American Journal of Medical Genetics 1 :445-460 (1978)

Replication Patterns of Three lsodicentric X Chromosomes and an X lsochromosome in Human Lymphocytes Gordon Dewald, Jack L. Spurbeck, and Hymie Gordon May0 Clinic and Mayo Foundation, Rochester, Minnesota

Chromosomes from four patients with variants of the Turner syndrome were investigated by G- and C-banding and DNA replication techniques. Their karyotypes were: 1) 46,X,idic(X)(q28), 2) 45,X/46,X,idic(X)(q24), 3) 45,X/ 46,X,idic(X)(pl l ) , and 4) 46,XJXq). In Patients 1, 2, and 3, the abnormal X was isodicentric, with different break-and-fusion points in each case. In each, t h e G-band pattern on one side of the breakpoint was a mirror image of that on the other side. Each had two distinct C-bands, only o n e of which was associated with a primary constriction. The fourth patient had an isochromosome of the long arm of an X in which only one C-band could be discerned. Replication studies were done o n lymphocyte cultures by incorporating a thymidine analogue and staining with acridine orange. In addition, replication patterns of normal early- and late-replicating X chromosomes were studied in two normal females. In the four patients, all t h e normal X chromosomes had normal early-replication patterns. The two idic(X) chromosomes with breakand-fusion points o n their long arms almost always had symmetric replication patterns, which demonstrates that t h e corresponding bands replicated synchronously. In contrast, many of the idic(X)(pll) and i(Xq) chromosomes showed asymmetric or asynchronous replication. In each, t h e replication pattern of the abnormal X was similar t o the equivalent portions of a normal late-replicating X. Key words: replication, isochromosome, dicentric, isodicentric, X chromosome INTRODUCTION

Recently, several methods have been developed t o detect the incorporation of halogenated thymidine-base analogues into replicating chromosomes [19,21,22,28] . This development has made it possible t o determine the relative time of replication of individual bands along chromosomes [23] and to describe the replication pattern for all 24 human chromosome types [ l l , 13,201. The replication patterns of t h e X chromosome have been

Received for publication July 25, 1977; accepted December 8, 1977. Address reprint requests to Dr Gordon Dewald, Mayo Clinic and Mayo Foundation, Rochester, MN

55901.

0148-7299/78/0104-0445f102.90 @ 1978 Alan R . Liss, Inc

446

Dewald, Spurbeck, and Gordon

described in detail, and the sequence in which the individual bands replicate on normal early-replicating and late-replicating X chromosomes has been established [17,23,36] . This sequence varies to some extent, depending on the tissue source [35] . In addition, the replication patterns of abnormal X chromosomes have been studied, including isodicentric X chromosomes [idic(X)] and isochromosomes of the long arm [i(Xq)] [2, 12,241. The findings have not been consistent in respect to i(Xq): Whereas Latt, Willard, and Gerald [24] reported the replication patterns to be symmetric on the two arms, Dutrillaux [12] and Baranovskaya et a1 [2] found the patterns to be different. The observation by Latt, Willard, and Gerald [24] is supported by autoradiography of i(Xq), which shows similar distribution of grains on the two arms [25,26]. Latt, Willard, and Gerald [24] and Dutrillaux [12] have reported the replication patterns of idic(X). (They actually refer to them as X-X translocations.) In the two cases reported by Latt, Willard, and Gerald, the replication patterns of the two components of the idic(X) were symmetric. In the five cases reported by Dutrillaux, the replication patterns apparently were asymmetric. Isodicentric X chromosomes are of special interest to cytogeneticists because they possess two C-band regions but usually exhibit only one primary constriction and normal anaphase segregation. These observations have prompted the suggestion that one of the centromeres somehow becomes nonfunctional, either by a process similar to gene inactivation or by deletion [ 5 , 10, 311 . Isodicentric X chromosomes are of special interest also because they present an opportunity to study the replication behavior of duplicate copies of the same bands on the same chromosome. Because of the reported inconsistency of the replication patterns of i(Xq) chromosomes and the sparsity of information about idic(X) chromosomes, we investigated the replication patterns of the X chromosomes from lymphocyte cultures of four patients with variants of the Turner syndrome, including two with idic(X) chromosomes with breakpoints on the long arms, one with an idic(X) with a breakpoint on the short arm, and one with an i(Xq) chromosome. In addition, the replication patterns of the early-replicating and the late-replicating X chromosomes of two normal females were studied. CASE REPORTS Patient 1

This patient is a 24-year-old woman who has primary amenorrhea. She is rather tall (height 175 cm), with relatively long limbs. Little spontaneous female secondary sex development occurred, but the response to treatment with estrogen has been satisfactory. Her serum follicle-stimulating hormone (FSH) level is abnormally elevated (1 11 pg/dl). Her buccal smear showed unusually large and some bifid X bodies in 46% of nuclei; there were no Y bodies. Her karyotype is 46,X,idic(X)(q28) in her peripheral blood and tissue fibroblasts. A total hysterectomy and bilateral salpingectomy and gonadectomy were performed. Histologic examination showed bilateral streak gonads. Patient 2

This patient, a 17-year-old girl with primary amenorrhea, is 160 cm tall and relatively long-limbed. Her secondary sex characteristics are moderately well developed. Her serum estrogen level is abnormally low (3.2 ng/dl), and her serum FSH level is abnormally elevated (181 pgldl). Her buccal smear showed only 7% X bodies and no Y bodies.

