Chromosome Abnormalities and Genetic Counseling , 3rd Edition

4. Autosomal Reciprocal Translocations

Reciprocal translocations are common, and every counselor can expect to see translocation families. The usual form is the simple, or two-way, reciprocal translocation: only two chromosomes, usually autosomes, are involved, with one breakpoint in each. It is this category we consider in this chapter. The special cases of translocations involving sex chromosomes and those of complex translocations are dealt with in separate chapters.

The translocation heterozygote (carrier) may have a risk to have a child who would be mentally and physically abnormal because of a “segmental aneusomy.” Typically, the imbalance is due to a segment of one of the participating chromosomes being duplicated, and a segment of the other chromosome being deleted. This confers a partial trisomy and a concomitant partial monosomy. A few translocations are associated with a high risk, as much as 20% or rarely higher, of having an abnormal child. Many translocations imply an intermediate level of risk, in the region of 5%–10%. Some carriers have a low risk, of 1% or less; but the woman who is a carrier or the partner of a male carrier may have a high miscarriage rate. Other translocations imply, apparently, no risk of having an abnormal child, but the likelihood of miscarriage is high. Still others, discovered fortuitously, seem to be of no reproductive significance, with carriers having no difficulties in conceiving or carrying pregnancies and having normal children. The counselor needs to distinguish these different functional categories of translocation, in order to provide each family with tailor-made advice.

BIOLOGY

Simple reciprocal translocations arise when a two-way exchange (hence reciprocal, t) of material takes place between two chromosomes. The process of formation follows the physical apposition of a segment of each chromosome, which may have been promoted by the presence in each segment of a similar DNA sequence. A break occurs in one arm of each chromosome, and the portions of chromosome material distal to the breakpoints switch positions. The portions exchanged are the translocated segments; the rest of the chromosome (which includes the centromere) is the centric segment. The rearranged chromosome is called a derivative (der) chromosome. It is identified according to which centromere it possesses, as in the der(5) and der(10) depicted in Figure 4-1. When no loss or perturbation of genetic material occurs—in other words, the translocation is balanced—the phenotype of the heterozygote is normal, other things being equal. Approximately 1 person in 500 is a reciprocal translocation heterozygote (Van Dyke et al., 1983; Jacobs et al., 1992). The translocation may have arisen de novo in the consultand, or may be widespread throughout a family, with many carriers, and sometimes of centuries-long duration. Koskinen et al. (1993) trace a t(12;21) in western Finland back to a couple born in 1752!

Figure 4-1. Reciprocal translocations demonstrating (a) double-segment and (b) single-segment exchange. The translocations are t(5;10)(p13;q23.3) and t(1;4)(q44;q31.3). (Cases of M. A. Leversha and N. A. Monk.)

When one of the translocated segments is very small and comprises only the telomeric cap of a chromosome arm—and thus we suppose contains no genes—this is regarded as being, effectively, a single-segment exchange. The t(1;4) translocation shown in Figure 4-1, involving a substantial piece of no. 4 long arm exchanging positions with the terminal tip of a no. 1 long arm, exemplifies single-segment exchange. When both translocated segments are of substantial size, we refer to this as a doublesegment1 exchange. The translocation shown in Figure 4-1 between a no. 5 and a no. 10, with breakpoints in about the mid-short arm of chromosome 5 and a little below the middle of the chromosome 10 long arm, is an example of a double-segment exchange. The translocation involving breakpoints right at or actually within the centromere, with exchange of entire arms, is a particular and rare type of double-segment exchange known as the whole arm translocation (Kleczkowska et al., 1986; Gravholt et al., 1994; Tümer et al., 1995).

DETAILS OF MEIOTIC BEHAVIOR

At meiosis I, the four chromosomes with segments in common come together as a four-some—a “quadrivalent.” To match homologous segments, the four chromosomes must form a cross-shaped configuration. This is most clearly seen when the chromosomes are at the pachytene stage (Fig. 4-2). As meiosis progresses, the four components of the quadrivalent release their points of attachment except at the tips of the chromosome arms, and they form a ring; if attachment fails, or if one of the erminal pairings release, a chain forms instead of a ring. With breakdown of the nuclear envelope, spindles forming at each pole of the cell can track to the equator and seek attachment to the centromeres. A cellular motor comes into play, and the chromosome travels to one or other pole. According to which spindle attaches to which centromere—and this may in part be influenced by the configuration of the ring or chain—the distribution of the four homologs is determined. The distribution of homologs to either pole is referred to as segregation. The expression 2:2 segregation describes two chromosomes going to one cell and two to the other. In 3:1 segregation, three chromosomes go to one cell, and one to the other. In 4:0 segregation, all four chromosomes go to one cell and none to the other.

Modes of Segregation

Within these broad categories we can list the particular modes of segregation, according to which chromosomes actually go where. Referring to the four chromosomes of the quadrivalent as A, B, A, and B (Fig. 4-2) the modes of segregation are summarized as follows:

One gametocyte with

Other gametocyte with

Segregation mode

2:2 Segregations

A & B

A′ & B′

Alternate segregation

A & B′

B & A′

Adjacent-1 segregation

A & A′

B & B′

Adjacent-2 segregation

3:1 Segregation

A B & A′

B′

3:1 segregation with tertiary trisomy or monosomy

A B & B′

A′

A′ B′ & A

B

3:1 segregation with interchange trisomy or monosomy

A′ B′ & B

A

4:0 Segregation

A B & A′ B′

None

4:0 segregation with double trisomy or monosomy

Figure 4-3 depicts five of the eight possible pairs of daughter gametocytes. Other things being equal, the chromosomal combination is conserved through meiosis II, and the mature gamete forms. Gametes from alternate segregation are normal or balanced. Conceptions from adjacent-1 gametes have trisomy for one translocated segment and monosomy for the other. Vice versa, adjacent-2 conceptions have trisomy for one centric segment, and monosomy for the other. Tertiary aneuploidies have trisomy, or monosomy, with respect to the combined chromosomal content of one of the derivative chromosomes. Interchange aneuploidies have a full autosomal trisomy or a full monosomy. In 4:0 segregation, there is a double trisomy or a double monosomy. Some of the gametes with these unbalanced combinations may be viable, in the sense of being “capable of giving rise to a conceptus, which would proceed through to the birth of a child.” Mostly, in fact, they are not. The combinations producing interchange monosomy and the (theoretical) 4:0 segregants are never viable.

Figure 4-2. Pachytene configuration, in simplified outline. The two normal (A, B) and two translocation (A′, B′) homologs align corresponding segments of chromatin during meiosis I.

Recombination at meiosis I, and asymmetric segregation at meiosis II, can complicate the story. If recombination occurs in the interstitial segment (between the centromere and the breakpoint), further unbalanced combinations are generated, most of which would not be remotely viable. This phenomenon may possibly have some practical relevance in preimplantation genetic diagnosis, since testing is done at a stage when there has been little opportunity for selective pressure to have applied. Scriven et al. (1998) list many of these recombination possibilities, and Van Hummelen et al. (1997) diagram the process with respect to a particular translocation on which they had undertaken sperm studies (and illustrate the point that a normal/balanced gamete can be restored following recombination in adjacent-1 segregation). The most telling evidence that recombination can happen comes from the observation of a meiosis I chromosome having one normal and one derivative chromatid, and polar body analysis has enabled such an observation to be made (Munné et al., 1998a). At meiosis II, asymmetric segregation may lead to two copies of a derivative chromosome being transmitted, as noted below in the section Meiosis II Nondisjunction.

Figure 4-3. The categories of 2:2 and 3:1 segregation that may occur in gametogenesis in the translocation heterozygote. Pairs of complementary daughter gametocytes are placed side by side. In the four 3:1 categories, only one of the two possible combinations in each category is depicted (both of each are shown in Fig. 4-4).

Alternate Segregation

In 2:2 alternate segregation, looking at each centromere in turn around the quadrivalent, one centromere goes to one pole, and the next centromere goes to the other pole. In other words, each centromere goes alternately to one or the other pole. Thus, the two daughter cells come to contain, respectively, the two normal homologs in one, and the two derivative chromosomes in the other. Note that alternate segregation is the only mode that leads to gametes with a complete genetic complement—one with a normal karyotype, the other with the reciprocal translocation in the balanced state. All other modes can be classified as malsegregation.

Adjacent Segregation

In 2:2 adjacent segregation, adjacent centromeres travel together (“adjacent” in the sense of centromeres being next to each other, in their positions around the quadrivalent). There are two categories. In adjacent-1 segregation, adjacent chromosomes with unalike (nonhomologous) centromeres travel to the same daughter cell (an aide-mémoire: in adja-cent-1, the daughter cells get one of each centromere). Overall, adjacent-1 is the most frequently seen mode of malsegregation in the children of translocation heterozygotes. In ad-jacent-2 segregation, which is rather uncommon, adjacent chromosomes with like (homologous) centromeres go to the same daughter cell (another aide-mémoire: in adjacent-2, the two homologous centromeres go together). Thus, adjacent-2 segregation rather resembles nondisjunction.

3:1 Segregation

This is also referred to as 3:1 nondisjunction. Gametes with 24 chromosomes and 22 chromosomes are formed, and the conceptuses therefore have 47 or 45 chromosomes. Almost always, the 47-chromosome conceptus is the only viable one. Two categories exist: either the two normal chromosomes of the quadrivalent plus one of the translocation chromosomes go together to one daughter cell (tertiary trisomy) or, rarely, the two translocation chromosomes and one of the normal chromosomes segregate (interchange trisomy). Tertiary monosomy, with a 45-chromosome conceptus, is extremely rare. Interchange monosomy has never been seen (except at preimplantation genetic diagnosis, PGD).

4:0 Segregation

In autosomal translocations, 4:0 segregation has been regarded as being of academic interest only. But it may come to have some practical relevance in PGD.

In theory, 16 possible chromosomal combinations could be produced in the gametes of the autosomal translocation heterozygote. For the most part, we can ignore four of these (3:1 interchange monosomies and 4:0 segregants), because they are never viable. The two balanced gametes (2:2 alternate segregants) are always viable, other things being equal. Of the remaining 10 possibilities, it is common for none to be viable, with spontaneous abortion the universal outcome. If a translocation heterozygote does have the possibility of viable imbalance in an offspring, it is most likely that there will be only one such combination (in 99% and 100% of translocations, in the considerable experience of two groups; Scriven et al., 1998). Usually, this sole survivable imbalance will be one that endows a partial trisomy. Infrequently two and, very rarely, more than two may be viable. Figure 4-4 depicts the various combinations that may be considered (using the previously discussed t(1;4) translocation as an example). In a review of 1159 translocation families, Cohen et al. (1994) found the proportions of chromosomally unbalanced offspring as follows: 71% adjacent-1, 4% adjacent2, 22% tertiary trisomy/monosomy, and 2.5% interchange trisomy.

Gamete Studies

Apparently, it is the norm for the heterozygote to produce gametes in which many of the possible chromosomal combinations occur, although the proportions may differ for some different translocations. Sperm karyotyping results from 27 men heterozygous for a translocation are summarized in Table 4-1, along with oocyte karyotyping data (in most indirectly via polar body analysis) from 7 women. The great majority, if not all, of these studied individuals would have presented to the clinic because of reproductive difficulty, and so the data may possibly be biased in the direction of unbalanced forms, compared to the whole population of translocation heterozygotes. On average, alternate and adjacent-1 segregants are the predominant types in spermatogenesis, occurring in fairly similar fractions (47% and 37%, respectively). Adjacent-2 at 12% and 3:1 at 5% are less frequently seen; and just one individual had a single 4:0 segregant sperm. Considerable variation occurred: some heterozygotes had no 3:1 segregants, and one had 21%; for adjacent-2, the range is 0% to 31%. Very similar figures are recorded in the review of Guttenbach et al. (1997), totaling 4445 sperm analyzed from 36 men. It would not be surprising if a fairly similar distribution and range of germ cell abnormalities were produced by their heterozygous sisters, although early data from preimplantation diagnosis research do suggest somewhat of a (possibly age-related) propensity in oogenesis for 3:1 segregation (Tables 4-1 and 4-2).

Figure 4-4. The full range of segregant gametes that may be produced by the translocation heterozygote, using the t(1;4) depicted in Figure 4-1 as an example. No. 1 chromatin is shown open, no. 4 chromatin is cross-hatched.

Conceptions

It might be expected that the distribution of normal and abnormal conceptions would reflect the distributions of karyotypes in the gametes. Thus, the two men in Table 4-1 with more than 60% alternate segregant forms, who were heterozygous for a t(6;14) and a t(10;12), respectively, might logically be presumed to have a better chance of having a normal child than, for example, the man with a t(1;11) in whose sperm only 33% showed a normal or balanced karyotype. This may be true in some cases, but not necessarily so in all instances, according to some preliminary data. A brother and a sister reported in Coonen et al. (2000), both with a t(3;11)(q27.3;q24.3), each suffered reproductive loss. The brother had 45% alternate forms of segregation on sperm analysis, but of 18 biopsied day 3 cleavage embryos, only 3 (17%) were normal or balanced. (One of these tested embryos was implanted, and a normal daughter with the balanced translocation subsequently was born.) Thus, the fraction of normal/balanced karyotypes fell by one-third from the gametes to the conceptions. The predominance of adjacent forms in the sperm of the t(11;22) heterozygote listed in Table 4-1 is not reflected in embryo analyses, in which, overall, alternate segregations are more common (Table 24-2). A definitive answer to this general question is awaited. Data from embryo studies in the IVF clinic such as those presented in Tables 4-2 and 24-2 will be instrumental; thus far, in comparing the pooled segregant distributions of gametes and embryos in each sex, the numbers overall are not grossly discrepant, except perhaps in the 3:1 category (cf. the average fractions in Tables 4-1 and 4-2).

