Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Evidence from Turner's syndrome of an imprinted X-linked locus affecting cognitive function


Turner's syndrome is a sporadic disorder of human females in which all or part of one X chromosome is deleted1. Intelligence is usually normal2 but social adjustment problems are common3. Here we report a study of 80 females with Turner's syndrome and a single X chromosome, in 55 of which the X was maternally derived (45,Xm) and in 25 it was of paternal origin (45,Xp). Members of the 45,Xp group were significantly better adjusted, with superior verbal and higher-order executive function skills, which mediate social interactions4. Our observations suggest that there is a genetic locus for social cognition, which is imprinted5 and is not expressed from the maternally derived X chromosome. Neuropsychological and molecular investigations of eight females with partial deletions of the short arm of the X chromosome6 indicate that the putative imprinted locus escapes X-inactivation7, and probably lies on Xq or close to the centromere on Xp. If expressed only from the X chromosome of paternal origin, the existence of this locus could explain why 46,XY males (whose single X chromosome is maternal) are more vulnerable to developmental disorders of language and social cognition, such as autism, than are 46,XX females8.


An increasing number of mammalian genes are known to be subject to genomic imprinting, defined as parental origin-specific differential gene expression5. No imprinted gene has yet been described on the X chromosome in humans9, although the Xist gene has been shown to be imprinted in the mouse10. We considered that it should be possible to identify the effects of an X-linked imprinted locus by comparing classes of females with Turner's syndrome. In this chromosomal disorder all, or a substantial part, of one X chromosome is missing as a result of non-disjunction (chromosome loss during gametogenesis or early cleavage of the zygote). In 70% of monosomic (45,X) Turner-syndrome females, the single X chromosome is maternal in origin1; in the remainder it is paternal. Normal females (46,XX) possess both a maternally derived X chromosome (Xm) and a paternally derived X chromosome (Xp), one of which is randomly inactivated in any given somatic cell7. In monosomy X, the single chromosome is never inactivated. Differences in physical or behavioural phenotype between 45,Xp and 45,Xm Turner-syndrome subjects might therefore indicate the existence of an imprinted genetic locus.

Impaired social competence and adjustment are frequent in Turner's syndrome3, but a minority have good social skills11. Intelligence is usually normal in monosomic (45,X) cases2. We wished to test the hypothesis that 45,Xp females would be distinguishable from 45,Xm females by their social behaviour.

We karyotyped 80 monosomic (45,X) females and eight with deletions of the short arm of one X chromosome (46,XXp-). The parental origin of the normal X chromosome was determined by comparing proband and parental DNA polymorphisms located on distal Xp, in a region that was deleted in both 45,X and 46,XXp- patients. Of the 80 45,X females, 25 were 45,Xp and 55 were 45,Xm, with ages from 6 to 25 years. Clinical records did not reveal any significant group differences in terms of physical phenotype or in the provision of hormone-replacement therapy12. From a first-stage screening survey13 of parents and teachers, using standardized instruments14,15,16, we discovered that 40% of 45,Xm girls of school age had received a statement of special educational needs, indicating academic failure, compared with 16% of 45,Xp subjects (P < 0.05); the figure in the general population is just 2%. We also found that clinically significant social difficulties affected 72.4% of the 45,Xm subjects over 11 years of age (21 of 29), compared with 28.6% of 45,Xp females (4 of 14) (P < 0.02).

Such phenotypic variability between the two classes of monosomy X subjects could indicate the existence of an imprinted genetic locus, at which gene(s) that influence social adjustment are expressed only from the paternally derived X chromosome. On the maternally derived X chromosome, the corresponding locus would be silenced. This could account for the excess of social and learning difficulties among 45,Xm females compared with the 45,Xp variant. Pilot interviews and observations showed that 45,Xm females in particular lacked flexibility and responsiveness in social interactions. We therefore devised a questionnaire relevant to social cognition to summarize the main features of their behaviour (Box 1). Parents of our sample of Turner's syndrome females and the parents of age-matched normal male and female comparisons completed the questionnaire. The results for subjects aged from 6 to 18 years of age confirm there are significant differences between 45,Xm and 45,Xp females in the predicted direction (Fig. 1). We found that 45,Xm subjects obtained significantly higher scores than 45,Xp females on our measure of social-cognitive dysfunction. Normal boys also obtained significantly higher scores on the questionnaire than did normal girls, indicating poorer social cognition (Fig. 1). The magnitude and direction of this difference are compatible with the hypothesis that there is an imprinted locus on the X chromosome that influences the development of social cognitive skills (although the finding is of course also compatible with other explanations of gender differences in behaviour17). If the putative locus only expressed a gene (or genes) from the X chromosome of paternal origin (Xp), we would expect a tendency for normal females to have superior social cognitive skills than males. Because males (46,XY) invariably inherit their single X chromosome from their mothers, the genetic locus would be silenced. In contrast, the gene(s) would be expressed from Xp in approximately half of the cells of normal females if it were inactivated, and from all cells if it escaped X-inactivation7.