Replication in idie(X) and i(Xq)

447

Chromosome analysis of blood and tissue fibroblasts showed mosaicism: 45,X/ 46,X,idic(X)(q24). Laparoscopy showed an infantile uterus and fallopian tubes with streak gonads confirmed histologically. Patient 3

This patient, a 28-year-old woman with primary amenorrhea, is short (height 147 cm) but normally proportioned. Slight secondary sex development began at age 15 years. Cyclic treatment with estrogen and progestin resulted in substantial secondary sex development. The buccal smear showed X bodies in just 2-3% of nuclei, with no Y bodies. The karyotype in her peripheral blood is 45,X/46,X,idic(X)(pll). Examination of her tissue chromosomes and histologic analysis of her gonads have not been done. Patient 4

This patient, a 21-year-old woman, has primary amenorrhea. Her height is 144 cm and her female secondary sex development is minimal. Her serum FSH level is abnormally elevated (1 61 pgldl). Her buccal smear was reported to have normal X bodies and no Y bodies. Her peripheral blood karyotype is 46,X,i(Xq) with no indication of mosaicism. Chromosomes from other tissues have not been examined, and the gonads have not been studied histologically.

METHODS

For all four patients, peripheral blood lymphocyte cultures were established on Difco chromosome medium for 72 hours. Fibroblast cultures were established by the method of Titus [32] from skin and gonad biopsy specimens for Patients 1 and 2 , on whom laparoscopy was done. All cultures were harvested by using demecolcine (Colcemid@) (0.5 pg/ml) for one hour, hypotonic KC1 (0.75 M), and methanol-glacial acetic acid (3: 1) fixative. For chromosome analysis, preparations were made with G-bands by trypsin [30], with C-bands by NaOH [ l ] ,and with the Giemsa nonbanding method. Replication patterns of the three idic(X) chromosomes were studied by autoradiography and by 5-iododeoxyuridine (IdU) incorporation. Replication of the i(Xq) chromosome was studied only by 5-bromodeoxyuridine (BrdU) incorporation. One of the normal subjects was studied with IdU and the other with BrdU. All the replication studies were done on peripheral blood lymphocytes grown in Difco chromosome medium for 72 hours at 37°C. IdU or BrdU (30 pg/ml) was added t o each culture 3.5 hours before harvesting (B-pulse) [36], after which, to avoid photolyses of DNA containing IdU or BrdU, the cultures were kept in opaque containers [18] . Each culture was treated with demecolcine (0.5 pg/ml) for one hour and then harvested by using hypotonic KCl (0.075 M) and methanol-glacial acetic acid (3: 1) fixative. Slide preparations were dried in air, stained with acridine orange (0.1 mg/ml deionized water), placed in a Sorensen phosphate buffer (pH 6.8) for 2 minutes, and then mounted in the same phosphate buffer. The slides were examined with an incident light fluorescent microscope system equipped with a Leitz KP 490-nm excitor filter and a 51 0-nm barrier filter. Suitable metaphases were photographed with Kodak Panatomic-X@ film and developed with Microdol-X@. Metaphases were accepted as suitable for photography when there was evidence of replication on any of the chromosomes and there were few or no overlapping

448

Dewald, Spurbeck, and Gordon

chromosomes. No attempt was made to select metaphases solely on the quality of the X chromosomes. Thus, for each patient we obtained a relatively random collection of photomicrographs containing X chromosomes that had incorporated IdU or BrdU for various periods less than three and one-half hours. Those regions of the chromosomes that had incorporated IdU or BrdU during replication fluoresced dully, in contrast to the bright fluorescent regions that had completed their replication before the addition of the thymidine-base analogue. All analyses were done on 8- by 10-inch enlargements of each metaphase. The X chromosomes in each metaphase were recognized by their relative size, centromere index, and replication pattern. The dull-appearing replication bands correspond approximately to bright Q-bands and to dark G-bands and thus can be used to identify chromosomes [ 4 , 1 4 , 2 2 ] . In each metaphase, the X chromosome with the more numerous and broader areas of dull fluorescence was considered to be the late-replicating X. The Paris Conference model of the X chromosome was used as a guideline; regions that fluoresced dully were assigned band positions based on visual approximations to the model. When all the available metaphases for each patient had been analyzed, the chromosomes were placed into groups having similar replication patterns and a single model was drawn which represented the general pattern for the group. Because this type of study is subjective, individual bands were analyzed independently by two investigators. The few metaphases about which there were irreconcilable differences of opinion were excluded from the study. To reduce subjectivity of analysis further, the band-replication studies were done for all four patients and the two control subjects within a two-day period. Autoradiography was used to determine whether the structurally abnormal X was a late-replicating chromosome, but it was not useful in establishing the replication sequence of individual bands. [3H] thymidine (1 pCi/ml) at a specific activity of 6.7 Ci/mmole was allowed to become incorporated into replicating chromosomes of lymphocyte cultures for three and one-half hours. Autoradiography was done by first preparing G-banded metaphases and then destaining the slides sequentially with 95% ethyl alcohol for 10 minutes, with 10% trichloroacetic acid for 8 minutes, and with 95% ethyl alcohol for 10 minutes. The slides were then coated with Kodak NTB-2 nuclear emulsion, exposed for about five days, and developed with D-19. The slides were restained with Giemsa for microscopic examination. R ESU LTS Normal X Chromosomes

