsley et al., 2001). Holinski-Feder et al. (2000) report a notable example of “ATR-16” due to a subtelomeric translocation, which had escaped detection on multiplex-FISH, and which only came to light after a pedigree analysis showed linkage to 16p.
|
|
|
Figure 17-8. Two chromosome 16p deletion syndromes, α-thalassemia/retardation syndrome (left), and Rubinstein-Taybi syndrome (right). Both deletions are in the distal-most band, 16p13.3, the ATR-16 region being distal to R-T. These deletions are difficult or impossible to see on routine banding, but are clearly apparent on FISH with the appropriate probe, as shown here, with one of the no. 16 chromosomes in each case showing nonhybridization (indicated by the shorter arrows; longer arrow is normal homolog). |
del 16p13.3: Rubinstein-Taybi Syndrome
The Rubinstein-Taybi syndrome (RTS) has a distinctive phenotype, and the facies and the broad thumbs are very characteristic. In about one-tenth of cases, the basic defect is a deletion of the gene CBP (cyclic AMP–regulated enhancer binding protein) (Wallerstein et al., 1997; Petrij et al., 2000a). It is thus a single-locus disorder rather than a contiguous gene syndrome, and point mutation within this gene can also cause the syndrome. The basic defect in CBP leads to a generalized dysregulation of expression in a number of target genes. The deletion can be seen on FISH using cosmid probes (Blough et al., 2000) (Fig. 17-8). There is no obvious clinical distinction between those RTS patients with or without the microdeletion (Taine et al., 1998). The range of observed severity presumably reflects a variable expressivity of the abnormal genotype, and the case of identical twins with RTS having rather different neurobehavioral phenotypes supports this suggestion (Preis and Majewski, 1995). The oldest putative case, from about A.D. 500–900, is that of a skeleton excavated at the Yokem site in Illinois (Wilbur, 2000); some kind of record would be set were this case ever to yield to a paleocytomolecular genetic analysis!
The sites of recombination in the majority of translocation and inversion forms of RTS lie within a breakpoint-cluster region in the 5′ part of the CBP gene. This region is also involved in somatic rearrangement, and, for example, the translocation t(8;16)(p11;p13.3) can be a contributory event in the genesis of acute myeloid leukemia (Petrij et al., 2000b). The CBP gene thus joins the ranks of the small number of genes known to cause congenital malformation if abnormal during embryogenesis, and cancer if the abnormality is acquired in postnatal life (NSD1 being another example; see del 5q35 above).
Chromosome 17
del 17p13.3: Isolated Lissencephaly Sequence, and Miller-Dieker Syndrome
A deletion of the brain morphogenesis gene LIS1 produces lissencephaly (“smooth brain”), a severe neuronal migration defect, although agyria/pachygyria (absence/thickness of gyri) may be a more accurate description. There are neuroradiological features of this 17p lissencephaly that enable distinction from the other major genetic type, the X-linked syndrome. Deletions are of variable extent, and may be intragenic or remove the entire gene (Cardoso et al., 2002). All FISH-detectable deletions confined to LIS1 found thus far have been de novo, implying a very low recurrence risk. Prenatal diagnosis with ultrasonography is unreliable, and parents wishing complete reassurance could choose to have CVS or amniocentesis with FISH (Pilz et al., 1998a).
|
|
|
Figure 17-9. Miller-Dieker syndrome. Chromosomes show deletion in distal 17p, at band p13.3. The FISH probe does not hybridize to the deleted chromosome (shorter arrow; longer arrow is normal homolog). |
Loss of an adjacent locus or loci adds in defects of other systems and characteristic dysmorphogenesis, and this constitutes the Miller-Dieker syndrome (MDS) karyotype and phenotype (Chong et al., 1997; Grimm et al., 1999; Gambello et al., 1999) (Fig. 17-9). The brain malformation is more severe than in the isolated LIS1 deletion, which may reflect the contribution of another brain morphogenesis locus in distal 17p. Some MDS cases have been due to a parental rearrangement (inversion or translocation), not necessarily recognizable on routine cytogenetic testing, and this possibility should be very carefully assessed (Yokoyama et al., 1997; Joyce et al., 2002). Mutchinick et al. (1999) describe a child with deletion involving the MDS region within 17p13.3, but with the LIS1locus remaining intact. The child also had a 5q trisomy, but they considered that they could dissect out the phenotypic components of the two aneusomies, and sheet home the specific non-LIS1 aspects of the MDS phenotype.