Viability in Utero

Most unbalanced combinations would produce such enormous genetic imbalance that the conceptus would be lost very early in pregnancy (occult abortion) or even fail to implant. Moderate imbalances would proceed to the stage of recognizable miscarriage or to later fetal death. Only those conceptuses with lesser imbalances may result in the birth of an abnormal child.

Table 4.1. Chromosome Segregations in Gametes of 34 Reciprocal Translocation Heterozygotes

t

Alt
(%)

Adj-1
(%)

Adj-2
(%)

3:1
(%)

4:0
(%)

Male Heterozygotes

46,XY,(1;2)(q32;q36)

41

42

6

11

0

46,XY,(1;4)(p36.2;q31.3)a

46

38

7

9

0

46,XY,(1;4)(p36.2;q31.3)a

39

50

8

3

0

46,XY,(1;9)(q22;q31)

46

38

13

4

0

46,XY,(1;11)(p36.3;q13.1)

33

43

16

8

0

46,XY,(2;5)(p11;q15)

5

21

0

46,XY,(2;9)(q21;p22)

43

28

24

4

0

46,XY,(2;17)(q35;p13)

56

33

11

0

0

46,XY,(2;18)(p21;q11.2)

42

35

14

8

0

46,XY,(3;8)(p13;p21)

34

44

21

1

0

46,XY,(3;15)(q26.2;q26.1)

48

36

12

2

2

46,XY,(3;16)(p23;q24)

37

41

16

5

0

46,XY,(4;6)(q28;p23)

46

52

2

1/2

0

46,XY,(4;17)(q21.3;q23.2)

57

35

7

2

0

46,XY,(5;7)(q13;p15.1)

40

26

17

17

0

46,XY,(5;13)(q11;q33)

77

21

2

0

0

46,XY,(5;18)(p15;q21)

81

16

0

3

0

46,XY,(6;7)(q27;q22.1)

51

49

0

0

0

46,XY,(6;14)(p24;q22)

68

32

0

0

0

46,XY,(7;14)(q21;q13)

53

32

16

0

0

46,XY,(8;15)(p22;q21)

37

38

21

4

0

46,XY,(9;18)(p12;q12.1)

34

63

0

2

0

46,XY,(10;12)(q26.1;p13.3)

61

26

7

6

0

46,XY,(11;17)(p11.2;q12.3)

39

32

27

3

0

46,XY,(11;22)(q23;q11)

23

39

23

15

0

46,XY,(12;20)(q24.3;q11)

47

42

10

2

0

46,XY,(16;19)(q11.1;q13.3)

40

28

31

1

0

Average fractionsb (sperm)

47

37

12

5

0

Female Heterozygotes

46,XX,(2;14)(q23;q24)

11

22

22

44

0

46,XX,(2;14)(q31;q24)

14

57

14

14

0

46,XX,(4;14)(p15.3;q24)

27

55

0

18

0

46,XX,(6;21)(q13;q22.3)

0

0

0

100

0

46,XX,(7;20)(q22;q11.2)

50

17

0

33

0

46,XX,(9;11)(p24;q12)

100

0

0

0

0

46,XX,(14;18)(q22;q11)

40

0

60

0

0

Average fractionsb (ova)

30

30

14

26

0

Alt, alternate; adj-1, adjacent-1; adj-2, adjacent-2.
aCousins.
bSperm: average of the percentages. Ova: average of pooled oocyte observations.
Sources: Most of the sperm data are from references in Estop et al. (1995b); further sperm data are recorded in Guttenbach et al. (1997) and Faraut et al. (2000). Oocyte/polar body data, which naturally are based on much smaller num- bers (2–11 observations per woman), are from Munné et al. (1998c, 1998d), Conn et al. (1999), and Escudero et al. (2000b).

Viability is much more likely in the case of effective single-segment imbalance, with only one segment of substantial size. In the unbalanced state, a partial monosomy or trisomy for the other very small terminal segment is likely to contribute minimally (if it contains no genes or, at any rate, no “dosage-vulnerable” genes2) or not at all to the overall imbalance. This is of particular relevance in adjacent-1 segregation. Consider, for example, gamete (3) in Figure 4-4.

The material missing from the telomeric tip of no. 1 long arm, the telomeric cap, is so small that its loss is, as far as we can tell, of insignificant phenotypic effect. For practical purposes, we can ignore this partial monosomy. So the significant imbalance reduces to a partial 4q trisomy (trisomy 4q31.3–qter). This is well recognized as being a viable complement (and it is the imbalance in the children whose photograph appears in the frontispiece). In the double-segment exchange, by contrast, the imbalance contributed by each segment must be taken into account. Thus, adjacent-1 gametes have both a partial trisomy and a partial monosomy to a significant degree, and would produce a “phenotypic hybrid.” Very frequently, the combination is nonviable.

Table 4.2. Chromosome Segregations in Embryos of 16 Reciprocal Translocation Heterozygotes Studied at Preimplantation Diagnosis (Shown as Actual Numbers in Each Segregant Category). See also Table 24-2.

t

Alt

Adj-1

Adj-2

3:1

4:0

Male heterozygotes

46,XY,t(3;6)(q25;q23)

3

7

2

1

0

46,XY,t(3;7)(q23;q36)

3

1

3

0

0

46,XY,t(3;7)(q25.3;p22.1)

4

1

3

1

0

46,XY,t(11;22)(q23.3;q11.2)

9

4

1

0

1

Average fractions

43%

30%

20%

5%

2%

Female heterozygotes

46,XX,t(1;13)(q23;p11)

0

1

0

1

0

46,XX,t(1;19)(q32.1;q13.1)

11

1

3

6

0

46,XX,t(2;4)(p22.2;q33)

2

5

0

0

0

46,XX,t(3;5)(p12;q14.2)

2

1

0

0

0

46,XX,t(4;15)(q26;q13)

0

0

0

1

0

46,XX,t(5;14)(p15.1;q32.1)

4

2

0

3

0

46,XX,t(8;18)(p21.1;q21.1)

2

0

0

0

0

46,XX,t(9;20)(q34.2;q11.2)

5

3

0

0

0

46,XX,t(11;17)(p15.5;p13)

9

7

0

5

1

46,XX,t(11;22)(q23.3;q11.2)

0

0

0

2

1

46,XX,t(12;17)(p13;p13)

11

4

0

2

0

46,XX,t(14;22)(q11.2;q13.3)

6

0

3

1

0

Average fractions

50%

23%

6%

20%

2%

Alt, alternate; adj-1, adjacent-1; adj-2, adjacent-2. Average fractions are derived from pooling the data in each group.
Source: From Mackie Ogilvie and Scriven (2002).

If very early miscarriages could be karyotyped, one might expect to discover more of the imbalanced forms. Fritz et al. (2000) conducted such an exercise, using comparative genomic hybridization as the cytogenetic tool. They identified a family segregating a subtle t(4;12)(q34;p13), in which two children were born with 46,der(4),t(4;12)(q34;p13), giving a distal 4q monosomy. There had been five previous abortions, and archival pathology material (paraffin-embedded placental tissue) was available for analysis from three of these. A 12- and a 17-week abortion both showed the same karyotype as the surviving children. An 8-week abortion described as a hydatidiform mole karyotyped as a tertiary trisomy for almost the whole of chromosome 4: 47,XY,der(4),t(4;12)(q34;p13), combination (9) in Figure 4-4.

Predicting Segregant Outcomes

How can we determine, for the individual translocation carrier, which segregant outcomes, if any, might lead to the birth of an abnormal child? What might be the relative roles of an inherent tendency for a particular type of segregation to occur and of in utero selection against unbalanced forms? A useful approach is to imagine how the chromosomes come to be distributed during meiosis. Following Jalbert et al. (1980, 1988), we may draw, roughly to scale, a diagram of the presumed pachytene configuration of the quadrivalent and then deduce which modes of segregation are likely to lead to the formation of gametes, which could then produce a viable conceptus. The following, with reference to Figure 4-5, are the ground rules:

1. We assume that alternate segregation is (a) frequent and (b) associated with phenotypic normality.

2. The least imbalanced, least monosomic of the imbalanced gametes is the one most likely to produce a viable abnormal conceptus.

3. If the translocated segments are small in genetic content, adjacent-1 is the most likely type of malsegregation to be capable of giving rise to viable abnormal offspring (Fig. 4-5a).

4. If the centric segments are small in content, adjacent-2 is the most likely segregation to give a viable abnormal outcome (Fig. 4-5b).

5. If one of the whole chromosomes of the quadrivalent is small in content, 3:1 disjunction is the most likely (Fig. 4-5c). The small chromosome may be a small derivative chromosome, or a chromosome 13, 18, or 21.

6. If the quadrivalent has characteristics of both Rules 3 and 5, or of Rules 4 and 5, then both adjacent and 3:1 segregations may give rise to viable offspring.

7. If the translocated and centric segments both have large content, no mode of segregation could produce an unbalanced gamete that would lead to a viable offspring (Fig. 4-5d).

8. Subtelomeric translocations may not necessarily form a quadrivalent, and the pairs of homologs might simply join up as bivalents, each pair then segregating independently.

Figure 4-5. Prediction of likely viable segregant outcomes by pachytene diagram drawing and assessment of the configuration of the quadrivalent.

Some examples to illustrate these points follow.3

Adjacent-1 Segregation, Single-Segment Exchange

Many translocations involve an effectively sin-gle-segment exchange, with the “single” translocated segment comprising a fairly small amount of chromatin (1%–2% of the haploid autosomal length [HAL]). This is the classical scenario for adjacent-1 segregation to occur and for producing a phenotype which may be capable of postnatal survival. The father with the t(1;4) in Figure 4-1, whose children with partial 4q trisomy are shown in the frontispiece as discussed above, is an example.

Figure 4-6. Adjacent-1 segregation. (a) Pedigree of a family in which a t(3;11)(p26;q21) segregates, having the characteristics associated with adjacent-1 malsegregation. Two independently ascertained probands have a partial 11q trisomy, and a deceased relative, who died at age 18 in an institution for the retarded, had a similar appearance from photographs, and so very probably had the same karyotype. Filled symbol, unbalanced karyotype; half-filled symbol, balanced carrier; N in symbol, 46,N; small diamond, prenatal diagnosis; arrow, proband. (b) Partial karyotype of a translocation heterozygote (upper), showing the 3;11 translocation, and a child with the unbalanced complement (lower) (case of A. J. Watt). (c) The presumed pachytene configuration during gametogenesis in the heterozygote (no. 3 chromatin is open, no. 11 chromatin is cross-hatched). Arrows indicate movements of chromosomes to daughter cells in adjacent-1 segregation; heavy arrows show the combination observed in this family.

Consider now the family whose pedigree is depicted in Figure 4-6a, in which the individuals shown as heterozygotes have the balanced translocation 46,t(3;11)(p26;q21). A segment of chromatin consisting of almost half of the long arm of chromosome 11, and comprising 1.4% of the HAL, is translocated to the tip of chromosome 3 short arm (Fig. 4-6b). The telomeric tip of chromosome 3 short arm, which we imagine to comprise little or no phenotypically important genetic material, has moved reciprocally across to chromosome 11. The presumed pachytene configuration during gametogenesis in the heterozygote would be as drawn in Figure 4-6c. The adjacent-1 segregant gamete with der(3) plus normal 11 (heavy arrows) produces a conceptus that has a partial 11q trisomy, since the der(3) carries the segment 11q21–qter. The loss of the 3p telomeric tip in this der(3) we presume to have no effect. Two, probably three, children in the family were born with this karyotype. No individuals are known to have the other adjacent1 combination (Fig. 4-6c, light arrows)—that is to say, the 46,der(11) karyotype, which would endow a partial 11q monosomy. Consulting Schinzel (2001), viability for the segment 11q21–qter in monosomic state is recorded in only two cases. We assume, therefore, that it has a very high lethality in utero.

The scenario of a single survivable imbalanced form due to a partial trisomy from adja-cent-1 segregation in a single-segment translocation, as in this t(3;11) example, is, as mentioned above, the most commonly encountered circumstance in translocation families at risk for an abnormal child.

Infrequently, both the partial trisomic and the partial monosomic forms are observed. A good example of this is given by distal 4p translocations: both deletion and duplication for this segment are well recognized as having substantial in utero viability. Consider the translocation t(4;12)(p14;p13) described in a family study by Mortimer et al. (1980). A number of family members over three or more generations were balanced carriers, and abnormal children had been born with typical Wolf-Hirschhorn syndrome (all dying in infancy), while others presented the syndrome of partial 4p trisomy (all surviving at least well into childhood). The breakpoints of the translocation are in distal 4p and at the very tip of 12p (12pter). The presumed pachytene configuration would be as drawn in Figure 4-5a (with the no. 4 chromatin being open, and the no. 12 chromatin cross-hatched). With such short translocated segments (and very long centric segments), adjacent-1 segregation is the only possibility for viable imbalance. If we ignore the tiny contribution of a duplication or deletion for telomeric 12p—in other words, if we interpret this as an effective single-segment imbalance—the situation reduces to the two possible adjacent-1 outcomes being a partial 4p trisomy and a partial 4p monosomy. Both of these are recognized entities, as noted above, and apparently both have substantial viability in utero. The abnormal karyotypes would be written 46,der(12)t(4;12)(p14;p13) and 46,der(4)t(4;12)(p14;p13).

