American Journal of Medical Genetics 41:78-82 (1991)
Monozygotic Twins Discordant for Ullrich-Turner Syndrome Paul B. Kaplowitz, Joann Bodurtha, Judith Brown, and J. Edward Spence Departments of Pediatrics (P.B.K., Jo.B., J.E.S.) and Human Genetics (Jo.B., Ju.B., J.E.S.), Children’s Medical Center-Medical College of Virginia, Virginia Commonwealth University, Richmond We describe 9-year-old twin girls who were thought to be monozygotic but who differed greatly in physical appearance and growth pattern. One twin had Ullrich-arner syndrome (UTS),45,X/46,XX mosaicism in peripheral blood, and only 45,X cells in skin fibroblasts. The phenotypically normal twin also had 45,X/46,XX mosaicism in blood but only 46,XX cells in cultured fibroblasts. Analysis of DNA marker patterns in blood lymphocytes and in skin fibroblasts confirmed monozygosity with a probability of 99.97%. This case is compared with other reported cases of discordance for UTS in twins. It is concluded that essentially all of the differences between the two twins can be explained by loss of an X chromosome early in embryogenesis with complete separation of 45,X and 46,XX cell lineages at the time of the twinning event. The presence of mosaicism in the peripheral blood of both twins is presumably due to anastomoses between the placentae resulting in a mixture of the two cell populations in the hematopoietic tissue.
KEY
WORDS: Discordant monozygotic twins, mosaicism, anastomoses
INTRODUCTION Since the first cases of monozygotic twins discordant for Ullrich-Turner syndrome (UTS) were reported 30 years ago [Turpin et al., 19611, there have been several additional reports of this rare and puzzling phenomenon [Mikkelsen e t al., 1963; Potter and Taitz, 1972; Weiss et al., 1982; Al-Awadi et al., 1983; Uchida et al., 19831. These cases often present a diagnostic challenge because the results of peripheral blood karyotypes may not Received for publication September 14, 1990; revision received February 5, 1991. Address reprint requests to Paul Kaplowitz, M.D., W.D., Medical College of Virginia, Box 140 MCV Station, Richmond, Va 23298.
0 1991 Wiley-Liss, Inc.
correlate with their different phenotypes. We studied 91h-year-old girls who had been considered monozygotic (MZ) twins based on blood group analysis but who had obvious differences in their growth patterns and physical appearance. Blood karyotype was 45,X/46,XX in both, but analysis of skin fibroblasts showed only 45,X cells in the affected twin and only 46,XX cells in the normal twin. Analysis of DNA from the fibroblasts confirmed monozygosity with a high degree of certainty. Examination of such patients is of interest not only because of the different genetic mechanisms by which such discordance is produced, but because it provides a rare chance to observe the effects of X chromosome monosomy alone on the growth and physical development of two otherwise genetically identical individuals. CLINICAL REPORT The twins were born to a 22-year-old gravida 3, abortus 2 white woman after a 38-week-gestation pregnancy complicated only by monthly spotting for the first 5 months. Twin A presented vertex, weighed 2,570 g, and had Apgar scores of 9 and 10. Twin B was in a transverse lie and required extraction and resuscitation but no intubation. Her weight was 2,180 g, Apgar scores were 6 and 10, and a single umbilical artery was noted; the only neonatal problem was mild jaundice treated with phototherapy. Although neither twin had puffy hands or feet at birth, twin B developed puffy hands at age 4 months which persisted until approximately age 7 years. The placenta was monochorionic and diamniotic, and weighed 784 g. The pathology report did not indicate vascular anastomoses on the placenta; however, an experienced genetic observer had noted a superficial anastomosis. Fingertip patterns were identical on 9 of 10 digits. The calculated probability of monozygosity based on these findings and 11blood group markers was 0.975. Twin B’s physical and motor development lagged behind that of twin A, despite the fact that twin A was placed in hip abduction splints for a dislocated hip from 4-7 months of age. By age 11months twin A was walking but twin B was not sitting up. A neurologist saw twin B once, diagnosed mild cerebral palsy, and referred her to a n infant stimulation program for therapy. Twin B suffered multiple episodes of otitis media from age 25. Her verbal skills were good and she was reading by