Figure 1: Subscale scores (mean + s.e.) of questionnaire on social-cognitive impairment (Box 1).

Higher scores indicate poorer social cognitive skills. The 45,Xm Turner-syndrome females score higher than 45,Xp females and both normal groups (P < 0.0001). Normal males score higher than normal females (P < 0.001); the effect size of this difference is 0.58, implying that the upper 50% of females score higher than approximately 72% of males. The ratios of mean social-dysfunction scores male : female and 45,Xm : 45,Xp are very similar 2.2 : 1 and 2.1 : 1, respectively). The overall higher scores for the Turner-syndrome subjects, compared with normal females, may reflect the contribution made by visuospatial abilities to social cognition11. These abilities are impaired equally in both monosomic groups. No information regarding parental origin of the normal X chromosome was made available to parents, their consultants, or members of the research team gathering these or other data.

We hypothesized that an imprinted X-linked locus, either without a Y-linked homologue or with a Y homologue showing a lower level of expression, could also explain why males are markedly more vulnerable than females to pervasive developmental disorders affecting social adjustment and language, such as autism8. Accordingly, 45,Xm Turner females should be exceptionally vulnerable to disorders of language and social adjustment that are usually more common in males. Consistent with this hypothesis, we identified three Turner-syndrome females with autism (meeting ICD-10 criteria)18, out of the total unbiased sample assessed personally in the course of this investigation (3.75%). All three females had retained a normal maternal X chromosome. The population prevalence of autism is less than 1 per 10,000 females8.

We then sought corresponding differences in the results of neuropsychological testing of both normal and Turner-syndrome subjects. We found that 45,Xp females were significantly superior in verbal intelligence to 45,Xm females (Table 1). Verbal IQ was negatively correlated with the social dysfunction score in the sample as a whole (r = −0.41, P < 0.002). Both monosomic Turner subgroups were equivalently impaired in non-verbal (including visuospatial) abilities.

Table 1 Neuropsychological test results

We subsequently performed several more focused neuropsychological assessments. First, although we had found that verbal abilities were moderately good predictors of social cognition, we hypothesized that higher-order cognitive skills would be better predictors. These are not measured directly by conventional intelligence tests. The executive functions of the prefrontal cortex4 exert an important influence on social interactions, and include skills that allow for the development of strategies of action and the inhibition of distracting impulses when striving towards a goal. Developmental disorders of social adjustment and language are associated with impairment in measures of executive function19. Many monosomic Turner-syndrome subjects have executive-function deficits2. We also predicted that 45,Xp females would perform better than 45,Xm females in tests of executive function skills. Finally, we predicted that abilities in which 45,Xp subjects were superior to 45,Xm subjects would also distinguish normal females from males, and in the same direction.

We chose tests of both planning ability (Tower of Hanoi) and behavioural inhibition (Same/Opposite World). The results were consistent with the initial hypothesis. Although the social-cognitive impairment score correlated independently with verbal IQ and with both measures of executive function, only planning ability (r = −0.4; P < 0.006) and behavioural inhibition (r = −0.37; P < 0.015) retained significance when the three variables were forced into a multiple regression analysis. The second prediction was partly confirmed: there were significant differences between 45,Xp and 45,Xm subjects in terms of the behavioural inhibition task, although not on the test of planning ability. The third prediction was fully confirmed (Table 1): there was no significant difference in the highest level achieved on the planning task between normal males and females. However, on the behavioural-inhibition task, normal females were superior to males and their mean scores were very similar to those of the 45,Xp subjects. Previous reports have noted gender differences on inhibition tasks, and they have been conceptually linked to corresponding differences in social behaviour20.