del 17p11.2: Smith-Magenis Syndrome
The Smith-Magenis syndrome (SMS) encompasses a picture of dysmorphology, mental defect, and fractious behavior (Fig. 17-10). This latter can have the particular characteristics of sleep disturbance (associated with a reversal of the normal circadian pattern of melatonin secretion), diminished pain sensitivity, and self-mutilation, the latter manifesting, for example, as onychotillomania (pulling out nails) (Smith et al., 1998a; Potocki et al., 2000b). To the practiced eye, the facies is distinctive (Allanson et al., 1999). Most (>90%) patients have a common ~3.7 Mb deletion, resulting from an unequal meiotic cross-over (Shaw et al., 2002); a larger deletion may be associated with a more complicated phenotype (Natacci et al., 2000). No instance is known of recurrence in a family of SMS due to the common deletion. Prenatal diagnosis is achieved primarily by molecular methodology, using a marker such as D17S258, which is deleted in all SMS cases (Juyal et al., 1996).
|
|
|
Figure 17-10. Smith-Magenis syndrome. Chromosomes showing deletion in proximal 17p, at band p11.2. The FISH probe does not hybridize to the deleted chromosome (shorter arrow; longer arrow is normal homolog). This region is proximal to the segment deleted in hereditary pressure-sensitive palsy (Fig. 17-11). |
del 17p11.2: Hereditary Pressure-Sensitive Neuropathy
Hereditary pressure-sensitive neuropathy (HPSN) is the genetic countertype of Charcot-Marie-Tooth neuropathy (see below, Duplication). The deletion of a particular “nerve gene”—the PMP22 or peripheral myelin protein 22 gene—leads to abnormal myelination of the peripheral nerves, and this compromises their function. A typical presentation is the backpacker who complains of numbness (sensory nerves) and weakness (motor nerves) in the arms after a day's hiking; these symptoms are due to the pressure of the shoulder straps on the nerves leading to the arms. Alternative names are hereditary neuropathy with liability to pressure palsies (HNPP) and tomaculous neuropathy. Almost all HPSN is due to this type of deletion, and thus detectable using FISH (Fig. 17-11).6 The deletion can arise de novo, or, as is more usual, be transmitted from an affected parent. The risk to transmit the defect is 50%. A single family is on record in which the defect was due to a reciprocal translocation, t(16;17)(q12;p11.2), which disrupted the PMP22 gene, and a heterozygous mother and son had HPSN (Nadal et al., 2000).
Two distinct sex-dependent mechanisms are involved that produce the deletion or duplication of the chromosome 17 region associated with HPSN and Charcot-Marie-Tooth type IA, respectively (Lopes et al., 1998; Inoue et al., 2001a). Rearrangements arising from maternal gametogenesis, which can be either deletions or duplications, are due to an intrachromosomal mechanism, either an unequal sister chromatid exchange or, in the case of deletion, excision of an intrachromatid loop. If the rearrangement occurs in paternal gametogenesis (the more common scenario), it comprises a duplication and arises by unequal meiotic crossing-over between the two no. 17 chromosomes (cf. Fig. 17-1), an interchromosomal mechanism. Hereditary pressure-sensitive neuropathy is an example of the unusual circumstance of the monosomic state having a less severe phenotype than the trisomic: the neuropathy of Charcot-Marie-Tooth disease is more disabling than in HPSN. This might reflect an accumulation within the cell of degraded excess protein in the PMP22 duplication, compromising the cell's function, instead of a mere reduction in the amount of protein with the deletion (Ryan et al., 2002).
|
|
|
Figure 17-11. Countertype deletion and duplication in 17p11.2 leading to hereditary pressure-sensitive neuropathy (deletion, left) and Charcot-Marie-Tooth neuropathy (duplication, right). The deletion is evident through non-hybridization of the appropriate probe on FISH of metaphase chromosomes (shorter arrow, left; longer arrow is normal homolog). The duplication can only be discerned when the chromosomes are in their attenuated state in interphase, and two separate (but close) spots identifying the duplicated segment are appreciated (longer arrow, right) in a substantial fraction of cells, while the normal chromosome shows a single spot (shorter arrow, right). |
del 17q11.2: Neurofibromatosis Plus
The common condition of neurofibromatosis type 1 (NF1) is typically due to mutation of the neurofibromin gene at 17q11.2. Rare cases of NF1 with dysmorphism and other signs beyond the NF1 phenotype reflect deletions extending beyond neurofibromin (Riva et al., 2000).