Twins Discordant for Ullrich-Turner Syndrome the first grade, but she had trouble learning math, and because of her tendency to be easily distracted, she was enrolled in learning disability classes. At age 8V2, twin B was seen by an orthodontist because of malocclusion. Her tooth eruption pattern was found to be out of sequence and her teeth were crowded because of undergrowth of the maxilla and mandible. A transpalatal appliance and a mandibular holding arch were prescribed. At age g3/i2 years, twin B was referred for evaluation of her severe short stature. The mother had been told that the cerebral palsy was the cause of the child's slower growth and development but she decided to seek another opinion when the 5-year-old brother caught up with twin B in height. Her height was 111.3 cm ( - 3.3 S.D.) and her weight was 18.0 kg ( 2 5 4 0 t h centile for height). There was no webbing of the neck or low posterior hairline, but the palate was high-arched and there was a shield chest. Cubitus valgus was present (R>L) and her fingernails were upturned. Neurologically she was normal and there was no evidence of cerebral palsy. Twin A, whose height was 127.3 cm ( - 1.0 S.D.), had a normal exam except for slight cubitus valgus on the right. The differences in their physical appearance is shown in Figure 1. Peripheral blood chromosome analysis showed 45,X/46,XX mosaicism in both twins (Table I). Twin A had normal gonadotropin levels (follicle-stimulating hormone [FSHI, 3.8 U/L; luteinizing hormone [LHI, 1.6 U/L), whereas twin 3 had an FSH which was somewhat elevated for a prepubertal girl (10.4 UIL; normal, < 5 ) and a normal LH (3.2 UIL). Ultrasonography on twin B showed normal kidneys, but no ovaries could be identified. Skin biopsies to establish fibroblast cultures were obtained from both arms of each twin and 20 cells were
79
TABLE I. Number of Cells Counted With Different Chromosome ComrJlements Random loss Twin A Lymphocytes Fibroblasts Twin B Lymphocytes Fibroblasts
45,x
46,XX -
-
9
3
0
41 37
4
6 39
40 0
1
analyzed from each cell line. Since the findings from the left and right arm cultures were very similar, the results of the 40 cells counted were combined (Table I). Cells with a 45,X chromosome complement were present only in cultures from twin B. The results of quantitative dermatoglyphic analysis on the twins is shown in Table 11. Twin B had a total ridge count which was 31 greater than that of twin A; a difference of 28 or greater is observed in MZ twins only 6% of the time [Holt, 19681. To verify that these twins were MZ, DNA was extracted from the leukocytes of both girls. For all of the 8 DNA markers tested, the twins had identical patterns (one of the gels is shown in Fig. 2). Using available heterozygosity values for the DNA markers, without testing parental DNA samples, a pair of twins has a t least a 99.97% likelihood of being MZ if they are identical for all 8 markers. Because each twin had mosaicism in the peripheral blood which, in theory, could have resulted from placental anastomoses between DZ twins and could still have resulted in identical DNA patterns for the lymphocytes, we took the further step of analyzing the DNA from fibroblast cultures of each twin. The DNA marker pattern from the fibroblast DNA samples was identical in each twin and identical to the pattern from the lymphocyte DNA samples, and there was no evidence of faint or anomalous DNA marker hybridization that might indicate mosaicism in either twin. To examine the effect that lack of one X chromosome had on somatic growth, length and height measurements on both twins at well-child visits were compared (Fig. 3). Twin A was slightly larger than twin B at birth, and this difference persisted and widened slowly during the first 2 years of life. Between ages 4 and 9, when the growth rate is normally relatively stable, twin A grew TABLE 11. Quantitative Dermatoglyphic Analysis Twin A Digital Ridge Digit pattern count RT UL 16 R1 RL 8 R2 UL 12 R3 UL 6 R4 UL 14 LT UL 14 L1 RL 12 L2 UL 10 L3 UL 9 L4 UL 11 Total ridge count 112 ~~
~~
Fig. 1. Twin A (left) and twin B (right) at age 101/iz years.
Twin B Digital Ridge pattern count UL 18 RL 9 UL 15 UL 14 UL 20 UL 19 UL 11 UL 13 UL 8 16 UL 143
80
Kaplowitz et al. 180
-
160 I50 -
170
-
140
-
130
-
E
2 120k
I 110
'3
-
-
"1
g loo-
70
Fig. 2. DNA marker patterns (autoradiographs)for leukocyte DNA from both twins. High molecular weight DNA was prepared using published methods [Spence et al., 19871. Each DNA sample (5 Fg) was digested with the restriction endonuclease PstI (Bethesda Research Laboratories) for 16 hours at 37°C.Electrophoresis was performed with 0.8% agarose gels and DNA was transferred to nylon membranes (Schleicher and Schuell, Keene, NH) by standard methods [Southern, 1975;Maniatis et al., 19821. Filters were prehybridized and hybridized at 67°C using the following series of highly polymorphic DNA markers: (group 1)D2544, D17S30, D6S44, D15S24; (group 2) DlS57, D4S125; (group 3) INS, D14S13. Each of the three groups of DNA marker mixtures was radiolabelled using the random-primer labelling procedure [Fineberg and Vogelstein, 19831 and hybridized in sequence, with alkali-stripping of the previous radiolabelled marker mixture between each hybridization. Washing of the hybridized filters was performed using previously described conditions [Maniatiset al., 19821, and autoradiography was performed at - 70°C. This autoradiograph shows the pattern for the group 1 markers, with twin A and twin B in lanes 1 and 2 respectively, and a different twin pair in lanes 3 and 4 for comparison.
29 cm (5.8 cmlyear) and twin B grew 21 cm (4.2 cmiyear, or 72% of the velocity of twin A). Between age 1 and 4 years, twin B's height remained 92-93% of that of twin A, but this decreased to 89%a t age 7 and to 87.5%at age 9%. Twin B was started on daily injections of biosynthetic growth hormone a t age 97/12 and after 9 months, her growth rate had improved to 8.2 cmtyear.