We then attempted to map the putative imprinted locus, provisionally, by studying eight females, ascertained as part of the Turner-syndrome project, who had large terminal deletions of the short arm of the paternally derived X chromosome (a 46,Xm Xp p-karyotype). These deletions all extended to a point proximal to the MLS gene at Xp22.3 (Table 2); the paternal X chromosome was consequently preferentially inactivated6. On examination, no-one in this series was found to have any significant learning difficulties. The mean verbal intelligence score for the group was 103.4 (s.d., 13.7), and their mean non-verbal IQ was 93 (s.d., 11.2). Their mean social dysfunction score was 3.75 (s.d., 3.1), a value very similar to that of the 45,Xp subjects (Fig. 1). We drew the following conclusions from these data. First, the imprinted locus had not been deleted on the structurally abnormal paternal X chromosome, and so it must lie on Xq or on Xp closer to the centromere than, UBE1 at Xp11.23. Second, the imprinted locus was not subject to X-inactivation, or the preferentially inactivated, partly deleted chromosome would not have expressed it. We already know of several genes that escape X-inactivation21.

Table 2 Cytogenetic and molecular information

We expected that isochromosomes of the long arm of the X chromosome [i(Xq)] would help in the mapping of the putative imprinted locus. However, virtually all i(Xq) chromosomes have been shown to contain proximal Xp sequences22. Duplication and consequent trisomy of the long arm of the X chromosome further complicate the interpretation of correlations between phenotype and genotype. Accordingly, this approach did not provide unambiguous evidence to assist in the deletion mapping of the locus.

An imprinted locus is not the only possible explanation for our findings. Among the 45,Xp females, there may have been a greater degree of cryptic mosaicism (with a normal 46,XX cell line) than among those who were 45,Xm. Some degree of mosaicism in apparently monosomic females may be essential for the fetus to avoid spontaneous abortion23. We examined both blood and cheek cells, tissues of mesodermal and ectodermal origin, respectively, and found two cryptic mosaics, but both were from the 45,Xm group.

Males are substantially more vulnerable to a variety of developmental disorders of speech, language impairment and reading disability, as well as more severe conditions such as autism8. Our findings are consistent with the hypothesis that the locus described, which we propose to be silent both in males and 45,Xm females, acts synergistically with susceptibility loci elsewhere on the genome to increase the male-to-female prevalence ratio of such disorders. Our data on normally developing children suggest it may also exert an effect on social and cognitive abilities in the normal range. These preliminary findings could thus provide evidence for the evolution of an imprinted X-linked locus that underlies the development of sexual dimorphism in social behaviour17.


Subjects. This study, which was approved by the local hospital ethics committee, involved 88 females with Turner's syndrome (80 monosomic and 8 partial X-chromosome deletions; age range, 6–25 years). They were selected from a national survey of Turner's syndrome and from records of the Wessex Regional Laboratory. The mean age of the 45,Xm females was 162.3 months (s.d., 57.6), that of the 45,Xp females was 164.5 months (s.d., 57.7), and that of the Xp- females was 185.8 months (s.d., 74.9). All subjects were healthy, with no significant neurological disease. Females with an Xp- chromosome were all referred for investigation because of short stature in middle childhood, with one exception who was karyotyped at birth. Neuropsychological test results are presented for subjects with verbal IQs 65 (three 45,Xm subjects and one 45,Xp subject had verbal IQs that fell out of range). Parents rated 70 normal males and 71 normal females (age range, 6–18 years) on the social-cognition scale. The neuropsychological test battery was used to assess 68 normal males and 91 normal females (age range, 6–25 years). Verbal IQs were in the range 65–151. All normal comparison subjects were recruited from urban and suburban schools (6–18 years) and from hospital staff (18–25 years).