The sequence in which the bands incorporated IdU was established for structurally normal X chromosomes by analysis of 30 different metaphases from a normal female. In each metaphase, the late-replicating X could easily be distinguished from its early-replicating homologue because of its more numerous and broader areas of dull fluorescence caused by the incorporation of greater quantities of IdU. Figure 1a shows five representative early-replicating X chromosomes arranged in order of decreasing number of dull fluorescing bands. The chromosome at the extreme left has incorporated the greatest amount of IdUand the chromosome at the extreme right has incorporated the least. Sometimes we observed IdU bands that differed in breadth from the Paris Conference diagram; we show them in our models as we saw them. By inspecting each chromosome or model, beginning on the right and proceeding to the left, it is possible to determine the sequence in which the bands replicate. In terms of the Paris

Replication in idic(X) and i(Xq)

449

Conference diagram, it is apparent that the last dark band to replicate is p21, preceded by q21, q27, q25, and q23, in that order. Using the three and one-half-hour €3-pulse method, we did not obtain sufficient incorporation of IdU t o evaluate the replication patterns of the light bands of the Paris Conference model. The replication patterns of late-replicating X chromosomes are shown in Figure Ib. The chromosomes are arranged in the same manner as in Figure l a . In this case, the last

Paris

a

Conference

b

Cohference

Paris

A

A

99m 0

C

B

C

D

D

E

E

Fig 1. a) Series of early-replicating X chromosomes from a normal female, arranged from left to right in order of decreasing number of replication bands. Each chromosome (above) is representative of a group of similar chromosomes, whose composite replication pattern is shown as a diagrammatic model (below). b) Series of late-replicating X chromosomes from a normal female, arranged as in la.

450

Dewald, Spurbeck, and Gordon

dark bands to replicate are p21 and q21; bands q25 and q27 replicate earlier but almost simultaneously. Band q23 usually replicates earliest, but sometimes we observed that it replicated almost as late as q21. The last light band to replicate appears to be q22, which is immediately preceded by q24, q26, and q28, in that order. The dark bands appear to “grow” lengthwise from the middle of a band outward to its borders, which suggests that band replication probably begins in the middle and proceeds outward. We noted an additional relatively late-replicating band at p22 in both early- and late-replicating X chromosomes. Although the Paris Conference diagram does not indicate it, we often see a distinct dark G-band at this site in our routine preparations. This same band has been reported in replication studies by Epplen, Siebers, and Vogel [I31 and by Willard [35]. We also saw replication bands at the q12 site, but their detection was hampered by their small size and their nearness to the centromeric constriction. Thus, they were seen only in less contracted chromosomes, and we were unable to determine the relative timing of replication objectively. Twenty-seven metaphases from the other normal female were studied by the BrdU B-pulse technique. The replication patterns of the early- and late-replicating X chromosomes were indistinguishable from those found with IdU. Patient 1: idic(X)(q28)

As shown in Table I, for Patient 1 420 of the 421 metaphases from skin, right and left gonadal regions, and peripheral blood lymphocytes had the karyotype 46,X,idic(X) (q28); a representative of the idic(X)(q28) chromosome is shown in Figure 2. The solitary exception had the karyotype 45,X. The loss of the idic(X) in this cell could be artifactual because there was no other evidence of mosaicism. No metaphases were seen with more than one idic(X) or with broken portions of an idic(X). In C-band preparations, two C-bands were observed on the abnormal X (Fig 2). One C-band corresponded to the primary constriction. The other C-band corresponded to a region of slight indentation in some cases or to a region with no visible indentation in other cases. As might be expected for an isodicentric chromosome, the G-band pattern is symmetric on the two sides of the q28 breakpoint (Fig 2). That the pattern is symmetric has also been confirmed by videodensitometric analysis [ 151 . The idic(X) was late-replicating, relative to its structurally normal homologue, in 73 metaphases examined by autoradiography and in 20 metaphases examined by IdU

TABLE I. Results of Chromosome Analysis for Three Patients With Isodicentric X Chromosomes and One Patient With an i(Xq) Chromosome

Case

Karyotype

1

46,X,idic(X)(q28)

2

45,X/46,X,idic(X)(q24)

3 4

45 ,X/46,X,idic(X)(pll) 46,X,i(Xq)

Tissue Lymphocytes Skin Left gonad Right gonad Lymphocytes Skin Left gonad Lymphocytes Lymphocytes

Metaphases 101 101 118 101 52 31 32 94 100

45 , x 1(1%)

...