DISCUSSION The occurrence of UTS in one or both of presumably MZ twins has been reported several times since 1961 and the major manifestations of these unusual cases have been reviewed [Weiss et al., 1982; Al-Awadi et al., 19831. Most of them fall into three categories: 1) those in which both twins are affected; 2 ) those in which there is a n apparently normal male and a n affected female; and 3) those in which there is one apparently normal female and one affected female. We will focus this discussion on the third category, which includes the present case and several other well-documented cases. Mikkelsen et al. [ 19631 described 31h-year-old twins who were a t the 10th centile in height but differed in
6o
50
TWIN A c . TWIN E
wI I
2 3 4
5 6 7 8 9 10 II 12 13 14 15 16 A G E (years)
Fig. 3. Growth curves of twins A and B from birth to age 10 years. At age 91h, twin B began receiving biosynthetic growth hormone a t a dose of 0.05 mg/kg 6 times per week.
that only one had webbing of the neck and other findings of UTS. In both twins, 113 of the lymphocytes were 45,X and 213 46,XX; in skin fibroblasts, the phenotypically normal twin had 45%of cells with 45,X and the affected twin had 80% of cells with 4 5 3 . The twins had a 98% chance of being identical based on blood group analysis. The authors could not readily explain the phenotypic differences observed in these twins with 45,X/46,XX mosaicism in both blood and somatic tissues. Potter and Taitz [ 19721 reported one-year-old twins who were identical in length and weight but differed in that one had webbing of the neck and edema of hands and feet. Monozygosity was suggested by detailed blood group analysis only. In blood the affected twin was 70% 45,X/30% 46,XX and the normal twin was 87% 45,X/13%46,XX. However, skin fibroblasts showed only 45,X cells in the affected sister and only 46,XX cells in the normal twin. The loss of a n X chromosome from one embryo was thought to have occurred almost immediately after the twinning event, and i t was proposed that the mosaicism in the blood of both girls resulted from a placental anastomosis between the two twins, allowing reciprocal transplant of hematopoietic precursor cells and a permanent mixture of the two cell types. Uchida et al. [1983] described twins discordant for UTS in which the affected twin died soon after birth with multiple congenital abnormalities which included generalized lymphedema and hypoplastic lungs. At au-
Twins Discordant for Ullrich-Turner Syndrome topsy, the left ovary was histologically normal but the right ovary showed virtually no germ cells and no primordial follicles. Blood karyotype was 3% 45,X/97% 46,XX in the normal twin and 2% 45,X/98%46,XX in the affected twin, while skin fibroblast cultures showed only 46,XX cells in the normal twin and only 45,X cells in the affected twin. The authors also proposed anastomoses across a common placenta as the explanation for the mosaicism in the peripheral blood. The twins reported by Weiss et al. [19821 had been followed for over 13 years, with the affected twin having multiple signs of UTS as well as the characteristic growth pattern. At age 6, the affected twin was 91% of the height of the normal one; at age 129/12, when the normal twin was at Tanner stage 111of puberty, height was 87% of that of the normal twin. These differences were very similar to those found in our patients. In addition, the affected twin had a smaller facial profile than the normal twin and was relatively retrognathic, similar to our twin B and to previous observations made in patients with UTS [Fillipson et al., 19651. Blood karyotype was 46,XX on the normal twin and 26% 4 5 3 7 4 % 46,XX on the affected twin. Permission to do fibroblast studies could not be obtained, so the distribution of 45,X and 46,XX cell lines in somatic tissues is unknown. The probability of monozygosity, based on 5 different methods but not including DNA analysis, was estimated to be >99.95%. Although the phenomonon of non-discordance for UTS in apparently identical twins has been described in over 10 reports, each case is unique in terms of the extent of phenotypic differences, the distribution of abnormal cell lines in blood and somatic tissues, and their relative proportions. One major variable responsible for the wide spectrum of findings is likely to be the timing of the loss of the X chromosome relative to the twinning event. If the loss of an X chromosome occurred during the first post-fertilization division and the twinning event occurred simultaneously, the result would be one twin with only 46,XX cells in the somatic tissues, and the other with only 45,X cells, which is what was found in our twins. The main difficulty with this explanation is that twinning occurs most commonly at the blastocyst stage (4-8 days after fertilization and before implantation), resulting in a single monochorionic diamniotic placenta as was observed is this case. Twinning occurs less frequently a t the morula stage ( < 3 days, 1 3 2 cells), which results in separate or fused dichorionic diamniotic placentae [Corner, 19551. If the X chromosome loss occurred at the two-cell stage or the morula stage and the twinning event occurred a t the blastocyst stage, the result would most likely be X/XX mosaicism in both embryos [e.g., Mikkelsen et al., 19631, assuming that the 45,X and 46,XX cell lineages were well-mixed in the cell mass before the twinning event occurred. However, it is conceivable that the different cell lines resulting from division of the progenitor 45,X and 46,XX cells selfaggregate, and that the twinning event could result in complete separation of the two cell lines, resulting in one 45,X and one 46,XX embryo, or partial separation leading to one mosaic embryo and one entirely normal embryo. Burn et al. [19861, in trying to explain the occur-
81
rence of Duchenne muscular dystrophy in one of MZ twin girls, suggested that X inactivation in females is followed by preferential aggregation of cells bearing the same active X within the inner cell mass prior to the twinning event. Nance [19901 proposes an additional mechanism, in that if the number of precursor cells in each embryo is unequal, the chance of seeing extreme X-inactivation profiles is greater in the twin arising from the fewest cells. It is therefore plausible that cell lines differing in the number of X chromosomes rather than the parental origin of the active X could also selfaggregate prior to the twinning event, or that the different cell lines might segregate almost completely by chance during an unequal twinning event. Finally, if the X chromosome loss occurred after the twinning event, there would be one completely normal embryo and one with X/XX mosaicism, the extent of which would depend on how many cells were present in that embryo at the time of the loss of an X chromosome from one cell lineage. The large variability in peripheral blood karyotype findings in twins discordant for UTS syndrome is best explained by the frequent presence of anastomoses between the two twin circulations when there is a single placenta, and perhaps differences in the growth rates of hematopoietic precursors with 46,XX or 45,X karyotypes. It is unlikely that mosaicism in the peripheral blood has any important clinical effect on either the affected twin or the normal twin. However, it must be noted that any determination of zygosity which relies heavily on blood groups or other methods using blood cells has the potential for error, since even DZ twins with fused placentae may rarely have anastomoses between the fetal blood supplies [Dunsford et al., 1953; Strong and Corney, 19671 and could be indistinguishable based on blood cell markers. F'rom the information available, we cannot be certain of the timing of the events by which both twinning and loss of an X chromosome occurred in our patients; however, the clinical and cytogenetic findings are consistent with a complete separation of 45,X cells and 46,XX cells during the twinning process, with the mosaicism in the peripheral lymphocytes being due to twin-to-twin transfusion. Monozygosity was established with a high degree of certainty with DNA typing on fibroblasts as well as lymphocytes. The twins had identical finger patterns in 9 of 10 digits, but twin B had a total finger ridge count which was 31 greater than twin A. Although a difference of this magnitude is not often seen in MZ twins, it is consistent with the observation that the average ridge count in UTS patients (169.3 2 41.6) is 42 greater than is found in control females (127.0 2 52.3) [Holt, 19681. Since the total ridge count is determined almost entirely by genetic factors (r = 0.95 for MZ twins, 0.46 for full sibs, and 0.23 for half-sibs [Phelan et al., 198111, the difference of 31 observed for the MZ twins described here may be a more accurate reflection of the effect of monosomy for the X chromosome on the dermal ridge count than the figure derived by comparing the mean value for UTS patients and control females. It was initially suspected, based on the finding of slight cubitus valgus on the right side only, that the phenotypically normal twin A might be a true mosaic for
82
Kaplowitz et al.