Adjacent-1 Segregation, Double-Segment Exchange

With a double-segment translocation, an adja-cent-1 imbalanced conceptus has both a partial trisomy and a partial monosomy (also called a duplication/deficiency, or duplication/deletion, abbreviated to dup/del). The combined effect of the two imbalances is more severe than either separately. Thus, it is infrequent that the carrier of a double-segment exchange can have a chromosomally unbalanced pregnancy proceeding through to term or close to term. Multiple miscarriage is the typical observation (e.g., Figure 4-16). But occasionally viability is observed for both of the dup/del combinations. The double-segment t(4;8)(p16.1;p23.1) depicted in Figure 4-7 has very small translocated segments: the tip of chromosome 4 and the tip of chromosome 8 have exchanged positions.4 In this family, each of the adjacent-1 segregant outcomes was observed: the index case with del(4p)/dup(8p), and his uncle with dup(4p)/del(8p). In the former, a Wolf-Hirschhorn gestalt was discernible, reflecting the del(4p) component. A similar example is seen in the family reported by Rogers et al. (1997), who provide a photograph of six siblings sitting on a sofa in 1958; one with a dup(11q)/del(4q) karyotype, two who have since died and are presumed to have been del(11q)/(dup(4q), and one girl carrying the family t(4;11)(q34.3;q23.1), who went on to have a del(11q)/(dup(4q) child in the next generation.

Exceptionally, both translocated segments can be of substantial size and yet be survivable, if barely, to term. The outlying points in Figure 4-17 reflect such cases. The double-seg-ment t(5;10)(p13;q23.3) exchange illustrated in Figure 4-1 provides an example, this translocation having been identified in a family following the death of a neonate with multiple malformations. The genetic abnormality comprises a deletion of 5p and a duplication of 10q, for a total imbalance of 2.5% HAL (1.1% HAL monosomy plus 1.4% HAL trisomy). When entire arms of chromosomes are translocated (whole-arm translocation), it is almost always the case that the unbalanced segregants are unviable. A rare exception exists in the case of Czakó et al. (2002), a t(18;20)(p11.1;p11.1), in which the abnormal child of a carrier father was effectively trisomic for all of 20p and monosomic for all of 18p (1.0% HAL trisomy plus 0.8% HAL monosomy).

Figure 4-7. Adjacent-1 segregation, double-segment translocation with very small segments. (a) Parent with the translocation t(4;8)(p16.1;p23.1). The index case, his child, has the karyotype 46,der(4) and so has a del(4p)/dup(8p) imbalance, and an uncle has the countertype dup(4p)/del(8p) imbalance due to the 46,+der(8) karyotype (case of C. E. Vaux). (b) The presumed pachytene configuration during gametogenesis in the heterozygote (no. 4 chromatin is open, no. 8 chromatin is cross-hatched). Arrows indicate movements of chromosomes to daughter cells in adja-cent-1 segregation. The upper combination (light arrows) would produce the dup(4p)/del(8p) imbalance, and the lower (heavy arrows), the del(4p)/dup(8p) imbalance.

The opportunity occasionally arises to provide direct evidence of early in utero lethality of a particular imbalanced state. In a family study of a t(8;18)(p21.3;p11.23), Cockwell et al. (1996) demonstrated in a severely malformed spontaneously aborted 11-week fetus one of the adjacent-1 conceptions, the dup(8p)/ del(18p) state. This chromosomal constitution caused a double-segment imbalance, with a trisomy for 8p21.3–pter, and a monosomy for 18p11.23–pter, giving a combined 1.2% HAL imbalance (0.8% for trisomy, 0.4% for monosomy). The countertype dup(18p)/del(8p) karyotype had produced a child with an abnormal phenotype in this family. Atypically, this viable form had more HAL monosomy than trisomy.

Adjacent-1 Segregation with Subtelomeric Double-Segment Exchange

A major cytogenetic development in recent times has been subtelomeric FISH. With this methodology, translocations involving segments in the (generally) gene-rich regions just below the telomeres, which are too small to be seen even on high-resolution banding, can now be identified (Slavotinek et al., 1999; Knight and Flint, 2000; Rossi et al., 2001). Consider the family reported in Vogels et al. (2000), the pedigree of which is presented in Figure 4-8a. Three individuals (two cousins, and the nephew of one of them) presented a similar picture of severe mental defect and obnoxious behavior, with facial dysmorphism and certain malformations; another with similar facial dysmorphism died of a heart defect as a neonate. This picture demanded close attention to the possibility of a chromosomal explanation, but high-resolution karyotyping was normal in all four, with one of the cousins having been tested on several occasions. Finally, FISH with subtelomeric probes provided a clear and unambiguous illumination (Fig. 4-8b). In testing the potential carriers in the family, a 5q subtelomeric probe and an 18q subtelomeric probe hybridized properly to one chromosome 5qter and to one chromosome 18qter; but the other 5qter hybridized with the 18q probe, and the other 18qter carried the 5q probe. Thus a double-segment reciprocal exchange between 5q and 18q was demonstrated: 46,t(5;18) (subqter;subqter).5 The affected individuals were trisomic for a very small segment of subterminal 5q and monosomic for a similarly small segment of 18qter, and the clinical picture suggested features of both the dup(5q) and the del(18q) syndromes. A number of similar families are now being reported, and many genetics services will be reviewing their “chro-mosomal-seeming families” in which G-band-ing, even of the highest resolution, failed to reveal any abnormality (Warburton et al., 2000). Not every such review will be fruitful, and along with the disconcerting observation of subtelomeric deletions in the normal population, some reservation about the broader significance of these findings is in order, at least until further experience has been gained (Joyce et al., 2001) (p. 244).

Figure 4-8. The subtelomeric translocation. (a) Pedigree of family. Filled symbol, abnormal individual with subtelomeric aneusomy; half-filled symbol, balanced carrier; dot, miscarriage. (From Vogels et al. [2000], courtesy J.-P. Fryns.) (b) FISH demonstration of a reciprocal translocation t(5;18), undetectable on high-resolution G-banding (middle), but obvious using probes for the 5q subtelomeric region (left) and the 18q subtelomeric region (right). (c) Segregation patterns at meiosis I in the heterozygote, making the assumption, which may or may not be correct, that the homologs align as independent bivalents, rather than as a quadrivalent. Segregation would thus be 1:1 for each pair of bivalents. If the pairs of bivalents happen to be disposed on either side of the equatorial plate as in upper section, normal and balanced daughter cells are produced. If they happen to be disposed as in the lower section, unbalanced cells result, either with del(5q)/dup(18q) (thin arrows), or, as observed in this family, with dup(5q)/ del(18q) (thick arrows), these outcomes thus resembling adjacent-1 segregation. The equatorial plate, in transverse section, is indicated by the dotted line.

FISH can cast light on a formerly obscure circumstance. Estop et al. (1995a) studied a child with an abnormal 17p, which they thought at first to be a de novo deletion. But on subsequent family study with FISH, a balanced double-segment t(9;17)(q34.3;p13.3) was discovered. Holinski-Feder et al. (2000) examined a kindred segregating mental retardation, in which the picture seemed to be that of a Mendelian condition showing anticipation, since the affected individuals were seen only in the two most recent generations. High-res-olution karyotyping and multicolor-FISH had been negative. A linkage study pointed to 16p. Directed FISH with subtelomeric probes finally revealed the subtle translocation t(3;16)(q29;p13.3), and the “anticipation” was seen actually to be due to unaffected carriers in earlier generations having affected children in following generations.

If a quadrivalent were to form in a subtelomeric translocation, in almost every case it would only be the adjacent-1 gametes that would be viable, besides the normal and balanced forms. But with such tiny chromosomal segments involved, it may well be that the homologs, the normal and the derivative, would simply pair up as in a normal bivalent. In that case, the expected segregations at meiosis would be normal:balanced:(dup/del):(del/dup) in the ratio 1:1:1:1.

Adjacent-2 Segregation

This is an uncommonly observed mode of segregation, typically limited to translocations in which the two participating chromosomes each has a short arm of small genetic content, and small enough that the whole short arm can be viable in the trisomic state. In fact, most cases involve an exchange between chromosome 9 and an acrocentric, or between two acrocentrics (Duckett and Roberts, 1981; Stene and Stengel-Rutkowski, 1988; Mangelschots et al., 1992; Cotton et al., 1993). The breakpoints characteristically occur in the upper long arm of one chromosome and immediately below the centromere in the long arm of the other (an acrocentric). Thus, the centric segments are small.

The t(9;21)(q12;q11) illustrated in Figure 4-9a exemplifies the adjacent-2 scenario. At meiosis I, the form of the quadrivalent would be as drawn in Figure 4-9b. The least imbalanced, least monosomic gamete from 2:2 malsegregation is that receiving chromosome 9 and the der(9) (heavy arrows). The conceptus will have, in consequence, a duplication of 9p (and a small amount of 9q heterochromatin) and a deletion of 21p (and a minuscule amount of subcentromeric 21q). Although comprising a substantial piece of chromatin (1.8% of HAL), 9p is qualitatively small in the trisomic state. Monosomy for 21p is without effect, and the 21q loss makes little if any contribution, and thus the picture is practically that of a pure 9p trisomy. This is a known viable aneuploidy. The countertype gamete with the der(21) causes monosomy 9p and is not viable. A very similar circumstance applies with the t(4;13)(q21;q21) described in Velagaleti et al. (2001); for this example, the open and crosshatched chromosomes in the cartoon karyotype (Fig. 4-9) could be regarded as chromosomes 4 and 13, respectively. The index case in this family was trisomic for all of 4p and the small segment 4cen-q21 (and monosomic for the tiny segment 13pter–13q21), having the karyotype 46,XY,der(4),13.

Figure 4-9. Adjacent-2 segregation. (a) Mother (upper) has a reciprocal translocation t(9;21)(q12;q11), and her child (lower) has the adjacent-2 karyotype 46,+der(9)t(9;21)(q12;q11) (case of C. M. Morris and P. H. Fitzgerald). (b) The presumed pachytene configuration during gametogenesis in the heterozygote (no. 9 chromatin is open, no. 21 chromatin is cross-hatched). Arrows indicate movements of chromosomes to daughter cells in adjacent-2 segregation; heavy arrows show the viable combination, as observed in this family.

The del(22)(q11) syndrome, so well known otherwise as being due to a simple deletion (see p. 288), can also arise from a familial translo-cation, and this provides an example of a dou-ble-segment imbalance with adjacent-2 segregation. Imagine a t(9;22)(q12;q11.21) with the 22q breakpoint just below the DiGeorge critical region (DGCR). If the cross-hatched chromosome in Figure 4-9b were a no. 22, then the der(9) would lack the DGCR. A 46,der(9),22 child from adjacent-2 segregation (the heavy arrows) would have the 22q deletion syndrome, superadded upon a 9p trisomy. Pivnick et al. (1990) and El-Fouly et al. (1991) describe children in whom these sepa-rate-and-together dup(9p) and del(22q) phenotypes could be distinguished.

A double-segment exchange with both adja-cent-2 segregants observed, and reflecting a parent-of-origin effect, is shown in the family reported by Abeliovich et al. (1996). The family translocation, carried by the father, was due to breakpoints in the long arms of chromosomes 15 and 21, t(15;21)(q15;q22.1). Both centric segments, 15pter–15q15 and 21pter–21q22.1, are of quite substantial size. One child had the karyotype 46,15,der(21), with a proximal partial 15q monosomy and a proximal partial 21q trisomy. The phenotype was predominantly that of the Prader-Willi syndrome (PWS), reflecting the lack of a paternally contributed PWS critical region, residing in 15q11–q13. There was no clearly apparent contribution from the partial trisomy for 21pter–21q22.1. The other child, with a dup(15q)/del(21q) combination, 46,der(15), 21, displayed a combination of features due to monosomy 21pter–21q22.1 and trisomy 15pter–15q15.

The reason so few examples of adjacent-2 segregants are seen is that most convey a lethal imbalance during early embryogenesis. Naturally, if the window of observation were to be shifted to this period of development, more cases would reveal themselves. An example is shown in Figure 4-10, this being the karyotype from the products of conception obtained at miscarriage in the first trimester from a woman who was herself a translocation carrier, 46,XX,t(13;16)(q12.3;q13). The karyotype of the cultured products, 46,XX,13,der(16), displays an overall HAL imbalance of 2.7%. Two previous miscarriages to this couple might also have had this karyotype. Earlier in the piece, at the 3-day embryo stage, selection pressures have not yet come to bear, and thus a preponderance of embryos at preimplantation diagnosis with an adjacent-2 imbalance, as seen from a 46,XY,t(10;18)(q24.1;p11.2) carrier reported in Munné et al. (2000b), is perhaps not too remarkable a finding.

3:1 Segregation with Tertiary Trisomy

Tertiary trisomy is fairly uncommon—or to be precise, fairly uncommonly seen in a term pregnancy—and may arise only when one of the derivatives is of small content. It exists in the abnormal individual as a supernumerary derivative chromosome, with the karyotype 47,+der. The centric segment will necessarily contain the whole short arm of the derivative chromosome, and it will necessarily be of a chromosome having a small short arm. Almost always, complete long arms (and in fact most complete short arms) contain too much material to allow viability in a supernumerary derivative chromosome. There is, as noted below, a significant maternal age effect in 3:1 imbalance.