Chromosome 18
del 18p, del 18q
Quite substantial deletions of chromosome 18 short arm and long arm were recognized in the very early days of medical cytogenetics; the small size of this chromosome facilitated recognition of these deletions on solid-stain cytogenetics (De Grouchy et al., 1964, 1966). The phenotype has been refined, and attempts made at a clinical correlation according to the nature of the deletion; sophisticated studies may be necessary to distinguish simple terminal deletions from more complex rearrangements (Cody et al., 1999; Brkanac et al., 1998; Gunn et al., 2003). One particular locus whose haploinsufficiency may contribute to the mental defect is MBP (myelin basic protein) at 18q23. Gay et al. (1997) studied 20 patients with 18q–, and scanned them with magnetic resonance imaging (MRI) to determine how well the white matter of the brain was developed. In 19 patients in whom the MBP gene was included in the deletion, reduced myelination was demonstrated, whereas in the one patient whose deletion did not include MBP, a normal appearance of the cerebral white matter indicated that myelination had proceeded, at least grossly, without compromise.
Not all “simple deletions” may in fact be so. Horsley et al. (1998) used a FISH assay to analyze the subtelomeric regions of all chromosomes in a case of supposed del(18p), whose phenotype was judged to be atypical. She turned out to have 2p subtelomeric sequences on the tip of 18p, and her true karyotype was actually 46,XX,der(18)t(2;18)(p25;p11.2), rather than simply 46,XX,del(18)(p11.2). Thus, in addition to the 18p partial monosomy, she also had a partial 2p trisomy. This sort of rearrangement may not be uncommon (p. 268).
Chromosome 20
del 20p12: Alagille Syndrome
The characteristic features of this syndrome are stenosis of the peripheral pulmonary arteries and insufficient development of bile ducts within the liver (thus, arteriohepatic dysplasia), along with certain eye and skeletal defects and a distinctive facies (Krantz et al., 1997). Most patients have a mutation in the gene JAG1, but a very few have a cytogenetic defect. In one series of 109 subjects, only three had visible chromosomal abnormalities—two deletions and one unbalanced 4;20 translocation (Crosnier et al., 1999). The phenotype did not vary between those with whole gene deletions and those with point mutations, suggesting that haploinsufficiency is the common mechanism. A single report exists of transmission from a mosaic parent (Laufer-Cahana et al., 2002).
Chromosome 21
del 21q: Partial Monosomy 21
There are 21q deletions of varying degree, and this allows an assessment of the contribution of different haploinsufficient segments to the observed range of phenotypes. Deletion of the segment encompassing the APP (amyloid precursor protein) and SOD1 (superoxide dismutase-1) loci is particularly important (Chettouh et al., 1995). Most cases are sporadic, but some have occurred in the setting of a parental balanced translocation (Huret et al., 1995). (Full monosomy 21 has, in the past, been reported; but restudy with more powerful methodology has shown these to be, in fact, partial 21 monosomies due to unbalanced translocations [West and Allen, 1998].)
Chromosome 22
del 22q11.21–q11.23: 22q11 Deletion Syndrome
Before their common cytogenetic basis (Fig. 17-12) was understood, the 22q11 deletion syndromes had a number of labels, including DiGeorge syndrome (DGS), velocardiofacial (VCF) syndrome, and Shprintzen syndrome (De Decker and Lawrenson, 2001). What used to be called Kousseff syndrome we can now see as 22q11 deletion with the phenotype including neural tube defect (Forrester et al., 2002). In true acknowledgement of the first definition of the syndrome, in the Czechoslovakian literature in 1955, Sedlácková syndrome may be the most fitting name (Turnpenny and Pigott, 2001). With a birth incidence of about 1 in 4000, this is the most common human site of deletion, and this vulnerability may be a reflection of certain qualities of the DNA structure in 22q11. Blocks of low-copy repeats (LCRs) in the region have a tendency to appose, setting the stage for erroneous recombination.