UTS. To increase the chances of finding the abnormal cell line, separate skin biopsies were taken from both the right and the left arms, but neither contained 45,X cells. It is conceivable that biopsies of multiple other tissues might have demonstrated a small number of 45,X cells in the normal twin, but it is unlikely that any 45,X cells which might be present have affected her growth, physical findings (exceptpossibly for the cubitus valgus), and ovarian function. Although the affected twin had evidence of “cerebral palsy” during her first year of life, it appears to have been mild since the motor deficits resolved spontaneously with time, and should not have affected significantly her growth and physical appearance. Therefore, it should be valid to compare the growth patterns and physical characteristics of these twins and to attribute the differences between them solely to the presence of monosomy for the X chromosome in the tissues of the affected twin.
REFERENCES Al-Awadi SA, Cuschieri A, Farag TI, Naguib K, Teebi AS, Al-Othman SA, Bahig AH (1983): Ullrich-Thrner syndrome in monozygotic twins. Am J Med Genet 15:537-542. Burn J, Povey S, Boyd Y, Munro EA, West L, Harper K, Thomas D (1986):Duchenne muscular dystrophy in one of monozygotic twin girls. J Med Genet 23:494-500. Corner GW [ 1955):Observed embryology of human single-ovum twins and other multiple births. Am J Obstet Gynecol 70:933-950. Dunsford I, Bowley CC, Hutchinson AM, Thompson JS, Sanger R, Race RR (1953):A human blood group chimera. Br Med J II:81. Fillipson R, Lindsten J , Almquist S (1965):Time of eruption of permanent teeth, cephalometric and tooth measurement and sulfation
factor activity in 45 patients with Turner syndrome and different types of X chromosome aberrations, Acta Endocrinol 48:91-113. Fineberg AP, Vogelstein B (1983):A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal Biochem 1325-13. Holt SB (1968): “The Genetics of Dermal Ridges.” Springfield, IL: Charles C Thomas Publishers, pp. 80, 158. Maniatis T, fiitsch EF, Sambrook J (1982): “Molecular cloning: A Laboratory Manual.” New York: Cold Spring Harbor Laboratory. Mikkelsen M, Froland A, Ellebjerg J (1963):XO/XX mosaicism in a pair of presumably monozygotic twins with different phenotypes. Cytogenetics 2%-98. Nance WE (1990):Invited editorial: do twin Lyons have larger spots? Am J Hum Genet 46:646-648. Phelan MC, Nance WE, Corey LA (1981):Determinants ofridge counts in MZ twin sibships. Ann Genet Med Gamellol 30:59-66. Potter AM, Taitz LS (1972):Turner syndrome in one of monozygotic twins with mosaicism. Acta Paediatr Scand 61:473-476. Southern EM (1975):Detection of specific sequences among DNA fragments separated by gel electrophoresis. J Mol Biol 98:503-517. Spence JE, Buffone GJ, Rosenbloom CL, Fernbach SD, Curry MR, Carpenter RJ, Ledbetter DH, O’Brien WE, Beaudet AL (1987): Prenatal diagnosis of cystic fibrosis using linked DNA markers and microvillar intestinal enzyme analysis. Hum Genet 76:5-10. Strong SJ,Corney G (1967):“The Placenta in Twin Pregnancy.” New York Pergamon Press, pp 78-89. Turpin R, Lejeune J, Lafourcade J, Chigot PL, Salmon C (1961):Presomption de monozygotisme au depit d’un dimorphisme sexuel: Sujet masculin XY et sujet haplo-X. CR Acad Sci [Dl (Paris) 252:2945-2946. Uchida IA, deSa DJ, Whelan DT (1983): 45,W46,XX mosaicism in discordant monozygotic twins. Pediatrics 71:413-417. Weiss E, Loevy H, Saunders A, Pruzansky S,Rosenthal I(1982):Monozygotic twins discordant for Ullrich-Turner syndrome. Am J Med Genet 13:389-399.