Figure 4-10. Adjacent-2 segregation, with an imbalance lethal in early pregnancy. The mother (upper) has the karyotype 46,XX,t(13;16) (q12.3;q13). Tissue from the products of conception of a spontaneous first-trimester abortion was cultured, and the chromosomal complement from these cells showed the karyotype 46,XX,-13,+der(16). There is monosomy of proximal 13q for a segment of HAL 0.6%, and partial trisomy 16 for a segment of HAL 2.1%. (Case of M. D. Pertile.)

Curiously enough, in the most common (by far) human reciprocal translocation, practically all abnormal offspring of the heterozygote have a tertiary trisomy, due to 3:1 meiosis I malsegregation (Shaikh et al., 1999). This is the t(11;22)(q23;q11)6 (Fig. 4-11a). The quadrivalent of this 11;22 translocation would have the form outlined in Figure 4-11b. The content of the smallest chromosome, the der(22), is small (respecting the requirement for the derivative is to have a small short arm, chromosome 22 easily qualifies), and its major genetic composition is accounted for by the distal 11q segment. The presence of this 47th chromosome does not impose a lethal distortion on intrauterine development, and a pregnancy could continue through to the birth of a child who would have trisomy for the segment 11q23–qter (and for the very small segment 22pter–q11), with the karyotype 47,+der(22),t(11;22)(q23;q11).7 The clinical picture has been recorded often enough that it may be accorded syndromal status.

This t(11;22) is the spectacular exception to the rule that, in different families, translocations arise at different sites. The great majority of families have a private translocation, and many may represent the first and only case in the whole of human evolution. Apparently, few hot spots for rearrangement exist; equally apparently, 11q23 and 22q11 are remarkably hot hot-spots, reflecting certain qualities of the DNA sequences in them, specifically, AT-rich palindromic repeats (McDermid and Morrow, 2002). Kurahashi and Emanuel (2001) studied normal volunteers, and, being able to test very large numbers of sperm, they were able to show that de novo t(11;22)(q23;q11) translocations must be being generated from time to time.

Figure 4-11. Tertiary trisomy. (a) The common t(11;22)(q23;q11) in the heterozygous state (upper), and in the typical unbalanced state (lower). (b) The presumed pachytene configuration during gametogenesis in the heterozygote (no. 11 chromatin is open, no. 22 chromatin is cross-hatched). Arrows indicate movements of chromosomes to daughter cells in a 3:1 tertiary segregation; heavy arrows show the viable trisomic combination.

The male t(11;22) heterozygote, at least, does produce other types of unbalanced gamete, as shown on sperm chromosome study (Table 4-1), but none of these is ever viable.8 Using FISH, Estop et al. (1999) showed in one subject 3:1 segregation in 40% of sperm (rather more than the 15% for (11;22)(q23;q11) noted in Table 4-1), and they note a predisposition of this translocation to segregate according to this mode. Going one stage further, Van Assche et al. (1999)undertook PGD for two t(11;22) carrier couples. From the couple with the male being the carrier, only 9 out of the 35 embryos were normal. Many of the abnormal embryos were mosaic, some “chaotically” so. Three embryos were obtained from the female carrier of the other couple (the male being 47,XXY), but all were unbalanced. By contrast, the majority of the embryos from the male carrier reported in Mackie Ogilvie and Scriven (2002) were due to alternate segregation (and see Table 24-2).

Note the point that probands in whom an extra structurally abnormal chromosome (ESAC) is discovered are often found, on parental study, to have a derivative chromosome reflecting a tertiary trisomy (Stamberg and Thomas, 1986). Braddock et al. (2000) describe a family in which an ESAC due to 3:1 malsegregation had initially escaped recognition as such. A child with “atypical Down syndrome” had been karyotyped as trisomy 21. On attending a Down syndrome clinic at age 9 years, the clinical picture raised doubt and his chromosomes were restudied. He turned out to have a tertiary trisomy for a der(21) which comprised much of chromosome 21 and a small part of distal 5p. His mother and several other relatives carried a t(5;21)(p15.1;q22.1), and a similarly abnormal aunt had the same tertiary trisomy, 47,der(21). This story has lessons for both cytogeneticists and genetic counselors.

3:1 Segregation with Tertiary Monosomy

If one derivative is very small, and the amount of material that is missing is “monosomically small,” the countertype 3:1 22-chromosome gamete may lead to a viable conceptus. Consider the 12;13 translocation t(12;13)(p13.32;q12.11) shown in Figure 4-12a. The large derivative chromosome is not far from being a composite of the two complete chromosomes. It is missing only subterminal 12p and pericentromeric no. 13. This is a “small” loss, and thus the 45,der(12) conceptus is viable (Fig. 4-12b). Any initially 45-count karyotype obliges consideration that there may, in fact, be a tertiary monosomy. For example, Courtens et al. (1994) describe an infant who died at birth with, at first sight cytogenetically, monosomy 21 (45,21). But with FISH and molecular studies, a 45,der(1) from a maternal 1;21 translocation was discovered.

Figure 4-12. Tertiary monosomy. (a) Mother (upper) has a reciprocal translocation between nos. 12 and 13, 46,t(12;13)(p13.32;q12.11). Two children (lower) inherited the derivative 12, but no normal chromosome 12 or 13 from the mother, and have the karyotype 45,der(12). They are thus monosomic for the tip of 12p and pericentromeric 13 (and only a mildly abnormal phenotype). Chorionic villus sampling in a subsequent pregnancy gave a 46,XX result; an elder sister was a balanced carrier. (Case of M. D. Pertile.) (b) The presumed pachytene configuration during gametogenesis in the heterozygote (no. 12 chromatin is open, no. 13 chromatin is cross-hatched). Arrows indicate movements of chromosomes to daughter cells in a 3:1 tertiary segregation; heavy arrowshows the monosomic complement. Alternatively, the three large chromosomes might form a trivalent, and the tiny der(13), being unattached, might segregate at random.

Sometimes the two phenotypes of the two contributing monosomies can be separately discerned. Thus, Reddy et al. (1996) describe children with a combined Wolf-Hirschhorn (WHS) and Di George (DGS) phenotype, having the karyotype 45,der(4)t(4;22)(p16.3; q11.2)mat. The large derivative chromosome comprised almost all of 4 and almost all of 22q, but it lacked the WHS and DGS critical segments. Similarly, Wenger et al. (1997) report a mother with a t(8;15)(p23.3;q13) whose child had the karyotype 45,der(8) and presented a phenotype with features of both 8p syndrome and Angelman syndrome, the latter due to loss of the maternally originating segment 15q11–q13.

Component phenotypes may also be discernible where the monosomy is very small and only a few contiguous loci are lost. An interesting historical example, in that it provided a key observation toward the discovery of the TSC2 locus, is that of a child with 45,der(16), who had monosomy for the segment 16p13–pter and had both tuberous sclerosis and polycystic kidney disease, due to loss and disruption, respectively, of the adjacent TSC2 and PKD1 loci. The heterozygous 46,t(16;22) family members had polycystic kidney disease, from the disruption of PKD1 (European Polycystic Kidney Disease Consortium, 1994).

3:1 Segregation with Interchange Trisomy

This mode of segregation can only produce a liveborn child when a trisomically viable chromosome (i.e., 13, 18, or 21) participates in the translocation (Fig. 4-13a). This chromosome (13, 18, or 21) accompanies the two translocation (interchange) elements of the quadrivalent to one daughter cell (Fig. 4-13b). Interchange trisomy 21 is rare, and interchange trisomies 13 and 18 extremely rare (Fryns et al., 1986; Stene and Stengel-Rutkowski, 1988; Daniel et al., 1989; Smith et al., 1989; Kotwaliwale et al., 1991; Teshima et al., 1992; Koskinen et al., 1993). Interchange trisomy 7 has been identified at cytogenetic analysis of the products of conception from a 6-week miscarriage, the mother carrying a t(1;7)(p22;q21.2) (Cockwell et al., 1996).

Figure 4-13. Interchange trisomy. (a) Mother (upper) has a reciprocal translocation between nos. 12 and 21; her child (lower) inherited the maternal translocation chromosomes and a “free” chromosome 21. The breakpoints are 12q13.1 and 21p13; an apparent gap, comprising satellite stalk, can be discerned between the centromere of the der(21) and its 12q component. (Case of R. Oertel.) (b) The presumed pachytene configuration during gametogenesis in the heterozygote (no. 12 chromatin is open, no. 21 chromatin is crosshatched). Arrows indicate movements of chromosomes to daughter cells in 3:1 interchange segregation; heavy arrows show the trisomic combination.

Theoretically, uniparental disomy can be a consequence of interchange trisomy, if one of the “trisomic” chromosomes is subsequently lost postzygotically, and if this chromosome came from the non-carrier parent. If this chromosome is one that is subject to imprinting according to parent of origin, phenotypic abnormality will be the consequence, notwithstanding the apparently balanced karyotype, the same as the parent's. Thus, for example, a 46,XY,t(8;15) father, such as the man with the t(8;15)(p22;q21) listed in Table 4-1 and 4% of whose sperm were 3:1, could have a 46,t(8;15) child with Angelman syndrome. A comparable case is shown in Figure 7-5 of a man with a 45,t(8;15) telomeric fusion chromosome who had a 45,t(8;15) child with UPD15 Angelman syndrome (Smith et al., 1994). Actual examples of this type of mechanism are extremely rare (Dupont et al., 2002).

3:1 Segregation with Interchange Monosomy

Autosomal monosomy is typically associated with very early pregnancy failure (see Nonimplantation and Occult Abortion, p. 343). Only with PGD does a practical relevance possibly emerge, since there has not yet been the chance for post-zygotic selection pressure to have operated. In a PGD case reported in Conn et al. (1999), the woman being a t(6;21) heterozygote, a transferred embryo that implanted only transiently may have had an interchange monosomy 6. Actually, it appears that many monosomiescause embryo arrest before the blastocyst stage, and this is a factor encouraging IVF laboratories to develop the technique of blastocyst culture, since using this procedure would allow enough time for these monosomic conceptions to have aborted (p. 384).

4:0 Segregation

A total nondisjunction of the quadrivalent complex is rare indeed: only four examples are listed in Tables 4-1 and 4-2. In one FISH sperm study of a man with 46,XY,t(1;10) (p22.1;q22.3), as many as 4036 sperm were analyzable, and none had all four quadrivalent chromosomes (Van Hummelen et al., 1997). If 4:0 segregation should happen, very earlylethality would be the likely consequence. Out of interest, the reader may care to note how a double trisomy of 18 plus 21, based on the combination (15) in Figure 4-4, and potentially associated with some in utero survival (Reddy, 1997), could come from the t(18;21) shown in Figure 4-15.

More Than One Unbalanced Segregant Type

Sometimes a reciprocal translocation has characteristics associated with more than one type of malsegregation, so each type may be seen in the family (Niazi et al., 1978; Abeliovich et al., 1982). Consider the 11;18 translocation t(11;18)(p15;q11) shown in Figure 4-14. First, the translocated segments are small: 18q is known to be viable in the trisomic state, and the tip of 11p contributes a minimal/nil imbalance (i.e., this is regarded as a single-segment imbalance). Thus, one of the adjacent-1 segregants is presumed to be viable. Second, two component chromosomes of the pachytene configuration, the der(18) and no. 18, are of small overall genetic content. Thus, 3:1 segregation with either tertiary trisomy or interchange trisomy is possible. In the event, the two unbalanced karyotypes in this family reflected adjacent-1 and 3:1 tertiary trisomy segregation. The t(9;21) discussed above as an example of adjacent-2 segregation could also, in theory, produce a second viable complement, interchange trisomy 21.

Rather more spectacular is the translocation illustrated in Figure 4-15. A mother had the karyotype 46,XX,t(18;21)(q22.1;q11.2): these breakpoints are toward the end of 18q and immediately below the centromere in 21q. She had a stillborn child with tertiary monosomy, a miscarriage with adjacent-1 malsegregation (and two other unkaryotyped miscarriages), and a surviving child with tertiary trisomy. These three karyotyped pregnancy outcomes were, respectively, 45,der(18), 46,der(18), and 47,+der(18).

An uncle said to have had Down syndrome (DS) may have had the 46,der(18) karyotype (the der(18) includes the segment of 21 that contributes substantially to the DS phenotype), or possibly interchange trisomy with 47,+21,t(18;21). Some of the other possible imbalanced segregants could theoretically be viable, and the reader may wish to determine which ones these would be. These segregants could be viable because many of these combinations have a genetically small imbalance. All partial trisomies and some partial monosomies for segments of chromosomes 18 and 21 can be viable as a single imbalance, and when two different imbalances occur in combination, e.g., partial trisomy 21 plus partial monosomy 18, a pregnancy may still be capable of proceeding substantially along its course.

Figure 4-14. More than one viable segregant form. (a) Pedigree. Filled symbols, unbalanced karyotype, as shown; half-filled symbols, heterozygote. (b) Mother and one daughter have a reciprocal translocation of chromosomes 11 and 18, t(11;18)(p15;q11) (upper). Each had one unbalanced offspring, one having 47,der(18) due to 3:1 tertiary trisomy (middle), and the other 46,der(11) from adjacent-1 segregation (lower). The former had a complete trisomy 18p, and the latter a partial 18q trisomy. (Case of C. Ho and I. Teshima.) (c) The presumed pachytene configuration during gametogenesis in the heterozygote (no. 11 chromatin is open, no. 18 chromatin is cross-hatched). Heavy arrows indicate one adjacent-1 segregant movement of chromosomes, and light arrows indicate movements of chromosomes to daughter cells in a 3:1 tertiary trisomy segregation, each of which occurred in this family. (From Gardner et al., 1978.)