|
|
|
Figure 17-12. Two chromosome 22q deletion syndromes, 22q11 (left) and 22q13 (right). The DiGeorge critical region probe identifies the proximal 22q11.21 deletion (left). One no. 22 homolog shows normal hybridization (longer arrow, left), while the deleted chromosome fails to hybridize (shorter arrow, left). Arrowheads show control probe. The distal 22q13.33 deletion (right) is indicated by the smaller arrow; longer arrow is normal homolog. |
There is a range of phenotypes for this syndrome, and Wulfsberg et al. (1997) maintain that the deletion produces variations on a theme, not a set of different themes. Increasing clinical experience may lead to a better recognition of more subtly dysmorphic cases. Zori et al. (1998) studied a group of patients presenting with velopharyngeal insufficiency, of whom 30% proved to have a deletion, and 13 medical geneticists who studied full-face photographs at a meeting achieved only a 62% correct assignment to del 22q11. Mehraein et al. (1997) suggest that about a quarter of children with an “isolated” complex cardiac malformation who test positive for a 22q11 deletion in fact show other features of the phenotype (especially the facial/endocrine/im-mune components), if these are carefully sought. Psychiatric disorder can be associated, and bipolar disorder and schizophrenia are particular concomitants (Bassett et al., 1998; Yan et al., 1998; Swillen et al., 2000). Some motor neurological dysfunction (drooling, abnormal speech, dysphagia) may result from a specific failure of development of the motor cortex of the brain (Bingham et al., 1997). In the familial case, a parent can, for example, show mild features of the condition, or have a predominantly Shprintzen facial and palatal phenotype, with a child showing a characteristic DGS cardiac and endocrine phenotype (Devriendt et al., 1997).
There seems not to be a clear correlation between the extent of the deletion and the clinical phenotype, and the fact that most monozygous twins are discordant is a perplexing observation (Amati et al., 1999; Singh et al., 2002b). It may be that there is not one universal DGS gene, but that various combinations of haplo-insufficiencies can lead to a similar, or a not so similar, clinical picture. Attractive candidates as haplo-insufficient genes include YPEL1, coding for a nuclear protein, and TBX1, a transcription-factor gene (Farlie et al., 2001; McDermid and Morrow, 2002). Alternatively, Singh et al. (2002b) speculate that variable second-hit epigenetic changes affecting gene expression on the remaining no. 22 might be contributory.
Most cases are de novo, but about 10% are inherited. It may be that earlier estimates of a larger fraction of affected parents were biased due to studying more remarkable families (Swillen et al., 1998). Indeed, Smith and Robson (1999) report only 5% of parents to have the deletion in an Australian series of 59 cases. Typically, 22q–parents show poor social functioning, and some have frank psychiatric disease. Presumed parental gonadal mosaicism has been described, and prenatal diagnosis using FISH in a subsequent pregnancy will cover this possibility (Hatchwell et al., 1998; Kasprzak et al., 1998; Bergman and Blennow, 2000; Sandrin-Garcia et al., 2002).
del 22q13
The particular trait is a failure to develop expressive language, and high pain tolerance is also notable; the physical phenotype comprises fairly soft dysmorphism. Most cases may be sporadic, but inherited translocations have been identified, as has an instance of parental mosaicism (Phelan et al., 2001). Of the contiguous genes involved, loss or compromise of proSAP2 may be an important contributor to the phenotype (Anderlid et al., 2002b). The diagnosis has been made incidentally when a distal 22q internal control probe, used at the time of diagnostic testing for the 22q11 deletion, failed to hybridize. As Phelan et al. observe, the simple expression “22q deletion syndrome” is now ambiguous; the proper name for this condition is 22q13 deletion syndrome.