Monozygotic Twins Discordant for Ullrich-Turner Syndrome Paul B. Kaplowitz, Joann Bodurtha, Judith Brown, and J. Edward Spence Departments of Pediatrics (P.B.K., Jo.B., J.E.S.) and Human Genetics (Jo.B., Ju.B., J.E.S.), Children’s Medical Center-Medical College of Virginia, Virginia Commonwealth University, Richmond We describe 9-year-old twin girls who were thought to be monozygotic but who differed greatly in physical appearance and growth pattern. One twin had Ullrich-arner syndrome (UTS),45,X/46,XX mosaicism in peripheral blood, and only 45,X cells in skin fibroblasts. The phenotypically normal twin also had 45,X/46,XX mosaicism in blood but only 46,XX cells in cultured fibroblasts. Analysis of DNA marker patterns in blood lymphocytes and in skin fibroblasts confirmed monozygosity with a probability of 99.97%. This case is compared with other reported cases of discordance for UTS in twins. It is concluded that essentially all of the differences between the two twins can be explained by loss of an X chromosome early in embryogenesis with complete separation of 45,X and 46,XX cell lineages at the time of the twinning event. The presence of mosaicism in the peripheral blood of both twins is presumably due to anastomoses between the placentae resulting in a mixture of the two cell populations in the hematopoietic tissue.
KEY
WORDS: Discordant monozygotic twins, mosaicism, anastomoses
INTRODUCTION Since the first cases of monozygotic twins discordant for Ullrich-Turner syndrome (UTS) were reported 30 years ago [Turpin et al., 19611, there have been several additional reports of this rare and puzzling phenomenon [Mikkelsen e t al., 1963; Potter and Taitz, 1972; Weiss et al., 1982; Al-Awadi et al., 1983; Uchida et al., 19831. These cases often present a diagnostic challenge because the results of peripheral blood karyotypes may not Received for publication September 14, 1990; revision received February 5, 1991. Address reprint requests to Paul Kaplowitz, M.D., W.D., Medical College of Virginia, Box 140 MCV Station, Richmond, Va 23298.
0 1991 Wiley-Liss, Inc.
correlate with their different phenotypes. We studied 91h-year-old girls who had been considered monozygotic (MZ) twins based on blood group analysis but who had obvious differences in their growth patterns and physical appearance. Blood karyotype was 45,X/46,XX in both, but analysis of skin fibroblasts showed only 45,X cells in the affected twin and only 46,XX cells in the normal twin. Analysis of DNA from the fibroblasts confirmed monozygosity with a high degree of certainty. Examination of such patients is of interest not only because of the different genetic mechanisms by which such discordance is produced, but because it provides a rare chance to observe the effects of X chromosome monosomy alone on the growth and physical development of two otherwise genetically identical individuals. CLINICAL REPORT The twins were born to a 22-year-old gravida 3, abortus 2 white woman after a 38-week-gestation pregnancy complicated only by monthly spotting for the first 5 months. Twin A presented vertex, weighed 2,570 g, and had Apgar scores of 9 and 10. Twin B was in a transverse lie and required extraction and resuscitation but no intubation. Her weight was 2,180 g, Apgar scores were 6 and 10, and a single umbilical artery was noted; the only neonatal problem was mild jaundice treated with phototherapy. Although neither twin had puffy hands or feet at birth, twin B developed puffy hands at age 4 months which persisted until approximately age 7 years. The placenta was monochorionic and diamniotic, and weighed 784 g. The pathology report did not indicate vascular anastomoses on the placenta; however, an experienced genetic observer had noted a superficial anastomosis. Fingertip patterns were identical on 9 of 10 digits. The calculated probability of monozygosity based on these findings and 11blood group markers was 0.975. Twin B’s physical and motor development lagged behind that of twin A, despite the fact that twin A was placed in hip abduction splints for a dislocated hip from 4-7 months of age. By age 11months twin A was walking but twin B was not sitting up. A neurologist saw twin B once, diagnosed mild cerebral palsy, and referred her to a n infant stimulation program for therapy. Twin B suffered multiple episodes of otitis media from age 25. Her verbal skills were good and she was reading by
Twins Discordant for Ullrich-Turner Syndrome the first grade, but she had trouble learning math, and because of her tendency to be easily distracted, she was enrolled in learning disability classes. At age 8V2, twin B was seen by an orthodontist because of malocclusion. Her tooth eruption pattern was found to be out of sequence and her teeth were crowded because of undergrowth of the maxilla and mandible. A transpalatal appliance and a mandibular holding arch were prescribed. At age g3/i2 years, twin B was referred for evaluation of her severe short stature. The mother had been told that the cerebral palsy was the cause of the child's slower growth and development but she decided to seek another opinion when the 5-year-old brother caught up with twin B in height. Her height was 111.3 cm ( - 3.3 S.D.) and her weight was 18.0 kg ( 2 5 4 0 t h centile for height). There was no webbing of the neck or low posterior hairline, but the palate was high-arched and there was a shield chest. Cubitus valgus was present (R>L) and her fingernails were upturned. Neurologically she was normal and there was no evidence of cerebral palsy. Twin A, whose height was 127.3 cm ( - 1.0 S.D.), had a normal exam except for slight cubitus valgus on the right. The differences in their physical appearance is shown in Figure 1. Peripheral blood chromosome analysis showed 45,X/46,XX mosaicism in both twins (Table I). Twin A had normal gonadotropin levels (follicle-stimulating hormone [FSHI, 3.8 U/L; luteinizing hormone [LHI, 1.6 U/L), whereas twin 3 had an FSH which was somewhat elevated for a prepubertal girl (10.4 UIL; normal, < 5 ) and a normal LH (3.2 UIL). Ultrasonography on twin B showed normal kidneys, but no ovaries could be identified. Skin biopsies to establish fibroblast cultures were obtained from both arms of each twin and 20 cells were