No Unbalanced Mode Possible

Finally, for the translocation in which the quadrivalent is characterized by long translocated and long centric segments, no mode of segregation could produce a viable unbalanced outcome. We emphasize the point that many reciprocal translocations (including whole-arm translocations) are in this category. Consider the family depicted in Figure 4-16, in which a 4;6 translocation t(4;6)(q25;p23) was discovered by chance at amniocentesis. The quadrivalent would have the form depicted in Figure 4-5d. It possesses none of the criteria that would allow a viable imbalance to result, by whatever mode of segregation. The translocated segments are both large (leading to double-segment imbalance); the centric segments are very large; and the content of all four chromosomes is large. Miscarriage is as far as any unbalanced conceptus could ever get. The large kindred of Madan and Kleinhout (1987) graphically illustrates this circumstance: 11 carriers had two or more miscarriages and numerous normal children, but none had an abnormal child. In some such translocations identified fortuitously, for example, at amniocentesis for maternal age, there may be little or no history of apparent reproductive difficulty.

Figure 4-15. Several viable unbalanced forms. The karyotype is illustrated (top) of a mother carrying the translocation t(18;21)(q22.1;q11.2). She had a miscarriage due to adjacent-1 segregation, an abnormal child with a tertiary trisomy, and a stillborn child with a tertiary monosomy, as depicted in the cartoon karyotypes. An uncle with Down syndrome may have had the same adjacent-1 karyotype as in the second row, or possibly interchange trisomy 21, as depicted in the bottom row. (Case of M. D. Pertile.)

Figure 4-16. No unbalanced product viable. (a) Pedigree of a kindred in which mother and daughter have had multiple miscarriages, each having (b) the karyotype 46,XX,t(4;6)(q25;p23) (case of A. J. Watt). The presumed pachytene configuration during gametogenesis in the heterozygote would be as in Figure 4-5d (no. 4 chromatin is open, no. 6 chromatin is crosshatched) and, with large centric and translocated segments, the translocation has none of the features that enable viability of any unbalanced segregant combination.

Meiosis II Nondisjunction. The great majority of segregant forms will have been determined at meiosis I. Meiosis II is not to be completely overlooked, however. A balanced complement may have been transmitted at meiosis I, but a nondisjunction at the following second meiotic division could then produce a gamete with an extra copy, or no copy, of one or other of the derivative chromosomes. In consequence, the conception would have either a partial trisomy of the component parts of the additional derivative chromosome or a partial monosomy. Illustrating the former possibility, Masuno et al. (1991) report a mother with a 46,XX,t(5;13)(p15;q21), whose fetus had the same translocation plus an additional copy of the der(13), that is, 47,XY,t(5;13)(p15;q21),+der(13)mat. The similar scenario with respect to the common t(11;22) is noted in note 7. This type of secondary nondisjunction is very rarely observed.

Meiotic Drive. As well as the effect of in utero survivability discussed above, the nature of the quadrivalent may, of itself, influence segregation. The propensity for a particular segregation outcome may reflect a particular geometry of the quadrivalent and what sort of ring or chain it forms. Quadrivalents that have translocation chromosomes with short translocated segments have the quality of being more likely to generate adjacent-1 gametes, while those with short centric segments predispose to the formation of adjacent-2 gametes (Faraut et al., 2000). This predisposition to form particular classes of segregant gamete may be considered a form of meiotic drive.

As we have had cause to comment more than once, each translocation is entitled to its individuality, and need not necessarily follow the rules set out above. Faraut et al. (2000) identified a few translocations that should have produced sperm with certain expected proportions of adjacent-1 and adjacent-2, but which did not. We have seen a remarkable family in which, over some 10 years of marriage, the woman had innumerable very early miscarriages, about eight at 12–14 weeks, one 16week miscarriage, and one phenotypically normal son. The husband (and the son) had the translocation 46,XY,t(12;20)(q15;p13). Perhaps, the quadrivalent was configured in such a way that alternate segregation was very difficult to achieve, and so almost all sperm had an unbalanced complement. De Perdigo et al. (1991) report a possibly similar case, in which they propose that heterosynapsis in the quadrivalent permitted spermatogenesis to proceed, but at the cost of producing many unbalanced gametes. In a family reported by Groen et al. (1998) with a mother having the karyotype 46,XX,t(5;6)(q35.2;q27), seven sequential retarded siblings of hers are presumed to have had a dup(5)/del(6) karyotype, and only the two eldest and the youngest were phenotypically normal. Observations from PGD (Chapter 24) are further illustrating the point that translocation carriers with very poor reproductive histories may indeed reflect a very high rate of meiotic malsegregation. The patient in Conn et al. (1999) mentioned above, she having the karyotype 46,XX,t(6;21)(q13;q22.3), had had four miscarriages and one child with interchange trisomy 21. She came to PGD, and not one of two oocytes and nine embryos was chromosomally normal (mostly 3:1, some adjacent segregations).

Failure to Form Quadrivalent. Where very small segments are involved, the imperative may lack for the coming together of the four chromosomes with segments in common. This might pertain, for example, to a translocation such as the t(14;15)(q12;q12) in Burke et al. (1996), in which the derivative chromosomes each comprise almost an entire chromosome 14 and chromosome 15, respectively. The 14 and der(14), and the 15 and der(15) might simply synapse as bivalent pairs. The same may very well apply to the general case of the subtelomeric translocation, as discussed above. If that were indeed so, then a segregation ratio of 1:1:1:1 would presumably operate for normal, balanced, and the two imbalanced out-comes—clearly, a high-risk circumstance. Different grounds for the nonformation of a quadrivalent may exist if one chromosome is a very small one. While the three other chromosomes could have come together as a trivalent, the fourth very small one might fail to be captured by the meiotic mechanism. That being so, it could then segregate at random. This could imply a high risk, and might be the reason, for example, that the t(12;13) carrier mother in Figure 4-12 had two out of her four children with a tertiary monosomy. But this is speculative. Detaching of the small derivative from the quadrivalent is an alternative possibility, as discussed in the next section.

Parental Origin and Parental Age Effect. There are more women who have been mothers (whether the children are normal or not) than there are men who have been fathers in translocation families. In their review of 1597 children in 1271 translocation families, Faraut et al. (2000) found the mother to be the carrier parent in 61% of the adjacent-1 children, in 70% of the adjacent-2 children, and in as many as 92% of the unbalanced offspring from 3:1 segregations. This 3:1 association may re-flect an actual maternal predisposition. With advancing maternal age, and after some decades of being held in meiosis I prophase, the small supernumerary chromosome may be increasingly likely to detach from the quadrivalent and then to migrate at random to one or other daughter cell, when meiosis reactivates in that particular menstrual cycle. No maternal age effect applies, however, to adjacent-1 or adjacent-2 offspring. Here, the maternal excess may more accurately be termed a paternal deficiency, due to reduction in fertility of the male heterozygote (see below). No paternal age effect is discernible in any segregation mode.

The Practical Problem of De Novo Apparently Balanced Translocations

A not uncommon problem encountered in the genetic clinic is that of the de novo apparently balanced translocation, which has been discovered in the course of investigation of a child with a nonspecific picture of cognitive compromise and sometimes also some dysmorphic signs. Is the translocation causative or simply coincidental? It is not wonderfully helpful (even if it would be true) to say to the parents that if their child should in the future have children and grandchildren with the same translocation who turn out to have, or not to have, the same abnormal phenotype, then one could know that the translocation was, or was not, the cause. One would rather be wise before the event. Families like those reported in Hussain et al. (2000) offer useful illustration: in this example, an apparently balanced translocation was cosegregating with a phenotype of non-syndromic mental retardation. Presumably this translocation, a t(1;17)(p36.3;p11.2), had been de novo at some prior point, possibly with the 65-year-old grandmother of their index case. In this family, there were seven children and grandchildren to bear witness to the harmful role of the translocation. Thus, the point is underlined: at least some apparently balanced translocations are indeed the cause of the nonspecific clinical picture with which they are associated. But the fact remains that, in most cases, when these translocations have arisen de novo, it will not be possible to state with certainty that the link with phenotypic abnormality was causative, merely that, quite probably, this could have been the case. These translocations are further mentioned below (Translocations with Breakpoints at Vital Loci) and on p. 274.

Infertility

Infrequently, the process of gamete formation in the male translocation heterozygote is disturbed to the extent that gametogenic arrest results. In the analysis of reproductive outcomes in the translocation families of Faraut et al. (2000), looking at prenatal diagnoses to avoid bias, 61% of all fetuses came from a carrier mother, versus only 39% from a carrier father; this ratio presumably reflects male infertility associated with the carrier state. This infertility is generally not something that is predictable from the nature of the translocation, and indeed the same translocation may compromise fertility in only some men in the family. Presumably there is, in addition, an effect of the genetic background otherwise (Rumpler, 2001). The detrimental process is considered to be a consequence of failure of pairing (asynapsis or heterosynapsis) of homologous elements in the translocation chromosomes during meiosis I, which promotes association of the quadrivalent with the X-Y bivalent, also known as the sex chromosome vesicle (Paoloni-Giacobino et al., 2000b). The more frequently this association occurs, the more marked the effect on sperm count. There may also be a high rate of mosaicism in abnormal embryos from translocation carriers (Blennow et al., 2001); this phenomenon may be an additional factor contributing, at a very early post-conception stage, to subfertility (in either parental sex).

The sex difference in susceptibility is striking in the family of Paoloni-Giacobino et al. (2000b). The mother was a t(6;21)(p21.1;p13) heterozygote, and she had eight children, four sons and four daughters (and two miscarriages). The four sons, each one 46,XY,t(6;21), were all married, one three times, and none had any children. Each had severe oligospermia or oligoasthenoteratospermia, two having testicular biopsies that manifested spermatogenic arrest at the spermatocyte I stage (meiosis I prophase), with extensive asynapsis of several chromosomes. Two sisters were 46,XX,t(6;21), and the one who was married had had two children (and two miscarriages). Oogenesis may not, however, be entirely immune to the translocation obstacle (Speed, 1988; Mittwoch, 1992). Tupler et al. (1994) report two women, one with primary and the other with secondary amenorrhea, who had a balanced reciprocal translocation. Ovarian biopsy in the former, whose translocation was a de novo one, showed absence of the follicular structures in the cortex. Such cases may exemplify a rare translocation effect in the female; equally, given the frequency in the population of the heterozygote, it does remain possible that the link is coincidental rather than causal.

Assisted Reproduction

Assisted conception may enable infertile men with a translocation to become fathers. But of course the translocation will, in any event, convey a genetic risk. Meschede et al. (1997) report a man with a t(1;9)(q44;p11.2) having intracytoplasmic sperm injection (ICSI), and two embryos were successfully transferred. At prenatal diagnosis, one twin had an adjacent-1 imbalance conferring a 9p trisomy, and the other was 46,XX, and the parents chose selective abortion. Belin et al. (1999) describe a triplet pregnancy achieved via ICSI, the father being a t(20;22) heterozygote. Two normal babies were born (one karyotypically normal, one with the translocation), but the third, with a dup(20p)/del(22q) imbalance, was severely malformed and died in the neonatal period. Meschede et al. (1997) suggest that the population of fertilizing sperm from carriers may differ between ICSI and natural conceptions, and that the risk criteria generally set forth in this chapter should be used with caution in couples needing ICSI. Could a chromosomally unbalanced but morphologically normal sperm have a better chance of being selected at ICSI than of succeeding in the open competition that obtains in vivo? We have yet to learn.

Rare Complexities

Translocations with Breakpoints at Vital Loci

The great majority of breakpoints in familial translocations between autosomes are apparently sited at points within the genome where they have no effect on its smooth running. Thus, the balanced carriers are phenotypically normal. Rarely, the act of breakage and reunion might compromise a gene or genes. Almost always, this is observed in the de novo, not the familial, case. Kumar et al. (1998) list the following possible explanations for the basis of an abnormality of phenotype: cryptic deletion or duplication; disruption of a gene; position effect; and uniparental disomy. They studied abnormal individuals with a de novo rearrangement, including one with 46,t(2;6)(p22.2;q23.1). With a molecular and FISH approach, they were able to identify a deletion of about 4 Mb on the der(6) chromosome. Presumably, this cryptic deletion was the cause of the clinical defects.

A detailed example of the gene disruption circumstance is provided in Kurahashi et al. (1998). A child with lissencephaly (a severe structural brain abnormality) had a de novo t(8;17)(p11.2;p13.3). The p13.3 breakpoint on the chromosome 17 was sited within intron 1 of the LIS gene, with the gene being split between the two derivative chromosomes: its 5 part on the der(8), and the rest of it on the der(17). The gene could not, in consequence, function. FISH can provide direct visual proof of a locus having been disrupted. For example, von Dadelszen et al. (2000) report an infant with features of Williams syndrome and having a de novo t(6;7)(q27;q11.23), the Williams critical region being at 7q11.23. They showed three separate hybridizations using a Williams syndrome probe: one on the normal chromosome 7, and one each on the der(6) and the der(7), at the sites of breakpoint. The latter two hybridizations represented the separated “halves” of the Williams syndrome region. The knowledge gained from 15q13 translocations that disrupt the SNRPN gene, in elucidating molecular mechanisms in Prader-Willi syndrome, is noted in Chapter 20.