Chromosome X
X chromosome deletions detectable on classical cytogenetics are discussed in Chapter 12.
del Xp21: Duchenne/Becker Muscular Dystrophy
In about two-thirds of Duchenne and Becker muscular dystrophy, the molecular defect is an intragenic deletion within the dystrophin gene at Xp21, and this can be identified by using a FISH probe made from cosmids that recognize particular exons (Ligon et al., 2000). Female carrier status may readily be appreciated if either both X chromosomes, or only one, shows hybridization with the particular probe that has shown nonhybridization in an affected male relative. This may be seen as an alternative approach to the use of intragenic DNA markers.
del Xp22.3: X-Linked Ichthyosis
A similar approach can be applied to the skin condition, X-linked ichthyosis, as most cases are due to large deletions of the steroid sulfatase gene and flanking sequences at Xp22.3 (Valdes-Flores et al., 2001).
DUPLICATION
Intrachromosomal Duplication: 46,dup
Most intrachromosomal duplications are presumed to arise de novo. Taking into account the possibility of parental gonadal mosaicism, a recurrence risk figure of around 1% (rather than the <0.5% we suggest for most rearrangements) may be appropriate. For the inverted duplication, 46,inv dup, we assume a low (<0.5%) recurrence risk. (Extremely rarely, an inverted duplication may, in fact, be due to recombination within a parental paracentric inversion [p. 157]. Paracentric inversions can be difficult to detect, and a careful and directed search may be appropriate.)
While there are certainly very many individual duplication cases on record, rather fewer duplication phenotypes have acquired eponymic (or acronymic) status than have deletions and we do not provide a catalog comparable to the listing of deletion syndromes above. We list merely the following few.
dup 14q23.4–q31
An interesting example of a duplication whose effect may depend on imprinting is reported by Robin et al. (1997). A father and daughter had the same chromosome abnormality, described as 46,dir dup(14)(q23.4q31). The father was normal, but the child was microcephalic with poor language development and a subtle facial dysmorphism. Perhaps the segment 14q23.4–q31 includes genes that are imprinted when maternally transmitted, but expressed when transmitted on the paternal chromosome. If so, the child would have had a functional partial paternal UPD 14, and this may be the explanation for her defects. Alternatively, this could be a harmless duplication, such as we discuss on p. 243, and the daughter's phenotype is coincidental. In any event, cases such as this with relatively small imbalances offer the opportunity to define specific chromosomal regions that may be subject to imprinting.
dup 15q11–q13: Syndrome of Retardation and Autism Without Dysmorphism
Duplications of the proximal region of 15q may cause a syndrome of intellectual impairment, ranging from borderline to severe (Bolton et al., 2001). Some individuals display behavior resembling autism, although this duplication makes only a minor (1%–2%) contribution to the generality of autism (van Karnebeek et al., 2002b; Keller et al., 2003). The poor behavior may fit the category of “pervasive developmental disorder, not otherwise specified” (PDD-NOS) (Thomas et al., 2003). Physical findings are usually unremarkable. The duplication includes, but is not as extensive as, the chromatin present in the large idic(15) (see above), and the phenotype apparently reflects this, in being not quite as severe as with the idic(15) (Torrisi et al., 2001). The Prader-Willi/Angelman critical region (PWACR) is contained within the duplicated segment, a point of practical usefulness in laboratory analysis. FISH analysis typically shows two spots on the abnormal chromosome, although in some instances molecular analysis is necessary to prove the point (Thomas et al., 2003). Some of these dup(15)s may have the same breakpoints as in the deletions of the Prader-Willi and Angelman syndromes (PWS/AS), possibly involving misalignment of the same duplicons, thus representing the countertype of the PWS/AS deletion. Both inter- and intrachromosomal mechanisms are implicated, and the process may occur on either parental homolog (Ji et al., 2000; Roberts et al., 2002). Note that this dup(15) is not to be confused with the 15q11.2–q13 euchromatic variant due to constitutional cytogenetic amplification, which does not involve the PWACR and which is nonpathogenic (p. 242).
As in the 15q11–q13 deletion, there is a parent-of-origin effect: the syndrome characteristically results when the duplication is transmitted from a heterozygous mother (or, if de novo, it is of maternal generation) but not when it is paternally transmitted (Yardin et al., 2002; Thomas et al., 2003). A biological correlate of this observation is that the UBE3A gene is expressed at a greater level from a maternal-originating duplication (Herzing et al., 2002). A very few cases of abnormal children with a dup(15)(q11q13)pat may reflect there having been an atypical imprinting in the region, or the observation may be coincidental. Whether a father heterozygous for the duplication has an increased risk of having a child with del(15) PWS is at this time quite speculative (Butler et al., 2002).