79
TABLE I. Number of Cells Counted With Different Chromosome ComrJlements Random loss Twin A Lymphocytes Fibroblasts Twin B Lymphocytes Fibroblasts
45,x
46,XX -
-
9
3
0
41 37
4
6 39
40 0
1
analyzed from each cell line. Since the findings from the left and right arm cultures were very similar, the results of the 40 cells counted were combined (Table I). Cells with a 45,X chromosome complement were present only in cultures from twin B. The results of quantitative dermatoglyphic analysis on the twins is shown in Table 11. Twin B had a total ridge count which was 31 greater than that of twin A; a difference of 28 or greater is observed in MZ twins only 6% of the time [Holt, 19681. To verify that these twins were MZ, DNA was extracted from the leukocytes of both girls. For all of the 8 DNA markers tested, the twins had identical patterns (one of the gels is shown in Fig. 2). Using available heterozygosity values for the DNA markers, without testing parental DNA samples, a pair of twins has a t least a 99.97% likelihood of being MZ if they are identical for all 8 markers. Because each twin had mosaicism in the peripheral blood which, in theory, could have resulted from placental anastomoses between DZ twins and could still have resulted in identical DNA patterns for the lymphocytes, we took the further step of analyzing the DNA from fibroblast cultures of each twin. The DNA marker pattern from the fibroblast DNA samples was identical in each twin and identical to the pattern from the lymphocyte DNA samples, and there was no evidence of faint or anomalous DNA marker hybridization that might indicate mosaicism in either twin. To examine the effect that lack of one X chromosome had on somatic growth, length and height measurements on both twins at well-child visits were compared (Fig. 3). Twin A was slightly larger than twin B at birth, and this difference persisted and widened slowly during the first 2 years of life. Between ages 4 and 9, when the growth rate is normally relatively stable, twin A grew TABLE 11. Quantitative Dermatoglyphic Analysis Twin A Digital Ridge Digit pattern count RT UL 16 R1 RL 8 R2 UL 12 R3 UL 6 R4 UL 14 LT UL 14 L1 RL 12 L2 UL 10 L3 UL 9 L4 UL 11 Total ridge count 112 ~~
~~
Fig. 1. Twin A (left) and twin B (right) at age 101/iz years.
Twin B Digital Ridge pattern count UL 18 RL 9 UL 15 UL 14 UL 20 UL 19 UL 11 UL 13 UL 8 16 UL 143
80
Kaplowitz et al. 180
-
160 I50 -
170
-
140
-
130
-
E
2 120k
I 110
'3
-
-
"1
g loo-
70
Fig. 2. DNA marker patterns (autoradiographs)for leukocyte DNA from both twins. High molecular weight DNA was prepared using published methods [Spence et al., 19871. Each DNA sample (5 Fg) was digested with the restriction endonuclease PstI (Bethesda Research Laboratories) for 16 hours at 37°C.Electrophoresis was performed with 0.8% agarose gels and DNA was transferred to nylon membranes (Schleicher and Schuell, Keene, NH) by standard methods [Southern, 1975;Maniatis et al., 19821. Filters were prehybridized and hybridized at 67°C using the following series of highly polymorphic DNA markers: (group 1)D2544, D17S30, D6S44, D15S24; (group 2) DlS57, D4S125; (group 3) INS, D14S13. Each of the three groups of DNA marker mixtures was radiolabelled using the random-primer labelling procedure [Fineberg and Vogelstein, 19831 and hybridized in sequence, with alkali-stripping of the previous radiolabelled marker mixture between each hybridization. Washing of the hybridized filters was performed using previously described conditions [Maniatiset al., 19821, and autoradiography was performed at - 70°C. This autoradiograph shows the pattern for the group 1 markers, with twin A and twin B in lanes 1 and 2 respectively, and a different twin pair in lanes 3 and 4 for comparison.
29 cm (5.8 cmlyear) and twin B grew 21 cm (4.2 cmiyear, or 72% of the velocity of twin A). Between age 1 and 4 years, twin B's height remained 92-93% of that of twin A, but this decreased to 89%a t age 7 and to 87.5%at age 9%. Twin B was started on daily injections of biosynthetic growth hormone a t age 97/12 and after 9 months, her growth rate had improved to 8.2 cmtyear.
DISCUSSION The occurrence of UTS in one or both of presumably MZ twins has been reported several times since 1961 and the major manifestations of these unusual cases have been reviewed [Weiss et al., 1982; Al-Awadi et al., 19831. Most of them fall into three categories: 1) those in which both twins are affected; 2 ) those in which there is a n apparently normal male and a n affected female; and 3) those in which there is one apparently normal female and one affected female. We will focus this discussion on the third category, which includes the present case and several other well-documented cases. Mikkelsen et al. [ 19631 described 31h-year-old twins who were a t the 10th centile in height but differed in
6o
50
TWIN A c . TWIN E
wI I
2 3 4
5 6 7 8 9 10 II 12 13 14 15 16 A G E (years)
Fig. 3. Growth curves of twins A and B from birth to age 10 years. At age 91h, twin B began receiving biosynthetic growth hormone a t a dose of 0.05 mg/kg 6 times per week.