As for the position-effect scenario, a number of examples can be cited, in which a syndrome is caused by a nearby intact gene failing to function: a chromosome 17q25.1 translocation whose breakpoint is 50 kb away from the SOX9 locus, producing campomelic dysplasia (Fig. 17-4); a translocation with an 11p13 breakpoint that can put the PAX6 gene into a chromosomal environment that does not permit its normal expression, resulting in abnormal development of the iris (aniridia); Saethre-Chotzen acrocephalosyndactyly with a 7p21 breakpoint, leading to TWIST gene non-functioning; and one genetic form of holoprosencephaly associated with the translocation breakpoint at 7q36, near the HPE3 locus (Wagner et al., 1994; Fantes et al., 1995; Roessler et al., 1997; Rose et al., 1997; Crolla and van Heyningen, 2002). Such translocations can be familial. The variation that may be seen in family members could be thought of as a form of incomplete penetrance (or variable expressivity). A salutary tale comes from the study of a family with an apparently dominantly inherited syndrome of skeletal anomalies, in which previous cytogenetic tests had given normal results (Stalker et al., 2001). Only after the birth of an infant with severe multiple malformations with an unbalanced karyotype was the fact revealed of a balanced t(13;17)(q22.1;q23.3) cosegregating with the phenotype of the syndrome in the family. There is a fair case for considering that a “bone locus” at 17q23.3 had been disrupted or otherwise influenced by the translocation. Stalker et al. rightly comment that a chromosome test is always worth doing in the investigation of an apparently new familial syndrome, earlier reports of normal cytogenetics notwithstanding, especially if the original laboratory material is not available for review. However, the simple possibility always remains that a breakpoint and a disease locus are closely linked, so that the translocation and the disease cosegregate in the family (Hecht and Hecht, 1984). The manic-depressive t(9;11) family noted below may represent such a case.

A recessive gene on one chromosome might be unmasked by a translocation on the other (Alley et al., 1995). Daïkha-Dahmane et al. (1998) describe a mother and abnormal fetus both with a t(6;9)(p12;q12), the fetal abnormality having been found by ultrasound at 17 weeks. Following termination of the pregnancy, the fetal defects were shown to be those of the rare recessive Casamassima-Morton-Nance syndrome of skeletal, anorectal, and urogenital defects. Possibly, a normal allele at one of the translocation breakpoints was disrupted, and the father was coincidentally a heterozygote.

Constitutional translocations might convey a risk for cancer if, for example, a tumor suppressor gene is disabled, or an oncogene is separated from its controlling region. Translocations possibly implying risks for breast cancer, renal cancer, and hematological malignancy are noted in the Genetic Counseling section.

Using Translocations to Track Genes. Translocations associated with Mendelian disease can serve as a very helpful signpost in mapping the gene. The t(16;22) segregating in a family with polycystic kidney disease noted above (under 3:1 Segregation with Tertiary Monosomy) enabled this gene, PKD1, to be isolated; the breakpoint was sited within the gene. Neurofibromatois type 1 and Sotos syndrome are other common genetic conditions that have been mapped this way (Ledbetter et al., 1989; Kurotaki et al., 2002). Less readily tractable is a condition such as manic-depressive disease (bipolar disorder), in which an incompletely penetrant Mendelian gene is a suspected contributory cause. Baysal et al. (1998) studied a family in which five persons in three generations were affected, each of whom carried (as did some unaffected relatives) a t(9;11) (p24;q23.1). A detailed molecular dissection failed to find disruption in any of a number of plausible candidate genes within the breakpoint regions of 9p or 11q; and so the question remains open whether segregation of the translocation with bipolar disorder is causal or coincidental. One of the more astonishing examples, even if the first discovery was via a linkage study, relates to a gene for language—a gene whose function sets our species apart from all others, with the prosaic name FOXP2. An individual with a de novo t(5;7)(q22;q31.2) had a severe expressive and receptive speech deficiency, apparently due to a compromise of both the neuromuscular apparatus of vocalization and the neural substrate of the speech centers of the brain. A disruption was shown in the FOXP2 gene, at the site of the 7q breakpoint (Lai et al., 2001).

When two cases are reported of a Mendelian disorder, both having a translocation with one breakpoint in common, naturally the case is very much stronger than with a single observation. Thus, there is merit in the proposition of McGhee et al. (2000) that the Coffin-Siris syndrome locus must be at 7q32–q34, as they had a patient with a de novo apparently balanced t(7;22)(q32;q11.2), and there was another child on record who had a de novo t(1;7)(q21.3;q34).

A formal review of the usefulness of apparently balanced rearrangements in leading the way to locus discovery is given in Bugge et al. (2000). A group of laboratories in Denmark and southern Sweden identified 216 “diseaseassociated balanced chromosome rearrangements” from a cytogenetic archive of 71,739 cases drawn from a population base of about 6.6 million. They categorized their case mate-rial as follows: disorders with proven candidate regions or actual known genes; disorders with suspected candidate regions/genes; presumed fortuitous rearrangements having no causal relationship to the phenotype; a miscellaneous group; eye disorders; malignancies; obesity; muscular disorders; behavioral disorders; autism; male and female infertility; and mental retardation with or without multiple congenital malformations. This latter group was by far the largest, and included 39 cases sharing a breakpoint with at least one other case in the same group (Table 4-3). Extrapolating their figures to a worldwide database involving some 292 laboratories, the Mendelian Cytogenetics Network, they conclude that around 7500 such rearrangements would be on record, and perhaps half or more of these might offer the potential of gene discovery.

Carrier Couple

Since reciprocal translocation heterozygotes are not uncommon in the population, on rare occasions both members of a couple will, by chance, carry a translocation (Neu et al., 1988b). For example, we have seen a couple who had had several miscarriages, from 5 to 9 weeks gestation. The husband's karyotype was 46,XY,t(7;11)(q22;q23) and the wife's was 46,XX,t(7;22)(p13;q11.2). Presumably, their history of miscarriage reflected at least one parent transmitting an unbalanced gamete with each pregnancy—many unbalanced karyotypes, as the reader can surmise, are possible!

A normal child is possible if each contributes a normal or a balanced gamete to the same conceptus. It should be reasonably likely in a given conception for the two contemporaneous gametes to have arisen from alternate segrega-tion—an educated guess puts the chance at about 20%—although when we saw this family, only miscarriage had occurred. A child of theirs having each parental translocation would qualify as having a complex chromosome rearrangement, and we shall follow that case in Chapter 11 (see Fig. 11-6).

Table 4.3. Breakpoints in De Novo Rearrangements Associated with Mental Retardation With or Without Multiple Congenital Malformations

1q31

4q35

9p24

1q44

5q13

9q34

2q23

6q27

12q15

4p16

7q22

13q34

4q22

7q36

14q32

These breakpoints were observed in more than one patient, from a large Scandinavian database (Bugge et al., 2000). Some of these breakpoints may indicate the sites of disease genes. The 15 breakpoints were seen in 39 patients among 216 with a “disease-associated balanced chromosome rearrangement” discovered in 71,739 cytogenetic analyses.

Mosaicism

Almost all balanced reciprocal translocations are seen in the nonmosaic state. This reflects that either the translocation was inherited from a carrier parent or the rearrangement arose preconceptually, in one or the other gamete. Rarely, a balanced translocation can be generated as a postzygotic event, and the person is a 46,t/46,N mosaic. In a literature review, Leegte et al. (1998) recorded 29 such cases. One of their subjects, for example, was a man who had presented with infertility, and he had the balanced karyotype 46,XY,t(9;15) (q12;p11.2). His mother had this translocation in a minority of cells on peripheral blood analysis, with the karyotype 46,XX,t(9;15) (q12;p11.2)/46,XX; thus, she was a somatic-gonadal mosaic. Wang et al. (1998) report a mother mosaic for an effectively whole-arm translocation, 46,XX,t(10q;16q)/46,XX, who had a child with a presumed uniparental disomy 16 phenotype from postzygotic “correction” of interchange trisomy 16. The abnormal child in Aughton et al. (1993) had a 46,XX,t(7;11) (q36;p11)/46,XX mosaicism, with about one-third of the cells having the apparently balanced t(7;11) on two separate blood samplings over a year apart. The question remained whether the translocation cell line might actually have been unbalanced and the phenotypic abnormality was the consequence of this. It could have been coincidental.

Gametic Complementation

Coincidentally abnormal gametes coming from each parent (Fig. 2-5c) is an extremely rare observation. Park et al. (1998) describe a unique example of this scenario in the context of a parental translocation. A father with a balanced t(3;15)(p25;q11.2) transmitted a sperm from adjacent-1 segregation, the der(15) being a very small chromosome. This would have led to a near-complete monosomy 15 in the conception. But this was corrected by the mother's transmitting an egg with disomy 15, most probably from a meiosis I error. The child had a partial trisomy for the very small segments 3p25–pter/15pter–q11.2 due to the der(15) as a 47th chromosome, but the apparently typical Prader-Willi clinical picture reflected the predominant contribution to his phenotype deriving from the maternal uniparental disomy 15.

Unstable Familial Translocation

Tomkins (1981) documents a family in which a mother with 46,XX,t(11;22)(p11;p12) had one daughter with the same translocation and another daughter with 46,XX,t(11;15) (p11;p12). A very few other similar cases are on record. Typically, the translocation breakpoints are at telomeres, centromeres, or in nucleolar organizing regions. There is some sequence similarity in these regions between different chromosomes, and this may set the stage for these very rare “second translocation” events (see also Jumping Translocation, p. 271).

GENETIC COUNSELING

The counselor may have to deal with these questions:

1. Is there a risk of having an abnormal child?

2. If so, what is the magnitude of the risk?

3. What would be the abnormality, and would the child survive?

4. What if the same translocation that I have is found at prenatal diagnosis?

5. What is the risk for pregnancy loss through abortion? Is pregnancy possible?

6. Is there anything else I should know?

P.88

Does a Risk Exist of Having an Abnormal Child?

If a family is ascertained through a liveborn aneuploid child, that very fact demonstrates viability for that particular aneuploid combination. It could happen again.

If, on the other hand, the family was ascertained by miscarriage or infertility or fortuitously, and there is no known family history of an abnormal child, the picture is less clear. Most likely, no aneuploid combination is viable. Alternatively, a viable imbalance may be possible but has not yet happened; or an imbalance could occasionally be viable but usually it is not, and (so far) has led only to abortion. The approach here is to determine the potentially unbalanced segregant outcomes, according to the favored mode of segregation—adjacent-1, adjacent-2, or 3:1—and check to see if any is on record in a pregnancy that produced an abnormal child. The definitive source of information is Schinzel's catalog (2001). When a single-seg-ment imbalance is a potential outcome in a conceptus from adjacent segregation, and if the potential imbalance comprises an aneuploidy equal to or less than one of these segments on record, viability must be assumed to be possible. If the potential imbalance comprises an aneuploidy greater than any on record, viability is unlikely, especially if the aneuploidy is much greater. The great majority of double-segment imbalances from adjacent segregation due to a translocation ascertained other than by a liveborn aneuploid child would be expected to lead to lethality in utero. Nearly always, a new dou-ble-segment exchange presenting at the clinic will truly be new, and there will be no literature record of exactly the same thing to which the counselor may appeal. Some tertiary trisomies from 3:1 segregation are listed in Schinzel's catalog, but in most instances one has to make an educated guess, erring on the side of caution, as to whether the combination of partial trisomies from a derivative chromosome might in sum be viable.

The Magnitude of Risk

If, in a family, it is judged that a risk to have an abnormal child exists, a broad estimate of the level of risk may be derived from a consideration of four factors: the mode of ascertainment of the family; the predicted type of segregation leading to potentially viable gametes; the sex of the transmitting parent; and the assessed imbalance of potentially viable gametes. Most risk figures fall in a range from 0% to 30%; higher risks are rare. These percentages are expressed in terms of abnormal livebirths as a proportion of all pregnancies, although there are other ways of looking at the risk (see below, Risk at Time of Prenatal Diagnosis, and Table 3-1).

A precise risk estimate needs to be based on the actual cytogenetic imbalance. Different chromosomal segments contain, of course, different genomic information. It is scarcely possible to come up with a unifying format, given that chromatin is not uniform; as Cohen et al. (1994) comment, “it would be hazardous to suggest a simple mathematical relationship between unbalance length and viability.” Some segments in the trisomic state impose a lesser degree of compromise on the process of embryonic development—such as, for example, 18p, and distal 5p. Other segments, although they may be of shorter length, are lethal during early pregnancy and lead to miscarriage. Some translocations can have their own peculiar segregation characteristics, which were a priori quite unpredictable. Nonetheless, it is interesting to attempt a correlation of quantitative chromatin imbalance with risk to have a liveborn affected child. Daniel et al. (1979), Cans et al. (1993), and Cohen et al. (1994) have compared the haploid autosomal length (HAL) with viability in translocation families. Most (96%) viable imbalances comprise up to 2% monosomy and up to 4% trisomy, with combinations of monosomy/trisomy viable only when the additive effect of x% monosomy plus y% trisomy falls within a triangular area defined by joining the 2% and 4% points on the x and y axes of a graph (Fig. 4-17). A few (4%) fall outside of this area, and these cases define the boundaries of a “surface of viable unbalances,” reflecting the effects of qualitative differences in different segments of chromatin.