... ... 34 (65%) 6 (19%) 8 (25%) 74 (79%)

...

Abnormal X line 100 101 118 101 1 8 (35%) 25 (81%) 24 (75%) 20 (21%) 100

R e p l i c a t i o n in idic(X) a n d i(Xq)

451

Fig 2. Representative abnormal X chromosomes, G- and C-banded, from each of four patients studied. Chromosomes from Patients 1 , 2, and 3 are isodicentric, showing two distinct C-band regions. Abnormal chromosome from Patient 4 is an isochromosome of long arm of X, in which only one (but larger than usual) C-band is visible.

incorporation. Fourteen photomicrographs of metaphases incorporating IdU were used to establish the replication sequence of individual bands. In each metaphase, the normal X had a replication pattern similar to that of the early-replicating X chromosomes of the normal subjects. All 14 metaphases contained an idic(X) in which the replication pattern was symmetric on both sides of the q28 breakpoint; we will refer to this as a symmetric replication pattern. Nine of the 14 metaphases had similar patterns of replicating bands; a representative chromosome and a composite model are shown in Figure 3. The other five idic(X) chromosomes also had symmetric patterns of replication but with more or fewer bands than shown in the model in Figure 3. The idic(X) chromosome exhibited a replication pattern similar to that of the equivalent portions of a normal late-replicating X.

452

Dewald, Spurbeck, and Gordon

P a ri s C o n f e re n c e Fig 3. An idic(X)(q28) chromosome from Patient 1, showing a symmetric replication pattern. Diagrammatic model below is a composite of general replication pattern found in one of the 14 idic(X) chromosomes studied in this patient. Broken line indicates position of presumed “inactive” centromere.

Patient 2: idic(X)(q24)

This patient was a chromosome mosaic, 45,X/46,X,idic(X)(q24). The idic(X) line was predominant in skin and gonadal-streak tissue; 45 ,X metaphases were more frequent in peripheral blood (Table I). A representative of the idic(X)(q24) chromosome is shown in Figure 2. The C-band method showed two distinct regions of centromeric constitutive heterochromatin (Fig 2). One of the C-bands was at the site of the primary constriction, but there was no constriction associated with the other C-band. By both inspection and videodensitometric analysis [15] ,the G-band pattern suggests that this is an isodicentric chromosome because the components on either side of the q24 breakpoint are the mirror image of each other.

Replication in idic(X) and i(Xq)

453

The idic(X) was later-replicating than the normal homologue in all five metaphases examined by autoradiography and in all 56 examined by IdU incorporation. Analysis of 34 photomicrographs showed the normal X to have an early-replicating pattern. Three different symmetric patterns of replication of the idic(X) were observed: One of these was observed in 19 photomicrographs and is illustrated by model A of Figure 4; another, observed in I 3 photomicrographs, is illustrated by model B of Figure 4; the third pattern, observed only once, is not illustrated. Chromosome model C is the only idic(X) in which asymmetric replication was suggested; one q22 band apparently replicated earlier than its counterpart on the other side of the breakpoint.

Fig 4. The two most common types of symmetric idic(X)(q24) replication patterns seen in Patient 2 are demonstrated by chromosome models A and B and their respective composite models below. Chromosome model C is the only idic(X) from this patient which demonstrated asymmetric replication. Broken line indicates portion of presumed “inactive” centromere.

4 54

Dewald, Spurbeck, and Gordon

Patient 3: i d i c ( X ) ( p l I ) .

The karyotype of Patient 3 was 45,X/46,X,idic(X)(pll). A representative of the idic(X)(pl 1) chromosome is shown in Figure 2. Analysis of 94 metaphases from the peripheral blood showed that 21 % ’ of the cells contained the idic(X) chromosome (Table I). This chromosome always exhibited two C-bands: one at the site of the primary constriction and the other at a site where there was no apparent attenuation (Fig 2). The G-band pattern was bilaterally symmetric around the p l 1 breakpoint, as expected in an isodicentric chromosome. This was apparent by both inspection and videodensitometric analysis ~ 5 1 . The idic(X) was later-replicating than the normal X in the six metaphases studied by autoradiography and in the 41 studied by IdU incorporation. The sequence of replication for individual bands was determined from inspection of the photomicrographs of 24 metaphases containing the idic(X). In each instance, the normal homologue showed a sequence of band replication similar to a normal early-replicating X. The replication patterns of 22 of the 24 idic(X) chromosomes were of four different categories (Fig 5): 2 type A, 7 type B, 7 type C, and 6 type D. The model of type B shows a symmetric pattern of replication, but in the other three models the replication pattern is asymmetric. Two other idic(X)

Paris

Conference

A

B

C

D

Fig 5 . Representative chromosomes, and composite models of groups they represent (below), of the four most common replication patterns for idic(X)(pll) chromosomes from Patient 3. Types A, C, and D show asymmetric replication; type B shows symmetric replication. Broken line indicates position of a presumed “inactive” centromere.