dup 15q26–qter: Syndrome of Overgrowth and Intellectual Disability
A growth control locus, the insulin-like growth factor receptor type 1 (IGFR1), is sited within distal 15q, at band q26. Duplication of this locus is associated with a syndrome of which overgrowth is a particular feature, along with macrocephaly and a degree of intellectual compromise (Nagai et al., 2002; Faivre et al., 2002). This contrasts with the growth retardation that characterizes deletions of this region (e.g. the ring 15 syndrome, see p. 181).
dup 17p11.2
This condition has the interesting history that it was not delineated in the traditional way in which a reasonably consistent phenotype is recognized, following which a specific karyotype is identified. Potocki et al. (2000a) looked at seven patients with the very nonspecific diagnosis of “developmental delay,” and found an increased band size in 17p11.2. (They were able to show that this was, in fact, the reciprocal of the Smith-Magenis deletion, as discussed above; see Fig. 17-1.) Most had a normal facies and mild to borderline mental retardation, and behavioral difficulties with hyperactivity and attention deficit disorder were commonly recorded. Even in retrospect, this clinical picture is barely sufficient to be labeled a “syndrome,” if that word is taken to mean, from its Greek derivation and de Gruchy's elegant interpretation, “un ensemble de signes dans un ensemble de personnes.”
dup 17p12: Charcot-Marie-Tooth Neuropathy
The most common form of Charcot-Marie-Tooth neuropathy (CMT) is due to the duplication of about 1.4 Mb in 17p11.2, which encompasses the PMP22 (peripheral myelin protein 22) gene (Nelis et al., 1999).7 It is the countertype of the deletion, noted above, which causes pressure-sensitive neuropathy. Typically, the duplication is not detectable on G-banding, but rarely it is subtly visible (Upadhyaya et al., 1993). The duplication leads to the production of a 150% amount of the PMP22 protein, and this excess affects the functioning of the peripheral nerve. The major effect is on the motor nerves, and weakness is the important consequence. The nerves to the peroneal muscles (on the outside of the leg, with tendons passing around the ankle to the foot) are particularly vulnerable, and an alternative name for the condition is peroneal muscular atrophy. A FISH probe recognizing sequences within the duplicated segment hybridizes twice to the duplicated chromosome and is seen as two adjacent fluorescent spots, with the third spot due to fluorescence from the other chromosome appearing elsewhere in the nucleus (Fig. 17-11) (Shaffer et al., 1997). It is necessary to do the analysis on interphase cells to achieve a recognizable separation of the closely adjacent FISH signals. Charcot-Marie-Tooth neuropathy due to this duplication is called type 1A; other genetic forms of CMT are, of course, not recognized by this FISH test.
An intriguing example is that of a de novo X-autosome translocation 46,X,der(X)t(X;17) (p22.1;p11.2) in a mildly retarded female who had CMT (King et al., 1998). The extra segment of 17p attached to Xp produced an attenuated picture of partial 17p trisomy, presumably reflecting an extension of inactivation into the 17p segment from the X inactivation center of the der(X). The PMP22 gene on the 17p segment was apparently fully functioning, however, since the neuropathy was typical for CMT. The inactivation process could be supposed to have “jumped over” the PMP22 region (this process discussed on p. 103).
Prenatal diagnosis can be done using FISH on noncultivated interphase chorionic villus cells (Lebo, 1998; Kashork et al., 1999). Lebo proposes that prenatal diagnosis should be made available for CMT, even though CMT can be a relatively mild handicap, and comments: “Given the slow rate of progress toward curing all forms of human genetic disease, patients with degenerative diseases who already have irreversible nerve pathology should not be offered undue hope for intervention by gene therapy.” Couples will make their own decisions.
dup 22q11
The countertype of the common del 22q11 is a duplication for the same segment (Edelman et al., 1999). Theoretically, the dup 22q11 should be similarly common, but it is not, or to be precise, it is not commonly recognized. Indeed, only one family is recorded, in which two of the three heterozygotes were phenotypically normal. Thus its apparent rarity may reflect that it is often innocuous, or nearly so.