For routine practice in the genetic clinic, we suggest starting off with the unvarnished empiric data for individual chromosome segments collected by Stengel-Rutkowski and colleagues, as set out in their invaluable monograph (Stengel-Rutkowski et al., 1988) and discussed in a review and further illustrated in practice (Stene and Stengel-Rutkowski, 1988; Midro et al., 1992). The figures set out in Tables 4-4 to 4-6 are summarized from their monograph; a better sense of how each figure has been derived is gained by consulting the original document. It will generally give a false sense of precision to use decimal points; a rounded figure will suffice. The paucity of information for some chromosomes has necessitated the lumping of data for considerable lengths of a chromosome arm; the risk figures derived in this way are, naturally, composites, and indicative rather than definitive. We assume that, in different families with (apparently) the same translocation, the genetic risks will be the same, regardless of what may have been the mode of ascertainment. And of course, the following principle always applies: if the counselee's family is large enough, do a segregation analysis to derive a private recurrence risk.

Figure 4-17. Viability of combined duplication/deletion states, according to amount of imbalance, measured as % HAL. Most (96%) fall within the triangular area whose hypotenuse lies between 4% duplication/0% deletion and 2% deletion/0% duplication, and a few outliers define an envelope of viable imbalances. (From Cohen et al., 1994, courtesy O. Cohen, and with the permission of Springer-Verlag.)

The figure given for a segment, say, q31/q34–qter—in other words, a lumped figure applying to a segment extending anywhere from q31–qter to q34–qter—might be given as 0.8%, a very small risk. (The “” sign in Tables 4-4 and 4-6 is used for risk estimates in those translocations where no additional aneuploid child has been born apart from probands.) But this figure might have been based mostly on data from families having a q31 breakpoint. A breakpoint at q34 might happen to exclude a dosage-sensitive region of major effect within q33, and thus imbalance for the slightly smaller segment q34–qter might be of considerably greater viability. The risk figure needs to be interpreted intelligently in light of what is otherwise known from the literature about the segments in question (consult Schinzel, 2001), and naturally from observation within the same family.

The reader consulting and using these figures, imperfect though they may be, will gain a good sense of the practical principles of estimating risk. Another tool is the HC Forum, provided on the Internet by Cohen et al. (1992) at hcforum.imag.fr, which uses computerized translocation data from their own and others' material. This can be used as a helpful check on the correctness of the counselor's pachytene diagram and listing of imbalanced gametes. Some risk estimates from one or other of these sources may need to be treated with reser-vation,9,10,11 and counselors should always make their own judgment based on first principles and on an intelligent assessment of the literature relevant to a particular potential imbalance.

A very few translocations, or at least some breakpoints, occur with sufficient frequency that specific risk data can be derived. The obvious example is the recurrent t(11;22). Enough 17p13 translocations were on record to enable Pollin et al. (1999) to work out a group risk figure for distal 17p trisomy or monosomy (the latter producing Miller-Dieker syndrome). The 17p segments ranged in size from 0.12% to 0.38% of HAL, while the segment from the other chromosome was 0.12%–0.67%, and the combined trisomic/ monosomic fractions from 0.24% to 0.9% of HAL. The risks were rather high: 19% for a child with an unbalanced karyotype, rising to 26% if unkaryotyped abnormal infants were included.

Table 4.4. Specific Risk Figures for Having a Liveborn Aneuploid Child Because of Single-Segment Imbalance from 2:2 Adjacent-1 Segregationa

Table 4.5. Specific Risk Figures for Having a Liveborn Aneuploid Child Because of Imbalance from 2:2 Adjacent-2 Segregation

Centric segment that would be imbalanced

RISKa

%

S.D.

4.

4pter→

q11–13

?0

 

8.

8pter→

q12–13

?

 

9.

9pter→

q11–13

18.4

4.5

10.

10pter→

q11–21

?

 

12.

12pter→

q11–13

?

 

13.

13pter→

q14–21

?

 

14.

14pter→

q21–22

?

 

15.

15pter→

q13–24

11.8

7.8

20.

20pter→

q11.1

27.3

13.4

21.

21pter→

q11.1–22

?

 

No obvious difference exists according to sex of parent. For segments not listed, no specific data are recorded in Stengel-Rutkowski et al. (1988).
aFigures are expressed as a percentage of all pregnancies.
S.D., standard deviation; ?, rare cases have occurred, but data are too few to derive a figure; ?0, probably no risk.
Source: From Stengel-Rutkowski et al. (1988).

Individual circumstances for different types of predisposing translocations are discussed below. The lowest risk, namely zero, applies in the case of imbalances of large genetic content. In families interpreted as being in this category, prenatal diagnosis could be seen as unnecessary (Vauhkonen et al., 1985).

Risk at Time of Prenatal Diagnosis

The risk of detecting an abnormality is higher at prenatal diagnosis than at the birth of a live baby. This is because there is differential survival throughout pregnancy, with spontaneous loss more likely (often much more likely) in an abnormal than in a normal pregnancy. Very unbalanced conceptions will abort before the time of prenatal diagnosis. Daniel et al. (1988) derived an overall figure of about 25% for carriers to have an unbalanced fetal karyotype detected at amniocentesis when ascertainment was through a previous aneuploid child, and about 5% when it was through recurrent miscarriage. The amniocentesis figure is at its highest, 35%, in the carrier whose risk of having an aneuploid live birth lies in the medium range (5%–10%) (Stengel-Rutkowski et al., 1988). To give an example from a specific chromosomal segment, Stengel-Rutkowski et al. record a 6% risk for an imbalance in the liveborn from translocations with a proximal 9p breakpoint, versus a 33% risk at amniocentesis. In a series of 57 pregnancies in 40 translocation couples, Barišić et al. (1996) determined an overall risk of 16% to discover an unbalanced karyotype at second-trimester amniocentesis, confirming a higher risk (32%) for couples who had previously had an abnormal child, versus a lower figure (12%) when ascertainment came from miscarriage. The counselor should be clear about these different types of risk figure being offered (Table 3-1). Ultrasonography can be used as an adjunctive procedure, with normal nuchal translucency in the first trimester and absence of structural anomalies in the second trimester suggesting a normal/balanced karyotype (Sepulveda et al., 2001).

Table 4.6. Specific Risk Figures for Having a Liveborn Aneuploid Child Because of Imbalance from 3:1 Single-Segment Segregation

Risks According to Likely Segregation Mode

Adjacent-1 Segregation, Single-Segment

Specific risk figures for individual single-seg-ment imbalances are set out in Table 4-4. A notable point is the number of risk figures that are very small, less than 1%. This most likely reflects the observation that many imbalances are almost always lethal in utero, and survival through to term is the exception. In fact, we can say that in order of frequency, there are imbalances that are (1) invariably lethal, (2) almost always lethal, (3) usually lethal, and, the least frequent category, (4) usually survivable. These risk figures are likely to be valid irrespective of the mode of ascertainment of the family or of the identity of the other chromosome contributing the telomeric tip, at least in the majority of translocations.

By way of example, imagine that the t(3;11)(p26;q21) carrier aunt of the male affected case in the family shown in Figure 4-6 had sought advice about her own risk to have a baby with the same defect. The single-seg-ment involved is 11q21–qter. According to the rules set out above, adjacent-1 segregation is the category that implies risk for viable imbalance in this family translocation. Consulting Table 4-4, therefore, we see that the exact segment 11q21–qter is not listed, but it comes within the compass of the segments 11qter–q13 to 11qter–q22. The best estimate of the group risk for this compass of segments is <2.6%, this risk probably referring only to the trisomic state.9 Some of the figures for liveborn aneuploidy risk in Table 4-4 may be underestimates, in that unkaryotyped early infant deaths and stillbirths had not been included. These figures are, however, noted in parentheses alongside the figure for known aneuploids, and may warrant some upward adjustment of the risk assessment. In this specific 11q21 case, no such data are noted.

Adjacent-1 Segregation, Double-Segment

Every double-segment translocation is likely to be a unique case (or at least no other described family is known), and risk assessment is less precise. Of course, if the family is large enough, a private segregation analysis will provide the best estimate. Otherwise, Stengel-Rutkowski et al. (1988) recommend considering each segment separately. They propose the rule of thumb that the risk will be half that of the smaller of the two risk figures. Even this may be an overestimate; in many cases, the dupli-cation/deficiency from a double-segment imbalance will invariably be lethal in utero—a risk of 0%—notwithstanding that each segment separately is on record with viability in the single-segment state.

Preimplantation genetic diagnosis requires a different viewpoint, since in utero lethality has not had the chance to operate, and the denominator of the risk figure is quite different: we are now concerned with the rate of abnormalities in the day 3 embryo. In a double-seg-ment translocation t(3;11)(q27.3;q24.3) carried by a brother and sister reported in Coonen et al. (2000), at least 15 out of 18 embryos of the brother were karyotypically unbalanced, and only 1 was normal or balanced. This one embryo was transferred, amniocentesis showed 46,XX,t(3;11), and a healthy carrier daughter was in due course born. His sister underwent two treatment cycles, with two out of six embryos being apparently normal, but neither transferred successfully.

Adjacent-1 Segregation with Subtelomeric Double-Segment Exchange

The viable outcomes will be due to one or possibly both of the dup/del and del/dup combinations, and the family history may well be informative. Either of the combinations could arise from adjacent-1 segregation, or from independent 1:1 segregation of each normal homolog and its derivative chromosome (see Fig. 4-8). From 1:1 segregation, if that occurred, a theoretical risk of 50% for unbalanced karyotypes would apply at conception. In one PGD case reported, relating to a couple, one of whom carried a t(2;17)(subqter;subqter), 13 of 18 embryos showed 2:2 segregation for the translocation (six alternate, seven adjacent-1), consistent with either two 1:1 events, or 2:2 disjunction from a quadrivalent (McKenzie et al., 2003). But the fact that the remaining five malsegregants displayed 3:1 disjunction suggests that a quadrivalent may indeed have formed, even if 3:1 is contrary to the “rules” of malsegregation set out on p. 68. More such cases will require study before a clearer picture can be drawn; but in any event a high risk, and as judged from the family history, is very probable.

Adjacent-2 Segregation

Very few translocations are capable of producing viable adjacent-2 segregant products, and the data on specific risk levels are limited (Table 4-5). When the potential imbalance has considerable viability, e.g., trisomy 9p and trisomy 21q, the risk is likely to be substantial, and may be in the range 20%–30%. The carrier mother in Figure 4-9 would have, from Table 4-5, an 18% risk for the recurrence of trisomy 9p.

3:1 Segregation, Tertiary Aneuploidy

In contrast to 2:2 segregation, the probabilities for unbalanced 3:1 outcomes differ between the sexes, with females having the greater risk (Table 4-6). While tertiary trisomy/monosomy is an uncommonly seen category, it is notable that the common 46,t(11;22)(q23;q11) practically always produces a viable abnormal baby only from 3:1 segregation with tertiary trisomy (Fig. 4-11). The risk for this outcome is 3.7% and 0.7%, respectively, for the female and male carrier.10 For other translocations in this category, the risk is generally small, and is less than 2%. Nevertheless, each translocation is entitled to its individuality, and atypically higher risks are possible, as may be exemplified in the t(12;13) noted above and shown in Figure 4-12, in which two out of four children had a tertiary monosomy. In this sort of case, it could be that the tiny derivative segregates independently, at random.

3:1 Segregation, Interchange Trisomy

The risk of having a child with Patau, Edwards, or Down syndrome from an interchange trisomy is remarkably small. It may be in the vicinity of 0.5% in the female, and less than this in the male (Stengel-Rutkowski et al., 1988). Upper limits of the estimated risks are given in Table 4-6. The figures for PGD can be much higher for 3:1 (see Table 4-2), and as illustrated by the case of Conn et al. (1999) noted above, in which a woman with the karyotype 46,XX,t(6;21)(q31;q22.3) had 9/9 embryos with chromosome imbalance, including two with interchange trisomy 21 and one with probable interchange monosomy 6.

More Than One Unbalanced Segregant Type

It is probably prudent to assume than when more than one mode of segregation can lead to a viable outcome, the overall risk will be cumulative and will be given by the sum of the individual risks. Thus, the carrier mother of the t(11;18)(p15;q11) shown in Figure 4-14 would have a risk comprising three components: duplication 18q11–qter due to adjacent-1; tertiary trisomy 18pter–q11 due to 3:1; and trisomy 18 due to 3:1 interchange. From Tables 4-4 and 4-6, and choosing the closest listed segments, these risks are 2.5%, 1.3%, and 0.2, respectively, for a total of up to 4.0%.11

Imprintable Chromosomes and Uniparental Disomy

In theory, any chromosome with an imprint-able segment is to be considered from this specific perspective. James et al. (1994) studied 21 individuals with an inherited balanced reciprocal translocation and various phenotypic (mostly functional neurological) abnormalities, and none had UPD. Engel and Antonarakis (2002) list 10 familial translocations from the literature associated with the birth of a child with a UPD syndrome, including Prader-Willi and Angelman syndromes (PWS and AS), Beckwith-Wiedemann syndrome, and UPD14. A theoretical risk for UPD following postzygotic correction is noted above (p. 79). What looks like alternate segregation in the fetus could actually have been 3:1 interchange trisomy, with a post-conceptual loss of the homolog in question. In practice, this appears to be an exceedingly rare outcome (Kotzot, 2001; Dupont et al., 2002). But this much is certain: any translocation involving chromosome 15 in particular is to be approached very circumspectly (see the discussion on PWS and AS, pp. 324 and 325). The translocation with a 15q11–q13 breakpoint is a special case, and unequal crossing-over can give a recombinant chromosome cytogenetically indistinguishable from the balanced state (Horsthemke et al., 1996).