Replication in idic(X) and i(Xq)

455

chromosomes showed different asymmetric patterns of replication. Thus, only seven of 24 idic(X) chromosomes from this patient exhibited symmetric replication; the rest all showed asymmetry. The replication patterns of the duplicate portions of these idic(X) chromosomes resembled those of the late-replicating chromosome of the normal subjects. In model A. asymmetry is apparent because band q24 on the “long” arm appears to have replicated earlier than its counterpart on the “short” arm. In models C and D, the asymmetry occurs because band q22 on the “long” arm has replicated slightly earlier than its counterpart on the “short” arm. The only difference between these two models is that band q22 on the “short” arm replicates earlier in C, whereas q22 replicates earlier on the “long” arm in D. Patient 4: i(Xq)

All 100 metaphases examined from a peripheral blood lymphocyte culture for Patient 4 contained an i(Xq) (Table I). Thus, the karyotype is 46,X,i(Xq). A representative of her i(Xq) chromosome is shown in Figure 2. A single large C-band was evident on her i(Xq) (Fig 2). There was no other evidence, such as abnormal anaphase segregation, to suggest that two functional centromeres were present. The G-band preparations showed two identical long-arm patterns on either side of the centromere region (Fig 2). This observation was also confirmed by videodensitometric analysis [I 51 . Using BrdU incorporation, we studied the replication patterns of the structurally normal X and the i(Xq) in 15 different metaphases. In each instance the normal X exhibited a normal early-replication pattern and the i(Xq) a late-replication pattern. Eleven of the 15 metaphases contained i(Xq) chromosomes that could be classified into one of four replication patterns (Fig 6). Model A (two chromosomes), model B (one chromosome), and model D (three chromosomes) have symmetric replication patterns. Model C (five chromosomes) has an asymmetric replication pattern in which band q22 on the lower arm replicates slightly earlier than its counterpart on the upper arm. Four i(Xq) chromosomes, not shown, also demonstrated evidence of asymmetric replication. In models C and D, the sequence in which the bands replicated on each arm of the i(Xq) chromosome was similar to the sequence on the long arm of a late-replicating normal X. In models A and B, band q27 appeared to replicate earlier than expected from our normal late-replicating X study; however, it is consistent with the reports of other investigators [17, 35, 361. DISCUSSION Normal X Replication

Our findings in respect to the early- and late-replicating normal X are in agreement with those of Willard and Latt [36], with two minor exceptions. First, Willard and Latt reported that in the late-replicating X chromosomes, band q25 begins replication later than q27, whereas in our study q25 and q27 appeared to replicate at about the same time. We agree with Willard and Latt that q27 begins DNA synthesis later than q25 on the earlyreplicating X chromosome. This may represent a notable difference in the replication patterns of early- and late-replicating X chromosomes. Second, our findings differ from those of Willard and Latt [36] in respect to the relative timing of replication of q23. It was our impression that q23 usually was one of the earliest bands to replicate on both the early- and the late-replicating X. In contrast, Willard and Latt regarded q23 as one of the last bands to replicate on the late-replicating X, completing its DNA synthesis later than q27, whereas they stated that q23 replicates before

456

Dewald, Spurbeck, and Gordon

Q

I!

h W

U

A

B

Paris Con f e renc e

C

D

Fig 6 . Representative chromosomes, and composite models of groups they represent (below), of four replication patterns for the i(Xq) chromosomes from Patient 4.Types A , B, and D show symmetric replication; type C shows asymmetric replication.

927 on the early-replicating X. The difference between their findings and ours may be caused by technical differences. Among the 60 metaphases analyzed, we observed considerable variation in the degree of acridine-orange fluorescence of the band patterns of different late-replicating X chromosomes, both with IdU and BrdU. Willard and Latt [36] observed a similar variation when using 33258 Hoechst fluorescence, and their results have been confirmed by others [17,35] ; together these investigators have examined about 1,000 cells. This variation was minimal in the early-replicating X chromosomes. lsochromosome or X-X Translocation

In an isochromosome, the band-pattern on one side of the centromere is an exact mirror image of the band-pattern on the other side. This could result from centric fission [6, 331, but if this were so, an isochrornosome of the long arms and an isochromosome of the short arms would be produced. This would lead to mosaicism, some daughter cells containing a long-arm isochromosome and others a short-arm isochromosome. To our knowledge, such mosaicism has never been recorded. In an isodicentric chromosome, exact mirror image band-pattern occurs on both sides of a hypothetical break-and-fusion point between two centromeres. This could result from either an isochromatid break in G 2 and