that only one had webbing of the neck and other findings of UTS. In both twins, 113 of the lymphocytes were 45,X and 213 46,XX; in skin fibroblasts, the phenotypically normal twin had 45%of cells with 45,X and the affected twin had 80% of cells with 4 5 3 . The twins had a 98% chance of being identical based on blood group analysis. The authors could not readily explain the phenotypic differences observed in these twins with 45,X/46,XX mosaicism in both blood and somatic tissues. Potter and Taitz [ 19721 reported one-year-old twins who were identical in length and weight but differed in that one had webbing of the neck and edema of hands and feet. Monozygosity was suggested by detailed blood group analysis only. In blood the affected twin was 70% 45,X/30% 46,XX and the normal twin was 87% 45,X/13%46,XX. However, skin fibroblasts showed only 45,X cells in the affected sister and only 46,XX cells in the normal twin. The loss of a n X chromosome from one embryo was thought to have occurred almost immediately after the twinning event, and i t was proposed that the mosaicism in the blood of both girls resulted from a placental anastomosis between the two twins, allowing reciprocal transplant of hematopoietic precursor cells and a permanent mixture of the two cell types. Uchida et al. [1983] described twins discordant for UTS in which the affected twin died soon after birth with multiple congenital abnormalities which included generalized lymphedema and hypoplastic lungs. At au-
Twins Discordant for Ullrich-Turner Syndrome topsy, the left ovary was histologically normal but the right ovary showed virtually no germ cells and no primordial follicles. Blood karyotype was 3% 45,X/97% 46,XX in the normal twin and 2% 45,X/98%46,XX in the affected twin, while skin fibroblast cultures showed only 46,XX cells in the normal twin and only 45,X cells in the affected twin. The authors also proposed anastomoses across a common placenta as the explanation for the mosaicism in the peripheral blood. The twins reported by Weiss et al. [19821 had been followed for over 13 years, with the affected twin having multiple signs of UTS as well as the characteristic growth pattern. At age 6, the affected twin was 91% of the height of the normal one; at age 129/12, when the normal twin was at Tanner stage 111of puberty, height was 87% of that of the normal twin. These differences were very similar to those found in our patients. In addition, the affected twin had a smaller facial profile than the normal twin and was relatively retrognathic, similar to our twin B and to previous observations made in patients with UTS [Fillipson et al., 19651. Blood karyotype was 46,XX on the normal twin and 26% 4 5 3 7 4 % 46,XX on the affected twin. Permission to do fibroblast studies could not be obtained, so the distribution of 45,X and 46,XX cell lines in somatic tissues is unknown. The probability of monozygosity, based on 5 different methods but not including DNA analysis, was estimated to be >99.95%. Although the phenomonon of non-discordance for UTS in apparently identical twins has been described in over 10 reports, each case is unique in terms of the extent of phenotypic differences, the distribution of abnormal cell lines in blood and somatic tissues, and their relative proportions. One major variable responsible for the wide spectrum of findings is likely to be the timing of the loss of the X chromosome relative to the twinning event. If the loss of an X chromosome occurred during the first post-fertilization division and the twinning event occurred simultaneously, the result would be one twin with only 46,XX cells in the somatic tissues, and the other with only 45,X cells, which is what was found in our twins. The main difficulty with this explanation is that twinning occurs most commonly at the blastocyst stage (4-8 days after fertilization and before implantation), resulting in a single monochorionic diamniotic placenta as was observed is this case. Twinning occurs less frequently a t the morula stage ( < 3 days, 1 3 2 cells), which results in separate or fused dichorionic diamniotic placentae [Corner, 19551. If the X chromosome loss occurred at the two-cell stage or the morula stage and the twinning event occurred a t the blastocyst stage, the result would most likely be X/XX mosaicism in both embryos [e.g., Mikkelsen et al., 19631, assuming that the 45,X and 46,XX cell lineages were well-mixed in the cell mass before the twinning event occurred. However, it is conceivable that the different cell lines resulting from division of the progenitor 45,X and 46,XX cells selfaggregate, and that the twinning event could result in complete separation of the two cell lines, resulting in one 45,X and one 46,XX embryo, or partial separation leading to one mosaic embryo and one entirely normal embryo. Burn et al. [19861, in trying to explain the occur-
81
rence of Duchenne muscular dystrophy in one of MZ twin girls, suggested that X inactivation in females is followed by preferential aggregation of cells bearing the same active X within the inner cell mass prior to the twinning event. Nance [19901 proposes an additional mechanism, in that if the number of precursor cells in each embryo is unequal, the chance of seeing extreme X-inactivation profiles is greater in the twin arising from the fewest cells. It is therefore plausible that cell lines differing in the number of X chromosomes rather than the parental origin of the active X could also selfaggregate prior to the twinning event, or that the different cell lines might segregate almost completely by chance during an unequal twinning event. Finally, if the X chromosome loss occurred after the twinning event, there would be one completely normal embryo and one with X/XX mosaicism, the extent of which would depend on how many cells were present in that embryo at the time of the loss of an X chromosome from one cell lineage. The large variability in peripheral blood karyotype findings in twins discordant for UTS syndrome is best explained by the frequent presence of anastomoses between the two twin circulations when there is a single placenta, and perhaps differences in the growth rates of hematopoietic precursors with 46,XX or 45,X karyotypes. It is unlikely that mosaicism in the peripheral blood has any important clinical effect on either the affected twin or the normal twin. However, it must be noted that any determination of zygosity which relies heavily on blood groups or other methods using blood cells has the potential for error, since even DZ twins with fused placentae may rarely have anastomoses between the fetal blood supplies [Dunsford et al., 1953; Strong and Corney, 19671 and could be indistinguishable based on blood cell markers. F'rom the information available, we cannot be certain of the timing of the events by which both twinning and loss of an X chromosome occurred in our patients; however, the clinical and cytogenetic findings are consistent with a complete separation of 45,X cells and 46,XX cells during the twinning process, with the mosaicism in the peripheral lymphocytes being due to twin-to-twin transfusion. Monozygosity was established with a high degree of certainty with DNA typing on fibroblasts as well as lymphocytes. The twins had identical finger patterns in 9 of 10 digits, but twin B had a total finger ridge count which was 31 greater than twin A. Although a difference of this magnitude is not often seen in MZ twins, it is consistent with the observation that the average ridge count in UTS patients (169.3 2 41.6) is 42 greater than is found in control females (127.0 2 52.3) [Holt, 19681. Since the total ridge count is determined almost entirely by genetic factors (r = 0.95 for MZ twins, 0.46 for full sibs, and 0.23 for half-sibs [Phelan et al., 198111, the difference of 31 observed for the MZ twins described here may be a more accurate reflection of the effect of monosomy for the X chromosome on the dermal ridge count than the figure derived by comparing the mean value for UTS patients and control females. It was initially suspected, based on the finding of slight cubitus valgus on the right side only, that the phenotypically normal twin A might be a true mosaic for
82
Kaplowitz et al.