Behavioural and cognitive measures. Initial screening was conducted by postal questionnaires using a well-standardized set of instruments14,15,16. These were completed by parents, teachers and the Turner-syndrome subjects themselves (11 years and over). The social cognition questionnaire (Box 1) was completed by parents only. In a survey of 175 Turner-syndrome subjects for whom we obtained parental ratings on two occasions, a mean of 2.7 years apart, the intraclass correlation coefficient was 0.81 (P < 0.01). Scores correlate with the self-rated social problem subscale of the YSR16 0.58 (P < 0.002), with the teacher rating on the TRF15 0.54 (P < 0.001), and with the parent-rated CBCL14 0.69 (P < 0.001). The range of scores was 0–23 in the Turner-syndrome sample and 0–21 in the normal sample (maximum score of 24). The CBCL14 was completed by 70 parents, the YSR16 was completed by 40 subjects over 11 years of age, and the TRF15 was completed by 45 teachers. Clinical significance of social problems was estimated according to clinical T scores14,15,16. Measures of cognition included the Wechsler Intelligence Scales for Children (WISC III-UK)24 and the Wechsler Adult Intelligence Scales–Revised (WAIS-R)25. The behavioural inhibition task was the Same–Opposite World subtest from the Test of Everyday Attention for Children26. This yields a time measure that ascertains the difference in latency for a subject responding to a series of stimuli on a task of sequential responses, which are named both as they appear and then opposite to their appearance. The subject reads a random series of numbers (1 and 2) saying ‘one’ to 1, and ‘two’ to 2. The subjects then repeat the task on a new series, but this time they have to inhibit the prepotent response and instead say ‘two’ to 1, and ‘one’ to 2, correcting any errors before proceeding. Test–retest reliability on a sample of 70 normal children gave an intraclass correlation coefficient of 0.62 (P < 0.001). The Tower of Hanoi task was based on the procedure described previously27. It was scored according to the most complex level of the problem the child could solve reliably. Test–retest reliability gave an intraclass correlation coefficient for the highest level achieved of 0.45 (P < 0.001), which is in line with expectations for a test that makes novel demands of this nature28.


  1. 1

    Jacobs, P. al. Acytogenetic and molecular reappraisal of a series of patients with Turner's syndrome. Ann. Hum. Genet. 54, 209–223 (1990).

    CAS  Article  Google Scholar 

  2. 2

    Pennington, B. al. The neuropsychological phenotype in Turner syndrome. Cortex 21, 391–404 (1985).

    CAS  Article  Google Scholar 

  3. 3

    McCauley, E., Ito, J. & Kay, T. Psychosocial functioning in girls with the Turner syndrome and short stature. J. Am. Acad. Child Psychiat. 25, 105–112 (1986).

    CAS  Article  Google Scholar 

  4. 4

    Damasio, A. R. On some functions of the human prefrontal cortex. Proc. N. Y. Acad. Sci. 769, 241–251 (1995).

    ADS  CAS  Article  Google Scholar 

  5. 5

    Barlow, D. P. Gametic imprinting in mammals. Science 270, 1610–1613 (1995).

    ADS  CAS  Article  Google Scholar 

  6. 6

    Ballabio, A. & Andria, G. Deletions and translocations involving the distal short arm of the human X chromosome: review and hypotheses. Hum. Mol. Genet. 1, 221–227 (1995).

    Article  Google Scholar 

  7. 7

    Lyon, M. F. Gene action in the X-chromosome of the mouse (Mus musculus L). Nature 190, 372–373 (1961).

    ADS  CAS  Article  Google Scholar 

  8. 8

    Bailey, A., Philips, W. & Rutter, M. Autism: towards an integration of clinical, genetic, neuropsychological and neurobiological perspectives. J. Child Psychol. Psychiat. 37, 89–126 (1996).

    CAS  Article  Google Scholar 

  9. 9

    Ledbetter, D. H. & Engel, E. Uniparental disomy in humans: development of an imprinting map and its implications for prenatal diagnosis. Hum. Mol. Genet. 4, 1757–1764 (1995).

    CAS  Article  Google Scholar 

  10. 10

    Zuccotti, M. & Monk, M. Methylation of the mouse Xist gene in sperm and eggs correlates with imprinted Xist expression and paternal X-inactivation. Nature Genet. 9, 316–320 (1995).

    CAS  Article  Google Scholar 

  11. 11

    McCauley, E., Kay, T., Ito, J. & Trader, R. The Turner syndrome: cognitive deficits, affective discrimination and behaviour problems. Child Dev. 58, 464–473 (1987).

    CAS  Article  Google Scholar 

  12. 12

    Saenger, P. Clinical Review 48: The current status of diagnosis and therapeutic intervention in Turner's syndrome. J. Clin. Endocrinol. Metabol. 77, 297–301 (1993).