Replication in idic(X) and i(Xq)

457

consequent fusion of the sister chromatids at their breakpoints or a break in DNA at GI with subsequent fusion of sister chromatids during or after replication. In the next cell division, the distal, acentric fragments will be lost and normal separation of the centromere will lead to the formation of an isodicentric chromosome [31]. Possibly, an isochromosome really is an isodicentric chromosome in which available methods cannot delineate what may be a very narrow intercentromeric region. In fact, in some preparations of so-called isochromosomes, a double or bifid centromeric band can be seen [7, 271 . The i(Xq) of Patient 4 shows only one C-band, but it is larger than the other normal-sized C-bands on the idic(X) chromosomes shown in Figure 2. A structure that is indistinguishable morphologically from an idic(X) or i(Xq) could result from an X-X translocation, if breakpoints occur at identical sites on the homologues [5, 311 . In this case, gene duplication would be homologous and not identical. Because the occurrence of identical breakpoints on two homologous chromosomes is a highly improbable event, it is unlikely that more than a small minority of so-called idic(X) or i(Xq) chromosomes are X-X translocations. In all four patients studied in the present investigation, the abnormal X chromosome is an isochromosome as defined above. Computer-based videodensitometric measurements [9] confirmed that, in each case, the band-patterns on one side of the presumed break-andfusion point was the mirror image of the pattern on the other side [ 151 . In Patients 1, 2, and 3, the isochromosome was dicentric because in each case there were two C-bands, which indicates two regions of centromeric constitutive heterochromatin. In each case, only one primary constriction was clearly delineated; occasionally, as in Patient 1, a lesser indentation sometimes was seen at the site of the other C-band. It is likely that only the major constriction represented a functional centromere and that the other centromere was suppressed by either molecular inactivation or deletion [31]. If both centromeres were functional, there would be abnormal anaphase segregation resulting in broken idic(X) chromosomes and cells with two idic(X) chromosomes, but we observed none of these anomalies. In Patient 4, the abnormal X chromosome has the characteristics of an isochromosome of the long arm, i(Xq). Only one C-band and one constriction were observed, but this does not preclude the possibility that this chromosome has two very closely located centromeres with an exceedingly narrow intercentromeric region. In this case, it is not necessary to postulate inactivation of one of the centromeres. The closeness of the two centromeres reduces the possibility of twisting of the chromosome at the intercentromeric region. If twisting were possible, the two arms of a chromatid, in metaphase, would become attached through their centromeres to spindle fibers on opposite sides of the metaphase plate, which would lead to abnormal anaphase segregation [37]. As expected, we found no evidence of abnormal anaphase segregation in this case. In Patients 2 and 3 there was chromosomal mosaicism, the 46,X,idic(X) line being accompanied by a 45,X line. This is to be expected if the idic(X) arose by sister-chromatid fusion during mitosis soon after conception. In Patient 1, however, all but one of the 42 1 cells from her peripheral blood and tissues had an idic(X) karyotype; the one cell with 45,X is likely to be an artifact. If this is so, the idic(X) in this patient probably was derived from sister-chromatid fusion during parental gametogenesis and was present in the patient at conception. In Patient 4, only the blood was examined; all the metaphases showed the i(Xq) karyotype. Presumably, her i(Xq) also was present from conception, as in Patient 1.