Phenotype and Survivability

A major degree of dysmorphogenesis involving several body systems, along with globally disordered brain function, is the typical picture in viable autosomal imbalance. Many patients will come with the knowledge of the particular phenotype of at least one of the viable segregant outcomes—the proband in their own family. The same imbalance in a future pregnancy would be expected to lead to a similar physical and mental phenotype. Survivability is less predictable because for many conditions there is a fine line between relative robustness and a fragile hold on existence, intrapartum and postnatally. Whether or not there is a heart defect (a frequent malformation in many chromosomal disorders) may be a major factor in this. As for the phenotype of potentially survivable outcomes other than those already exemplified in the family, reference to the chromosomal catalogs (de Grouchy and Turleau, 1982; Schinzel, 2001) and to the journal literature provides a guide. For imprintable chromosomes, there may be an influence of the parental origin of the aneuploid segment (see above).

The Parental Balanced Translocation in the Fetus

The conventional wisdom is that if the same (balanced) karyotype found in the carrier parent is detected at prenatal diagnosis, there is no increased risk for phenotypic abnormality in the child: like parent, like child. Some have doubted this, and Fryns et al. (1992a) measured a 6.4% risk of mental and/or physical defects in the heterozygous children of translocation carriers (this figure includes the background risk of 2%–3%). Others remain skeptical and impute ascertainment bias as the confounding factor (Steinbach, 1986). Theoretical mechanisms whereby an apparently balanced translocation could have a deleterious consequence, the parental normality notwithstanding, include the following four: a cryptic unbalanced defect beyond the resolution of routine cytogenetics; the postzygotic loss of a derivative chromosome in one cell line, converting an unbalanced to a mosaic balanced/unbalanced state; a position effect; and uniparental disomy.

Concerning the cryptic unbalanced defect, Wagstaff and Hemann (1995) provide a disconcerting example: an apparently balanced reciprocal translocation that turned out to be a complex chromosome rearrangement, with a tiny segment from the breakpoint of one of the translocation chromosomes inserted into a third chromosome (p. 192 and Fig. 11-7). In families in which the balanced translocation has been transmitted to numerous phenotypically normal individuals, such a scenario is most unlikely, since consistent cosegregation of the “cryptic chromosome” to give an overall balanced complement in all these persons would be improbable. Where the translocation is of more recent origin, perhaps de novo in the parent, the possibility may be less remote. The increasing sophistication of molecular cytogenetics may enable such cases to be teased out.

A single case reported by Dufke et al. (2001) illustrates the possible scenario of mosaicism. An abnormal child with the same balanced t(17;22)(q24.2;q11.23) as his mother on peripheral blood analysis showed on skin fibroblast culture a 47,t(17;22),+der(22) karyotype. This mosaic picture may reflect there having been an interchange tertiary trisomy7 complement in the conceptus, with postzygotic loss of one of the two der(22)s in blood-forming tissue. Of course, this report raises the following question: could the excess noted by Fryns et al. in a postnatal population be accounted for, at least in part, by this process? And, if so, could this be the basis of a misleading prenatal diagnosis? In fact, it could be imagined that, if the mother in Dufke et al. had had an amniocentesis, the unbalanced 47,t(17;22),+der(22) state would have been seen, since the sampling of amniocytes is somewhat equivalent to taking several skin biopsies. On that premise, it could be argued that a good number of normal cells/colonies from amniocentesis would indicate the unlikelihood of any such mosaicism.

The concept of position effect is discussed above (p. 85). A particular gene in the close vicinity of a translocation breakpoint may function normally in a parent; but, for whatever reason, in the child the gene in question may be silenced, because of an effect of the adjacent chromatin of the other participating chromosome. Of similar extreme rarity, there is the theoretical question of uniparental disomy following postzygotic “correction” of an interchange trisomy for an imprintable chromosome.

If an important additional risk due to one or other of the above scenarios really does exist, it is surely very small, perhaps no more than a fraction of a per cent, that a child with the “balanced” parental karyotype might have a defect of mostly unpredictable severity and extent. In the meantime, it remains true that in the great majority of cases the balanced translocation really is balanced, structurally and functionally, and will have, of itself, no detrimental effect. Thus, in practical terms, it is quite appropriate to advise continuing a pregnancy when the fetal karyotype is the same as that of the carrier parent, and with very considerable (if not absolute) confidence in the high probability of a normal outcome.

PREGNANCY LOSS AND INFERTILITY

Miscarriage

Conceptions with large imbalances will abort. Against the background population risk of 15% for a recognized pregnancy to miscarry,12 the risk for the translocation carrier is much greater, and is in the range 20%–30% (Stengel-Rutkowski et al., 1988). For a few, the risk is very high, well over 50%. An increasing viability of conceptuses implies a correspondingly diminishing likelihood of pregnancy loss by miscarriage. Not to diminish the distress felt at the loss of a welcomed and wanted pregnancy, patients can be heartened that miscarriage, in this setting, is the natural elimination of a severe abnormality, which provides the opportunity to make a fresh and perhaps more fortunate start. For a couple having lost all pregnancies to miscarriage, karyotyping in the previous generation may be helpful. The consultand would embody the proof in him- or herself that the heterozygote can have a normal child, should one of his or her parents also be a carrier. Optimism has to be muted, however, in the setting of a family history of many miscarriages, which may indicate a propensity for the production of unbalanced gametes. Recourse to preimplantation diagnosis (Chapter 24) may offer hope to some in that situation.

Infertility

Occasionally, some male translocation carriers are infertile with a spermatogenic arrest, as discussed above. There is little (if any) increased incidence of infertility in the female; oogenesis is a more robust process.

OTHER ISSUES

Cancer Risk

In rare familial translocations, the rearrangement may promote mitotic malsegregation, or disrupt a tumor suppressor gene, and thus comprise a “hit” in the cascade of events leading to the cellular phenotype of cancer. A well-recognized case is that of chromosome 3 translocations implicated in familial renal cancer, of which a number of examples have been published (Kanayama et al., 2001; Bodmer et al., 2002a, 2002b; Meléndez et al., 2003). According to one construction, a three-hit sequence is envisaged, the first hit being the actual inheritance of the balanced translocation. Then, the mechanism is a mitotic malsegregation in an embryonal kidney cell. The derivative chromosome containing the 3p segment is lost, and consequently one daughter cell, and thus the lineage from it, has only one copy of distal 3p, on the normal homolog (the second hit). Thereafter, on this remaining normal chromosome, a somatic mutation occurs in postnatal life at a tumor suppressor gene (VHL13 on distal 3p) in a kidney cell within this lineage (the third hit); now the stage is set for a renal cancer to come into being. According to another construction, in the case of a translocation of which the no. 3 breakpoint is at 3q21, disruption of a gene (DIRC2) sited at the actual breakpoint may comprise the first hit, and the consequential haploinsufficiency may of itself be sufficient to initiate the process of oncogenesis.

Otherwise, a t(3;6) is recorded which possibly predisposes to hematological malignancy (Markkanen et al., 1987). Two families are known with familial adenomatous polyposis being due to a constitutional reciprocal translocation, t(5;10)(q22;q25) and t(5;8) (q22;p23.1), the relevant tumor suppressor gene (APC) having been disrupted at the 5q22 breakpoint (van der Luijt et al., 1995; Koorey et al., 2000). Laureys et al. (1990) reported a child with a de novo t(1;17)(p36.2;q12) presenting with neuroblastoma; it may be that somatic loss of the der(17) and duplication of the normal no. 17 in cells of the nervous system set the stage for the tumor to arise (Van Roy et al., 2002). An excess of constitutional rearrangements in a series of children with various tumors suggests the possibility of a causative role for some of them (Betts et al., 2001). The possible increase in risk for breast cancer in the common t(11q;22q) has been noted above, although the case is yet to be confirmed (Lindblom et al., 1994). Where the cytogenetic–cancer associations are firm, heterozygotes should receive appropriate counseling, and entry into a cancer surveillance program is appropriate. Often, the associations appear to be no more than fortuitous (given that rearrangements are not uncommon, and cancer is very common).

Interchromosomal Effect

Originally, there was concern that a reciprocal translocation heterozygote might be prone to produce gametes aneuploid for a chromosome not involved in the translocation, specifically, in this context, chromosomes 13, 18, 21, or X. Warburton (1985) reviewed the associations of reciprocal translocations and trisomy 21 from unbiased (amniocentesis) data, and found no evidence to support the contention. Uchida and Freeman (1986) and Schinzel et al. (1992) studied families in which a child with trisomy 21 also had a balanced translocation, and while in several the translocation was of paternal origin, in fact the extra chromosome 21 came from the mother.

More directly, numerous sperm karyotyping studies have, for the most part, shown no increase in disomies unrelated to the translocation (Martin and Rademaker, 1990; Van Hummelen et al., 1997; Martini et al., 1998), although some workers have raised slight doubts (Estop et al., 2000; Blanco et al., 2000; Vegetti et al., 2000). Pellestor et al. (2001) suggest that carrier males with poor semen indices are the only ones in whom any such effect might exist; in which case, it might be the altered testicular environment, rather than the translocation of itself, that is the cause (Egozcue et al., 2000). Analysis of embryos at preimplantation diagnosis supports the case against any effect (Gianaroli et al., 2002). Possibly, some specific translocations may have a very small individual risk, but there seems little reason to withdraw from the generality of Jacobs' assessment from 1979, and in practical terms we expect this view to prevail: “there is no indication that parents with a structural abnormality are at an increased risk of producing a child with a chromosomal abnormality independent of the parental rearrangement … [and] their recurrence risk for such an event is the same as the incidence rate in the population.” Only with infertile men (needing ICSI for conception) might there really be a small risk, and this may be due to the infertility per se, rather than to the particular translocation.

Notes

1. There is scope for confusion in the use of these terms: of course, all reciprocal exchanges, by definition, involve two segments. A true single-segment exchange—that is, a one-way translocation—is generally considered not to exist, in that a segment of chromosome cannot attach to an intact telomere, although there are rare exceptions to this rule (Martin et al., 1986a; Rossi et al., 1993). The distinction begins to break down when a translocated segment is very small but could still contain genes (as in the subtelomeric translocation). Be this as it may, the terms double- and single-segment exchange, used knowledgeably, serve a practical purpose.

2. At least for chromosome 16p, a small subterminal deletion may be without any effect, other than to cause the harmless condition of α-thalassemia trait. Of the 15 genes in the top 250 kb of the chromosome, none, apparently, is susceptible to a haploinsufficient effect, except for the α-globin locus (Horsley et al., 2002).

3. The reader wishing to study further worked examples is referred to Midro et al. (1992), who analyze in some detail a series of translocations of differing risk potentials.

4. This same 4;8 translocation has been observed in a small number of unrelated families, and it may be, after the t(11;22) noted below, the most frequent human reciprocal translocation. This recurrence may reflect the presence in distal 4p and 8p of “olfactory-receptor clusters” which can act as recombination-predisposing duplicons (Giglio et al., 2002).

5. This is not official ISCN nomenclature.

6. One breakpoint may compromise the function of a tumor suppressor gene; there is preliminary evidence that the t(11;22) carrier has an increased risk for breast cancer (Lindblom et al., 1994), but a definitive study, looking at large numbers in several kindreds, has yet to be done.

7. As a point of rather academic interest, the same overall imbalance can exist in a child having both translocation chromosomes, plus an additional copy of the der(22), as a sort of “interchange tertiary trisomy.” The karyotype could possibly be written 47,t(11;22)(q23;q11),+der(22), a subtle distinction from the typical abnormal karyotype. Nondisjunction at meiosis II in a balanced spermatocyte is one route whereby this might arise, and 3:1 disjunction following a crossover in the 22q interstitial segment is another (Petković et al., 1996).

8. Except in the extraordinary setting of postzygotic rescue. Kulharya et al. (2002) report a t(11;22) carrier mother having had a child from presumed adjacent-1 segregation with 46,XY,der(22) at conception, and then mitotic loss of the der(22) in one cell and duplication of the normal 22, leading to 46,XY,der(22)/46,XY mosaicism.

9. The figure given on entering this translocation into the HC Forum is 20%–25%. This departs too much from observation in the family to accept. But referring to the original data in Stengel-Rutkowski et al. (1988), we see also that the <2.6% figure is based on rather limited data from only a few families, and so neither should it be accepted at face value, at least for this family. The moral: consider figures carefully, and interpret them critically.

10. These figures are from Stengel-Rutkowski et al. (1988), as listed in Table 4-6. In a very large collaboration, with data from 110 families seen in 15 countries (there being some overlap with the material in Stengel-Rutkowski et al.), the risk figures for the male and female heterozygote were 1.8% and 2.1%, respectively (Iselius et al., 1983). Notably, in most of these families the index case was the only one known definitely to have the unbalanced karyotype. However, it could be supposed that reported malformed stillborn infants in these families were likely also to have had the unbalanced karyotype, and making this assumption, the risk figures for a live- or stillborn affected infant would increase to 5% and 5.7%, respectively. Carrier mothers had considerably more offspring than did carrier fathers, pointing to a reduced fertility in the male heterozygote.

11. Intuitively, this figure may seem too low, given the well-known viability of partial and complete trisomy 18. The figure of 15% from HC Forum might seem to be more plausible.

12. This figure applies to clinically diagnosable miscarriage, mostly occurring in the period 8 to 16 weeks gestation. Severely imbalanced forms may be lost as very early, even occult, abortions (see p. 343).

13. VHL = the von Hippel Lindau locus; DIRC2 = the Disrupted in Renal Cancer 2 locus.