UTS. To increase the chances of finding the abnormal cell line, separate skin biopsies were taken from both the right and the left arms, but neither contained 45,X cells. It is conceivable that biopsies of multiple other tissues might have demonstrated a small number of 45,X cells in the normal twin, but it is unlikely that any 45,X cells which might be present have affected her growth, physical findings (exceptpossibly for the cubitus valgus), and ovarian function. Although the affected twin had evidence of “cerebral palsy” during her first year of life, it appears to have been mild since the motor deficits resolved spontaneously with time, and should not have affected significantly her growth and physical appearance. Therefore, it should be valid to compare the growth patterns and physical characteristics of these twins and to attribute the differences between them solely to the presence of monosomy for the X chromosome in the tissues of the affected twin.
REFERENCES Al-Awadi SA, Cuschieri A, Farag TI, Naguib K, Teebi AS, Al-Othman SA, Bahig AH (1983): Ullrich-Thrner syndrome in monozygotic twins. Am J Med Genet 15:537-542. Burn J, Povey S, Boyd Y, Munro EA, West L, Harper K, Thomas D (1986):Duchenne muscular dystrophy in one of monozygotic twin girls. J Med Genet 23:494-500. Corner GW [ 1955):Observed embryology of human single-ovum twins and other multiple births. Am J Obstet Gynecol 70:933-950. Dunsford I, Bowley CC, Hutchinson AM, Thompson JS, Sanger R, Race RR (1953):A human blood group chimera. Br Med J II:81. Fillipson R, Lindsten J , Almquist S (1965):Time of eruption of permanent teeth, cephalometric and tooth measurement and sulfation
factor activity in 45 patients with Turner syndrome and different types of X chromosome aberrations, Acta Endocrinol 48:91-113. Fineberg AP, Vogelstein B (1983):A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal Biochem 1325-13. Holt SB (1968): “The Genetics of Dermal Ridges.” Springfield, IL: Charles C Thomas Publishers, pp. 80, 158. Maniatis T, fiitsch EF, Sambrook J (1982): “Molecular cloning: A Laboratory Manual.” New York: Cold Spring Harbor Laboratory. Mikkelsen M, Froland A, Ellebjerg J (1963):XO/XX mosaicism in a pair of presumably monozygotic twins with different phenotypes. Cytogenetics 2%-98. Nance WE (1990):Invited editorial: do twin Lyons have larger spots? Am J Hum Genet 46:646-648. Phelan MC, Nance WE, Corey LA (1981):Determinants ofridge counts in MZ twin sibships. Ann Genet Med Gamellol 30:59-66. Potter AM, Taitz LS (1972):Turner syndrome in one of monozygotic twins with mosaicism. Acta Paediatr Scand 61:473-476. Southern EM (1975):Detection of specific sequences among DNA fragments separated by gel electrophoresis. J Mol Biol 98:503-517. Spence JE, Buffone GJ, Rosenbloom CL, Fernbach SD, Curry MR, Carpenter RJ, Ledbetter DH, O’Brien WE, Beaudet AL (1987): Prenatal diagnosis of cystic fibrosis using linked DNA markers and microvillar intestinal enzyme analysis. Hum Genet 76:5-10. Strong SJ,Corney G (1967):“The Placenta in Twin Pregnancy.” New York Pergamon Press, pp 78-89. Turpin R, Lejeune J, Lafourcade J, Chigot PL, Salmon C (1961):Presomption de monozygotisme au depit d’un dimorphisme sexuel: Sujet masculin XY et sujet haplo-X. CR Acad Sci [Dl (Paris) 252:2945-2946. Uchida IA, deSa DJ, Whelan DT (1983): 45,W46,XX mosaicism in discordant monozygotic twins. Pediatrics 71:413-417. Weiss E, Loevy H, Saunders A, Pruzansky S,Rosenthal I(1982):Monozygotic twins discordant for Ullrich-Turner syndrome. Am J Med Genet 13:389-399.