dup Xq22: Pelizaeus-Merzbacher Disease
This is a disease of the white matter of the brain, and presents a severe neurodegenerative clinical picture. It is X-linked, and the genetic defect resides in the PLP (proteolipid protein) locus at Xq22. Proteolipid protein is a major structural component of myelin in the central nervous system. A quite common mutational basis of Pelizaeus-Merzbacher disease is duplication of about a megabase of DNA at Xq22, which includes the PLP gene (Woodward et al., 1998). This is analogous to Charcot-Marie-Tooth neuropathy type 1A (CMT 1A): there is an additional copy of a gene that produces myelin, but in this case the myelin that sheathes of axons of the central nervous system. As with CMT 1A, interphase FISH in Pelizaeus-Merzbacher disease may show adjacent spots on the X, reflecting the tandem nature of the duplication. Since most cases can be shown to have arisen in the maternal grandfathers of affected boys, it appears that an intrachromosomal event in male gametogenesis is the usual source of the defect, and an unequal sister-chromatid exchange would be the likely mechanism. A single case is known of a pericentric inversion (X)(p11.4q22.1), in which the PLP gene was duplicated, with one copy at each inversion breakpoint—a noncontiguous duplication (Hodes et al., 2000). FISH can be exploited for prenatal diagnosis (Inoue et al., 2001b; Garbern and Hobson, 2002). (In about a quarter of cases a point mutation in the PLP gene is found, and in many, no genetic defect can be identified.)
Interchromosomal Rearrangement with Duplication: 46,rea
The great majority of these rearrangements arise de novo, presumably from illegitimate meiotic recombination between nonhomologs, and a low (<0.5%) recurrence risk applies. The possibility of occult parental gonadal mosaicism warrants consideration of prenatal diagnosis.
Extra Structurally Abnormal Chromosome: 47, + ESAC
Parental mosaicism is unlikely, but not completely excluded, if the parental lymphocyte karyotypes are normal. Although a risk for recurrence will be small (which we cannot precisely define, although we presume a low, single-digit per cent figure, possibly <1%), prenatal diagnosis may be appropriate.
Isochromosomes, and the “Robertsonian Isochromosome”
A couple having had a child with an isochromosome for a chromosome other than an acrocentric can generally be given encouraging advice, especially if the child is mosaic. The major mechanisms of generation operate either at meiosis II or postzygotically, and in either case no discernibly increased risk would be implied. Exceptions to this (noted in the Biology section above), presumably due to parental gonadal mosaicism, are extremely rare. The unbalanced “Robertsonian isochromosome” is a different category and is to be seen in a different light. Gonadal mosaicism has been directly demonstrated (p. 255) and a small increased risk would, in theory, apply.
TRIPLICATION
Triplications would be expected to be de novo events, and not predisposed to recur.
Notes
1. This word can be and is used in the context of a single locus. It is usually applied in cytogenetics to the more recently delineated “microdeletion” syndromes, but it can in principle refer to the classical syndromes with larger deletions.
2. Magenis et al. (1999) record the historical point that Pallister-Killian syndrome was first identified serendipitously, when fibroblasts taken for archival purposes were subject to routine cytogenetic analysis.
3. The presence of two supernumerary idic(15) chromosomes offers the rare opportunity to use the word “hexasomy” (Qumsiyeh et al., 2003). The same principles apply in terms of presence/absence of SNRPN, with the abnormal phenotype being aggravated in the hexasomic state.
4. NSD1 has a quite different role in cancer, as one of the two genes at the breakpoints of a somatic translocation t(5;11)(q35;p15.5) that leads to the production of a chimeric protein that may well be an initiating factor in childhood acute myeloid leukemia (Jaju et al., 2001).
5. Another deletion syndrome with exostosis is del(11)(p11.2p12), in which the EXT2 gene is presumably one of a set of contiguous genes removed (Ligon et al., 1998).
6. The diagnosis may also be made by molecular means. Wilke et al. (2000) used this locus to trial real-time polymerase chain reaction as a methodology, and suggested that this may have a wider applicability to other dup and del syndromes.
7. The duplicated segment of chromosome 17 contains certain other genes along with the PMP22 gene (Inoue et al., 2001a). But the genetic imbalance from the trisomic state of these genes is apparently without any phenotypic effect. These genes may belong (as it is beginning to seem) to a majority in which a 150% level of activity makes no appreciable difference.