458

Dewald, Spurbeck, and Gordon

Symmetric vs Asymmetric Replication

In each of our patients, autoradiography and IdU or BrdU showed that the abnormal X replicated later than its structurally normal homologue. The same observation was made by other investigators in respect to i(Xq) chromosomes [S, 1 6 , 2 5 , 2 6 , 2 9 ] and idic(X) chromosomes [24, 31, 341 . In our study, the i(Xq) chromosomes frequently had asymmetric replication patterns, which indicates that the timing of replication of corresponding bands on opposite sides of the breakpoint was asynchronous. Predominantly asymmetric replication patterns of i(Xq) chromosomes were observed by some previous investigators [2, 3, 111, whereas Muldal et a1 [25] and Ockey, Wennstrom, and de la Chapelle [26] found predominantly symmetric replication. Latt, Willard, and Gerald [24] found mostly symmetric patterns for i(Xq), but in 3 of 164 i(Xq) chromosomes from their three patients the patterns were asymmetric. In our case, the i(Xq) chromosomes had asymmetric patterns on 9 of the 15 chromosomes that we examined. In two of our patients with idic(X) chromosomes - Patients 1 and 2 with q28 and q24 breakpoints, respectively - the replication patterns almost always were symmetric; in only one metaphase (from Patient 2) was asymmetry observed. In Patient 3, in whom the breakpoint was on the short arm (pl l ) , an asymmetric replication pattern was observed in 17 of 24 metaphases. The cases of “X-X translocations” reported by Latt, Willard, and Gerald [24] and Dutrillaux [12] also could be regarded as idic(X) chromosomes according to our definition. These cases had predominantly symmetric replication. In both cases studied by Latt et a1 the breakpoints were on the long arms; in one, the breakpoint appears to be a t q24, similar to the breakpoint in our Patient 2. The finding of both symmetric and asymmetric replication in i(Xq) and idic(X) chromosomes may be caused by technical variations, differences in the pathogenesis of specific isochromosomes (chromatid fusion or X-X translocation), or differences in the genetic control of replication resulting from varied amounts of chromatin loss in different isochromosomes. The incorporation of a halogenated thymidine-base analogue, such as BrdU, is known to cause despiralization [38]. This effect is probably minor at 30 pg/ml but is readily apparent at much greater concentrations. Nevertheless, there may be some differential despiralization of certain corresponding bands on opposite arms of an isochromosome. Thus, a band may appear to have replicated while its counterpart may not. If there are differences in the degree of contraction of the two arms of an isochromosome, then it is possible that certain replicated bands may be overlooked because of relatively excessive contraction. In the case of idic(X) chromosomes, usually only one centromere region is constricted and presumably active. The regions of the chromosome nearer to the “active” centromere might be expected to be more contracted, thus distorting the replication pattern. Alternatively, there may be technical causes, such as uneven phase (wavelength) shifting of acridine orange during exposure to ultraviolet light, variations in photographic technique, or subjective differences in interpretation of replication patterns. In most cases, bands q22 and q24 were responsible for the asymmetric replication patterns. Inasmuch as these two bands are relatively late-replicating, they are best studied by using very short B-pulses. Therefore, any bias toward selecting metaphases showing considerable incorporation of IdU or BrdU might reveal primarily symmetric replication pat t ems. Baranovskaya et a1 [2] suggested that asymmetric replication patterns may occur on isochromosomes that arise by way of an X-X translocation. In contrast, symmetric replica-

Replication in idic(X) and i(Xq)

459

tion patterns may occur on isochromosomes that originate by some process of duplication (eg, chromatid fusion). They suggested that X-X translocations may exhibit asymmetric patterns simply because the initiation sites of DNA synthesis may not be genetically identical on homologous X chromosomes. In support of this, Latt, Willard, and Gerald [24] noted less concordance for the replication patterns of late-replicating X chromosomes in patients with three or more X chromosomes than for isochromosomes. Although this hypothesis is plausible, it seems unlikely, as discussed earlier, that X-X translocations could be responsible for so many apparent isochromosomes. Our study did not yield new information regarding the mechanism of the “inactivation” of one of the two centromeres on idic(X) chromosomes. In the patient with the idic(X)(pl 1) chromosome there were two types of asymmetric replication patterns, but if it were not for the recognition of a primary constriction, these models would have been regarded as identical. In one model, the overall replication seemed to be slightly ahead on the “short arm” side of the centromere. In the other model, the “long arm” side of the centromere was slightly advanced. It is not known whether this was because of random inactivation of one of the two centromeres early in ontogeny or simply because of random asymmetric replication. REFERENCES 1. Arrighi FE, Hsu TC: Localization of heterochromatin in human chromosomes. Cytogenetics 10:81-86, 1971. 2. Baranovskaya LI, Egolina NA, Zakharov AF, Tsvetkova TG: Isochromosome X in man: Different DNA replication patterns in the long arms. Hum Genet 33:55-60, 1976. 3. Benjush VA, Baranovskaya LI, Mirzajanz GG: Structural-functional characteristics of the iso-Xchromosome for the long arm in 45,X/46,X,i(Xq) patients. Genetica 9: 102-109, 1975. 4. Calderon D, Schnedl W: A comparison between quinacrine fluorescence banding and 3H-thymidine incorporation patterns in human chromosomes. Humangenetik 18:63-70, 1973. 5. Cohen MM, Rosenmann A, Hacham-Zadeh S, Dahan S: Dicentric X-isochromosome (Xqi dic) and pericentric inversion of No. 2 [inv(2) (p15 q21)] in a patient with gonadal dysgenesis. Clin Genet 8:ll-17, 1975. 6. Darlington CD: Misdivision and the genetics of the centromere. J Genet 37:341-364, 1939. 7. de la Chapelle A, Stenstrand K: Dicentric human X chromosomes. Hereditas 76:259-267, 1974. 8. de la Chapelle A, Wennstrom J , Hortling H, Ockey CH: Isochromosome-X in man. Part I. Hereditas 54~260-276, 1965. 9. Dewald GW, Robb RA, Gordon H: A computer-based videodensitometric system for studying banded human chromosomes illustrated by the analysis of the normal morphology of chromosome 18. Am J Hum Genet 29:37-51, 1977. 10. Distbche C, Hagemeijer A, Frederic J, Progneaux D: An abnormal large human chromosome identified as an end-to-end fusion of two X’s by combined results of the new banding techniques and microdensitometry. Clin Genet 3:388-395, 1972. 11. Dutrillaux B: Traitements discontinus par le BrdU et coloration par l’acridine orange: obtention d e marquages R, Q et interm6diaires. Chromosoma 52:261-273, 1975. 12. Dutrillaux B: Study of human X chromosomes with the 5-BrdU-acridine orange technique. Chromosomes Today 5:395--407, 1976. 13. Epplen JT, Siebers JW, Vogel W: DNA replication patterns of the human chromosomes from fibroblasts and amniotic fluid cells revealed by a Giemsa staining technique. Cytogenet Cell Genet 15:177-1 85, 1975. 14. Ganner E, Evans HJ: The relationship between patterns of DNA replication and of quinacrine fluorescence in the human chromosome complement. Chromosoma 35:326-341, 1971. 15. Gordon H, Dewald G, Dahl R, Soucheck R: Videodensitometric analysis of normal and abnormal X chromosomes by computer. In Vallet HL, Porter IH (eds): “Genetic Mechanisms of Sexual Development.” New York: Academic Press (In press).

460

Dewald, Spurbeck, and Gordon

16. Grumbach MM, Morishima A, Taylor JH: Human sex chromosome abnormalities in relation to DNA replication and heterochromatinization. R o c Natl Acad Sci USA 49:581-589, 1963. 17. Grzeschik K-H, Kim MA, Johannsmann R: Late replicating bands of human chromosomes demonstrated by fluorochrome and Giemsa staining. Humangenetik 29:41-59, 1975. 18. Kato H: Photoreactivation of sister chromatid exchanges induced by ultraviolet irradiation. Nature 249:552-553,1974. 19. Kim MA: Chromatidaustausch und Heterochromatinveranderungen menschlicher Chromosomen nach BUdR-Markierung: Nachweis mit Benzimidazolfluorochrom und Giemsafarbstoff. Humangene tik 25:179-188, 1974. 20. Kim MA, Johannsmann R, Grzeschik KH: Giemsa staining of the sites replicating DNA early in human lymphocyte chromosomes. Cytogenet Cell Genet 15:363-371, 1975. 21. Korenberg JR, Freedlender EF: Giemsa technique for the detection of sister chromatid exchanges. Chromosoma 48:355-360, 1974. 22. Latt SA: Microfluorometric detection of deoxyribonucleic acid replication in human metaphase chromosomes. Proc Natl Acad Sci USA 70:3395-3399, 1973. 23. Latt SA: Microfluorometric analysis of DNA replication in human X chromosomes. Exp Cell Res 86:412-415, 1974. 24. Latt SA, Willard HF, Gerald PS: BrdU-33258 Hoechst analysis of DNA replication in human lymphocytes with supernumerary or structurally abnormal X chromosomes. Chromosoma 57:135-153, 1976. 25. Muldal S, Gilbert CW, Lajtha LG, Lindsten J, Rowley J, Fraccaro M: Tritiated thymidine incorporation in an isochromosome for the long arm of the X chromosome in man. Lancet 1:861-863, 1963. 26. Ockey CH, Wennstram J, de la Chapelle A: Isochromosome-X in man. Part 11. Hereditas 54:277292, 1965. 27. Palmer CG, Reichmann A: Chromosomal and clinical findings in 110 females with Turner syndrome. Hum Genet 35:35-49, 1976. 28. Perry P, Wolff S: New Giemsa method for the differential staining of sister chromatids. Nature 251:156-158,1974. 29. Priest JH, Blackston RD, Au K-S, Ray SL: Differences in human X isochromosomes. J Med Genet 12:378-389,1975. 30. Seabright M : A rapid banding technique for human chromosomes (Letter to the editor). Lancet 2:971-972, 1971. 31. Therman E, Sarto GE, Patau K: Apparently isodicentric but functionally monocentric X chromosome in man. Am J Hum Genet 26:83-92,1974. 32. Titus JL: Chromosome study of autopsy tissues. In Ludwig J (ed): “Current Methodsof Autopsy Practice.” Philadelphia: WB Saunders, 1972, pp 217-219. 33. Todd NB: A theory of karyotypic fissioning, genetic potentiation and eutherian evolution. Mammal Chromosomes Newsletter 8:268-279, 1967. 34. Van den Berghe H, Fryns JP, Soyez C: X/X translocation in a patient with Turner’s syndrome. Humangenetik 20:377-380, 1973. 35. Willard HF: Tissue-specific heterogeneity in DNA replication patterns of human X chromosomes. Chromosoma 61:61-73, 1977. 36. Willard HF, Latt SA: Analysis of deoxyribonucleic acid replication in human X chromosomes by fluorescence microscopy. Am J Hum Genet 28:213-227, 1976. 37. Ying KL, Ives EJ: Mitotic behavior of a human dicentric Y chromosome. Cytogenetics 10:208218,1971. 38. Zakharov AF, Egolina NA: Differential spiralization along mammalian mitotic chromosomes. I . BUdR-revealed differentiation in Chinese hamster chromosomes. Chromosoma 38: 341 -365, 1972.

Edited by Holger Hoehn