    CAS  Google Scholar 

  13. 13

    Skuse, D., Percy, E. L. & Stevenson, J. in Growth, Stature, and Adaptation. Behavioral, Social, and Cognitive Aspects of Growth Delay (eds Stabler, B. & Underwood, L.) 151–164 (UCP, Chapel Hill, 1994).

    Google Scholar 

  14. 14

    Achenbach, T. M. Manual for the Child Behavior Checklist/4-18 and 1991 Profile (Department of Psychiatry, University of Vermont, Burlington, VT, 1991).

    Google Scholar 

  15. 15

    Achenbach, T. M. Manual for the Teacher's Report Form and 1991 Profile (Department of Psychiatry, University of Vermont, Burlington, VT, 1991).

    Google Scholar 

  16. 16

    Achenbach, T. M. Manual for the Youth Self-Report Form and 1991 Profile (Department of Psychiatry, University of Vermont, Burlington, VT, 1991).

    Google Scholar 

  17. 17

    Eagley, A. H. The science and politics of comparing men and women. Am. Psychol. 50, 145–158 (1995).

    Article  Google Scholar 

  18. 18

    World Health Organization The ICD-10 Classification of Mental and Behavioural Disorders: Clinical Descriptions and Diagnostic Guidelines (World Health Organizaiton, Genev, 1992).

  19. 19

    Pennington, B. F. & Ozonoff, S. Executive functions and developmental psychopathology. J. Child Psychol. Psychiat. 37, 51–87 (1996).

    CAS  Article  Google Scholar 

  20. 20

    Bjorklund, D. F. & Kipp, K. Parental investment theory and gender differences in the evolution of inhibition mechanisms. Psychol. Bull. 120, 163–188 (1996).

    CAS  Article  Google Scholar 

  21. 21

    Disteche, C. M. Escape from X inactivation in human and mouse. Trends Genet. 11, 17–22 (1995).

    CAS  Article  Google Scholar 

  22. 22

    Wolff, D. J., Miller, A. P., Van Dyke, D. L., Schwartz, S. & Willard, H. F. Molecular definition of breakpoints associated with human Xq isochromosomes: implications for mechanism of formation. Am. J. Hum. Genet. 58, 154–160 (1996).

    CAS  PubMed  PubMed Central  Google Scholar 

  23. 23

    Hassold, T., Pettay, D., Robinson, A. & Uchida, I. Molecular studies of parental origin and mosaicism in 45,X conceptuses. Hum. Genet. 89, 647–652 (1992).

    CAS  Article  Google Scholar 

  24. 24

    Wechsler, D. Wechsler Intelligence Scale for Children 3rd UK edn (Psychological Corporation, London, 1992).

    Google Scholar 

  25. 25

    Wechsler, D. Wechsler Adult Intelligence Scales-Revised (Psychological Corporation, New York, 1986).

    Google Scholar 

  26. 26

    Borys, S. V., Spitz, H. H. & Dorans, B. A. Tower of Hanoi performance of retarded young adults and nonretarded children as a function of solution length and goal state. J. Exp. Child Psychol. 33, 87–110 (1982).

    CAS  Article  Google Scholar 

  27. 27

    Manly, T., Robertson, I. H. & Anderson, V. The Test of Everyday Attention for Children (TEACh) (Thames Valley Test Company, Bury St Edmunds, in the press).

  28. 28

    Rabbitt, P. M. A. in Methodologies of Frontal and Executive Function (ed. Rabbitt, P. M. A.) (Psychology Press, Hove, in the press).

  29. 29

    Temple, C. M. & Carney, R. A. Patterns of spatial functioning in Turner's syndrome. Cortex 31, 109–118 (1995).

    CAS  Article  Google Scholar 

Download references


We thank E. Percy, S. Cave, A. O'Herlihy, R. South, J. Smith, M. Power and D.Robinson for assistance; M. Pembrey for comments and discussion; many paediatric consultants for assisting with the recruitment of patients, the schools who participated, and all of the subjects of our investigation and their families for their time. This research was supported by the Wellcome Trust and the Child Growth Foundation. Compilation of the national register of Turner syndrome was supported by the British Society for Paediatric Endocrinology and by Pharmacia.

Author information



Corresponding author

Correspondence to D. H. Skuse.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Skuse, D., James, R., Bishop, D. et al. Evidence from Turner's syndrome of an imprinted X-linked locus affecting cognitive function. Nature 387, 705–708 (1997).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing