1 23

Human Genetics ISSN 0340-6717 Hum GenetDOI 10.1007/s00439-011-1020-y

The role of the TCF4 gene in thephenotype of individuals with 18qsegmental deletions

Minire Hasi, Bridgette Soileau, CourtneySebold, Annice Hill, Daniel E. Hale,Louise O’Donnell & Jannine D. Cody

1 23

Your article is protected by copyright and

all rights are held exclusively by Springer-

Verlag. This e-offprint is for personal use only

and shall not be self-archived in electronic

repositories. If you wish to self-archive your

work, please use the accepted author’s

version for posting to your own website or

your institution’s repository. You may further

deposit the accepted author’s version on a

funder’s repository at a funder’s request,

provided it is not made publicly available until

12 months after publication.

ORIGINAL INVESTIGATION

The role of the TCF4 gene in the phenotype of individuals with 18qsegmental deletions

Minire Hasi • Bridgette Soileau • Courtney Sebold •

Annice Hill • Daniel E. Hale • Louise O’Donnell •

Jannine D. Cody

Received: 15 February 2011 / Accepted: 25 May 2011

� Springer-Verlag 2011

Abstract The goal of this study is to define the effects

of TCF4 hemizygosity in the context of a larger seg-

mental deletion of chromosome 18q. Our cohort included

37 individuals with deletions of 18q. Twenty-seven had

deletions including TCF4 (TCF4?/-); nine had deletions

that did not include TCF4 (TCF4?/?); and one individual

had a microdeletion that included only the TCF4 gene.

We compared phenotypic data from the participants’

medical records, survey responses, and in-person evalu-

ations. Features unique to the TCF4?/- individuals

included abnormal corpus callosum, short neck, small

penis, accessory and wide-spaced nipples, broad or

clubbed fingers, and sacral dimple. The developmental

data revealed that TCF4?/? individuals were only mod-

erately developmentally delayed while TCF4?/-

individuals failed to reach developmental milestones

beyond those typically acquired by 12 months of age.

TCF4 hemizygosity also conferred an increased risk of

early death principally due to aspiration-related compli-

cations. Hemizygosity for TCF4 confers a significant

impact primarily with regard to cognitive and motor

development, resulting in a very different prognosis for

individuals hemizygous for TCF4 when compared to

individuals hemizygous for other regions of distal 18q.

Introduction

It has recently been shown that hemizygosity of the TCF4

gene causes Pitt–Hopkins syndrome (MIM ID #610042)

through a haploinsufficiency mechanism (Zweier et al.

2007). The TCF4 gene is located at 18q21.1 and is,

therefore, also hemizygous in some individuals with larger

segmental deletions of 18q. The goal of this study is to

define the effect of TCF4 hemizygosity in individuals with

18q deletions.

The constellation of phenotypic features known as

Pitt–Hopkins syndrome is characterized by severe intel-

lectual disability, wide mouth with fleshy lips, a beaked

nose, and intermittent hyperventilation followed by apnea

(Pitt and Hopkins 1978). Since it was first described, the

phenotype has been shown to have three different

genetic causes. It can be caused dominantly by hemi-

zygosity of or inactivating mutations in the TCF4 gene,

or recessively by mutations in the CNTNAP2 gene on

chromosome 7q35 or the NRXN1 gene on chromosome

2p16.3 (Zweier et al. 2009). Because this phenotype has

multiple underlying molecular mechanisms, the phrase

‘‘Pitt–Hopkins’’ refers only to the clinically-defined

syndrome.

Web Resources Online Mendelian Inheritance in Man (OMIM),

http://www.ncbi.nlm.nih.gov/Omim/.

M. Hasi � B. Soileau � C. Sebold � A. Hill �D. E. Hale � L. O’Donnell � J. D. Cody (&)

Department of Pediatrics, UT Health Science Center,

7703 Floyd Curl Dive, San Antonio, TX 78229, USA

e-mail: cody@uthscsa.edu

D. E. Hale � J. D. Cody

CHRISTUS Santa Rosa Children’s Hospital,

San Antonio, TX, USA

L. O’Donnell

Department of Psychiatry, UT Health Science Center

at San Antonio, San Antonio, USA

J. D. Cody

The Chromosome 18 Registry and Research Society,

San Antonio, TX, USA

123

Hum Genet

DOI 10.1007/s00439-011-1020-y

Author's personal copy

The identification of the genetic bases of Pitt–Hopkins

has allowed a fuller appreciation of the range of phenotypic

features associated with mutations or deletions of these

genes. It has recently been realized that 2% of individuals

with phenotypic Angelman syndrome actually had TCF4

aberrations (Takano et al. 2010). In addition, Rosenfeld

et al. (2009) identified seven cases with TCF4 deletions

which were referred for chromosomal microarray analysis

due to intellectual disability. In this genotypically ascer-

tained group, a reevaluation of the phenotypic effects of

TCF4 hemizygosity was undertaken. Of these patients,

only 3 of 7 had a breathing abnormality and none had

seizures, indicating that the penetrance of these features is

significantly less than 100%. In their review of published

cases likely caused by haploinsufficiency of TCF4

(N = 36), they found that the Pitt–Hopkins facial appear-

ance was present in 91% of cases. Other features seen in

more than half of the subjects included: severe psycho-

motor delay (97%), hypotonia (88%), happy disposition

(85%) (though no elaboration of this behavior was pro-

vided), single palmar crease (69%), microcephaly (64%),

constipation (58%), and brain abnormalities (56%).

Interestingly, variations in all three genes responsible

for Pitt–Hopkins syndrome are associated with schizo-

phrenia and autism. Hemizygous deletions of NRXN1 and

CNTNAP2 as well as SNPs in both genes are associated

with an increased risk of schizophrenia, epilepsy, and

autism spectrum disorder (Kirov et al. 2009; Friedman

et al. 2008) and SNPs of TCF4 have been associated with a

slightly increased risk of schizophrenia (Stefansson et al.

2009). In addition, individuals with 18q deletions including

TCF4, NETO1 and FBXO15 are more likely to exhibit

autistic-like behavior (O’Donnell et al. 2010). These data

point to a potential functional relationship between the

products of these three genes as well as a causal relation-

ship between autism and schizophrenia.

The TCF4 gene produces a basic helix-loop-helix (bHLH)

transcription factor which acts through binding to E-box

consensus sequences in the promoter regions of the target

genes. It belongs to a family of proneural bHLH transcription

factors controlling differentiation of neuronal subtypes.

The temporal and spatial differentiation of the numerous

neural cell types is controlled by a relatively small number

of transcription factors acting as homo- and hetero-dimers.

The hetero-dimerization of transcription factors produces a

greater diversity of regulatory combinations, thereby a

greater diversity and specificity of gene expression (Guille-

mot 2007). It is likely that the TCF4 gene product interacts

with numerous other transcription factors. In the mouse, Tcf4

dimerizes with Math1, another bHLH transcription factor.

Interestingly, mice hemizygous for Math1 die shortly after

birth from central apnea (Rose et al. 2009).

One of the overall goals of our research group is to

determine which genes on 18q contribute to the 18q-

phenotype through haploinsufficiency. This information

will eventually allow the molecular karyotype to be pre-

dictive with regard to the phenotype. The aim of this study

was to determine the extent to which hemizygosity of

TCF4 contributes to the physical and behavioral phenotype

in people with segmental 18q deletions.

Methods

Subject recruitment

Potential participant families learn about the research study

from a variety of sources. Primarily, however, families are

referred to the Research Center from the Chromosome 18

Registry & Research Society. Eligibility for the study

requires a cytogenetic or molecular diagnosis of an 18q

deletion. This study was approved by the Institutional

Review Board of the University of Texas Health Science

Center at San Antonio. All families were and continue to

be involved in the informed consent process, which is

appropriately documented.

Phenotypic assessment

For all families enrolling at the Chromosome 18 Clinical

Research Center, phenotypic data are compiled from three

sources. First, upon enrollment, families provide the

Research Center with extensive medical records to doc-

ument the participant’s medical and developmental his-

tories. These data are entered into a relational database

which is updated annually with the most recent medical

and developmental information obtained from the fami-

lies, providing longitudinal data on each of the study

participants. Second, all families are solicited annually by

mail to complete psychological surveys. Parents are asked

to complete the following questionnaires and return them

by mail: Behavior Assessment System for Children-Sec-

ond Edition (BASC-2; Reynolds and Kamphaus 2004)

which provides information regarding the presence of

behavior problems and emotional disturbance; the Vine-

land Adaptive Behavior Scales-Second Edition (Sparrow

et al. 2005) which asks parents to rate communication,

daily living and socialization skills and the Gilliam Aut-

ism Rating Scale-First Edition (GARS; Gilliam 1995) or

Second Edition (GARS-2; Gilliam 2006) which provides

an overall probability of autism rating.

The third data source is the comprehensive clinical

evaluation at the Chromosome 18 Clinical Research Center

as previously described (Cody et al. 2009).

Hum Genet

123

Author's personal copy

The evaluation completed at the Research Center

includes a comprehensive neuropsychological evaluation

comprised multiple measures. To obtain a measurement of

estimated cognitive functioning, an individually adminis-

tered measure of ability is given. If the participant is able to

understand the task demands to the extent that it is possible

to obtain a reliable and valid estimate of ability, then

measures based on the chronological age are given: Bayley

Scales of Infant and Toddler Development-Third Edition

(Bayley 2006); Differential Abilities Scales-Second

Edition (Elliot 2007) or the Wechsler Adult Intelligence

Scales-Third Edition (Wechsler 1997). If it is not possible

to evaluate the intellectual functioning using tasks and

activities based on their chronological age, it is necessary

to employ standardized measures which are routinely used

to assess the cognition, language, and motor abilities of

infants and toddlers. Because some participants are sig-

nificantly older than the normative comparison group for

these evaluations no standard comparison was possible.

Therefore, estimates of cognitive functioning are generated

from age equivalent scores.

To gain additional information about the group of par-

ticipants described in this project, it was necessary to

re-contact participating families to fill in gaps in the data

and to address new questions. All families participating in

this study were contacted by telephone to obtain additional

information about the participant’s respiratory history,

including abnormal breathing patterns.

Genotypic assessment

The DNA of all participants was evaluated using custom

designed oligonucleotide microarray comparative genomic

hybridization as previously described (Heard et al. 2009).

All participants in the study cohort assessed here had a

hemizygous region of 18q without other major chromo-

somal copy number changes.

Results

To determine the contribution of TCF4 hemizygosity to the

overall phenotype of 18q-, we compared the phenotypes

of those with and without hemizygosity of TCF4. We have

a cohort of over 200 individuals with hemizygosity for a

portion of chromosome 18q, each with a unique hemizy-

gous region (Heard et al. 2009). Within this cohort, there

are 27 individuals with a deletion that includes TCF4

(TCF4?/-). As shown in Fig. 1 and pictures in Fig. 2 (a–

m), this is a very heterogeneous group including 13 people

with terminal deletions between 24.67 and 30.71 Mb in

size (Fig. 1, Panel b), 8 people with small interstitial

deletions (between 5.6 and 16.88 Mb in size) and 6 with

large interstitial deletions (between 24.02 and 27.58 Mb in

size) (Fig. 1, Panel c).

Within our cohort, there are 132 individuals with ter-

minal deletions of 18q with breakpoints distal to the TCF4

gene (TCF4?/?). The comparison group was selected from

these individuals. We selected the TCF4?/? individuals

with the specific aim of creating a comparison group of

similar size to the TCF4?/- group. Because we had cog-

nitive data on 8 of the TCF4?/- individuals and adaptive

behavior data on 11, we selected nine TCF4?/? individuals

with breakpoints as close as possible to TCF4 to serve as a

control group. Parenthetically, the next smallest deletion

contained several more genes than other members of the

control group. The photographs of six of these individuals

are shown in Fig. 2n–s.

In this study, the TCF4?/? and TCF4?/- groups were

compared with regard to both physical and behavioral

phenotypes. The TCF4?/- group included 27 individuals:

12 males and 15 females. Since this is a longitudinal study,

the information was gathered over a period of time.

However, the age range at the most recent assessment was

10 months to 24 years 7 months, with an average age of

10 years 10 months. The TCF4?/? group of nine individ-

uals included four males and five females with an average

age range from 10 months to 28 years 7 months with an

average age of 13 years.

We aimed to (1) identify which features are found only

in the TCF4?/- group, and (2) determine the incidence of

features associated with the Pitt–Hopkins phenotype in

both the TCF4?/- group and the TCF4?/? group. The

physical features that distinguish the TCF4?/- group from

the TCF4?/? group are listed in Table 1 (Zweier et al.

2007, 2008; Pitt and Hopkins 1978; Rosenfeld et al. 2009;

Amiel et al. 2007; Andrieux et al. 2008; Brockschmidt

et al. 2007; de Pontual et al. 2009; Giurgea et al. 2008;

Kalscheuer et al. 2008; Peippo et al. 2006; Singh 1993;

Van Balkom et al. 1998). Only abnormalities of the corpus

callosum (64%) were found in more than half of the

TCF4?/- group. Other features were found in fewer than

35% of participants with TCF4 hemizygosity, yet were

unique to this group.

We then compared the non-unique phenotypic features

between the two groups. The features previously described

as associated with Pitt–Hopkins syndrome are indicated in

italics (Rosenfeld et al. 2009). These findings are shown in

Table 2 and reveal that the majority are non-specific find-

ings often associated with many different conditions

(Zweier et al. 2007, 2008; Pitt and Hopkins 1978; Rosen-

feld et al; Amiel et al. 2007; Andrieux et al. 2008;

Brockschmidt et al. 2007; de Pontual et al. 2009; Giurgea

et al. 2008; Kalscheuer et al. 2008; Peippo et al. 2006;

Singh 1993; Van Balkom et al. 1998; Taddeucci et al.

2010).

Hum Genet

123

Author's personal copy

To isolate the effect of TCF4 hemizygosity on cognitive

functioning, we wanted to compare the TCF4?/- group

with terminal deletions (N = 13) to the control group of

nine TCF4?/? individuals who also had terminal deletions.

However, we had cognitive data from in-person assess-

ments on only 8 of the 13 individuals in the TCF4?/-

group. Table 3 presents the estimated intellectual abilities

of these two groups by age. As Table 3 indicates, the

estimated intellectual abilities for the TCF4?/? group range

from mild intellectual disability to low average cognitive

functioning. This is quite mild in comparison to the eight

persons in the TCF4?/-group. In this group, cognitive and

motor development is very significantly delayed with

standard scores within the profound to severe range of

intellectual disability. In fact, the individuals in the

TCF4?/- group were unable to do even the easiest items of

each intellectual assessment instrument designed to assess

typical children under 1 year of age. Therefore, the most

informative indication of their cognitive function was to

report their age equivalent; the age at which the skills are

acquired in a typically developing child. These pronounced

deficits were present in children under 5 years of age and

continue to be present across the lifespan into child and

young adulthood. In general, the cognitive and motor

growth pattern is essentially flat across the lifespan

meaning that they did not acquire additional skills beyond

that of a 12 months old.

We also compared the language development of these

same two groups of individuals. All persons in the

TCF4?/- group are nonverbal with receptive language

limited to reactions to sounds in the environment, calming

when spoken to and recognition of caregiver’s voice with

Fig. 1 Panel a Chromosome 18 ideogram. The box indicates the

region of the chromosome shown in panels b–d. Panels b–d Display

the aCGH data from study participants with distal 18q deletions. The

light bars indicated the presence of the intact chromosome. The darkband at the end of the light bar indicated the breakpoint region. The

participant’s study number is to the left of their data. Panel

b Individuals with terminal deletions of 18q, who have one copy of

TCF4. Panel c Individuals with interstitial deletions of 18q who have

one copy of TCF4. Panel d The nine individuals in our study cohort

with terminal deletions and breakpoints closest to but not including

TCF4. These individuals phenotypic data were used as the compar-

ison group

Hum Genet

123

Author's personal copy

increased motor movement (wiggling). Expressive lan-

guage consisted of undifferentiated throaty or nasal sounds.

In contrast, in the TCF4?/? group, six of the nine persons

communicate verbally; one child uses sign language to

communicate; and two individuals were under the age of

two at the time of testing and were not yet speaking. These

two toddlers do however communicate through the use of

engaged eye contact, facial expressions, gestures and sub-

vocalizations.

Parents in both terminal deletion groups completed an

adaptive behavior questionnaire which compared the

communication, daily living, and socialization skill

development of their children with a normative group of

same-age peers. We had data on 11 of the 13 individuals in

the terminal deletion TCF4?/- group using this instrument

(Table 4). Parental ratings of both groups across all

domains were congruent with the intellectual assessment

results obtained through one-on-one assessment. The

adaptive behavior functioning of the TCF4?/- group was

rated as severely impaired while the behavioral functioning

of the TCF4?/? group was rated as falling within the mild

range of impairment.

Parents also completed the BASC-2 which provides an

evaluation of problem behaviors which fall into three

general areas: externalizing problems (hyperactivity,

aggressiveness, conduct problems); internalizing problems

(anxiety, depression, somatization) and those in the

behavior index [atypicality (disconnection from reality),

withdrawal and attention problems]. Only problems with

attention were rated by parents in both groups as cause for

concern (average T-Score = 63.18 for the TCF4?/- group

and average T-Score = 60.55 for the TCF4?/? group).

Average parental ratings across all of the other domains

were within normal or typical limits compared with same

age peers.

To determine the presence of behaviors consistent with

autism, parents evaluated their children’s communication

and social skill functioning along with the presence of

stereotyped or perseverative behaviors using one of the

Gilliam Autism Rating Scales (GARS or GARS-2). The

Fig. 2 Individuals with terminal deletions of 18q, who have one copy

of TCF4. a 3 years 9 months, b 6 years, c 6 years 8 months,

d 8 years 4 months, e 4 months 29 days, f 9 months, g 14 months

22 days. Individuals with interstitial deletions of 18q who have one

copy of TCF4. h 19 years 10 months, i 4 years, j 1 year 5 months,

k 4 years 2 months, l 2 years 8 months, m 1 year 10 months.

Individuals with terminal deletions and breakpoints closest to but

not including TCF4. n 25 years 9 months, o 14 years, p 7 years

5 months, q 8 years 9 months, r 6 years 6 months, s 1 year 2 months.

One individual with a small interstitial deletion that includes only the

TCF4 gene. t 12 years 8 months

Hum Genet

123

Author's personal copy

GARS is a screening tool and should not be used alone to

make a definitive diagnosis of autism. Therefore, scores

from this instrument are cataloged as to the probability of

having an autistic diagnosis. Parental ratings of children in

the TCF4?/- group indicated a very likely or high proba-

bility of autism being part of the diagnostic picture while

the ratings of parents of children in the TCF4?/? group

indicated that it was a possibility.

Within the group of TCF4?/-, children who had been

evaluated at the Research Center, there is great genotypic

variability between the individuals with regard to the

number of other genes involved in the deletion. Although

Table 1 Unique features of

those with TCF4 hemizygosity

Features in bold italics are those

in the literature as TCF4phenotypic componentsa Zweier et al. (2007, 2008);

Pitt and Hopkins (1978);

Rosenfeld et al. (2009); Amiel

et al. (2007); Andrieux et al.

(2008); Brockschmidt et al.

(2007); de Pontual et al. (2009);

Giurgea et al. (2008);

Kalscheuer et al. (2008); Peippo

et al. (2006); Singh (1993); Van

Balkom et al. (1998)

TCF4 phenotypic componentsa Hemizygous for TCF4

Number %

Abnormal corpus callosum 16/25 64

Atrial septal defect 7/20 35

Sacral dimple 7/23 30

Clubbed or broad fingers 3/11 27

Short neck 5/19 26

Hypertonia 7/27 26

Genital abnormalities-small penis 3/12 25

Camptodactyly of the fingers 5/22 23

Wide spaced nipples 5/22 23

Toe-2nd and 4th overlapping 3rd toe bilaterally 4/21 19

Premature death (\22 y/o) 5/27 18

Overfolding of the ears 4/22 18

Short philtrum 4/23 17

Malrotation of intestine 4/25 16

Cortical visual impairment 4/26 15

Absence or flattening of the superior fork of the antihelix 3/22 13

Optic atrophy 3/26 12

Accessory nipple 2/22 9

Wolff–Parkinson–White syndrome 2/25 8

Kyphosis 1/22 4

Table 2 Features NOT unique

to TCF4 hemizygosity

Features in bold italics are those

in the literature as TCF4phenotypic componentsa Zweier et al. (2007, 2008);

Pitt and Hopkins (1978);

Rosenfeld et al. (2009); Amiel

et al. (2007); Andrieux et al.

(2008); Brockschmidt et al.

(2007); de Pontual et al. (2009);

Giurgea et al. (2008);

Kalscheuer et al. (2008); Peippo

et al. (2006); Singh (1993); Van

Balkom et al. (1998); Taddeucci

et al. (2010)

TCF4 phenotypic componentsa Hemizygous for

TCF4Nine largest terminal deletions with two

copies of TCF4

Number % Number %

Hypotonia 26/26 100 9/9 100

Microcephaly 16/26 62 3/9 33

Postnatal growth retardation 15/27 55 1/9 11

Single palmar crease 12/22 54 1/9 11

Seizures 14/27 52 4/9 44

Myopia 14/27 52 3/9 33

Intra uterine growth retardation 13/25 52 1/9 11

Constipation 11/25 44 2/9 22

Central apnea 9/25 36 2/9 22

Drooling 7/25 28 3/9 33

Genital abnormalities-cryptorchidism 3/12 25 1/4 25

Strabismus 6/27 22 2/9 22

Scoliosis 2/14 14 1/9 11

Hum Genet

123

Author's personal copy

the behavioral and cognitive data presented in Tables 3 and

4 are from only those individuals with terminal deletions,

we wished to evaluate whether other regions of hemizy-

gosity of 18q had an impact on the effect of TCF4 hemi-

zygosity. For this analysis, we now included the cognitive

data from both the interstitial as well as the terminal

deletions in the TCF4?/- group. Again, these data were not

available on every individual whose genotype data are

shown in Fig. 1 and phenotype data shown in Tables 1 and

2. Because there are no other genes on 18q specifically

identified as haploinsufficient, the comparison between the

TCF4?/- subgroups was made based on grouping them by

the size of their deletion. We created three sub-groups:

those with terminal deletions (N = 13, 8 with cognitive

data), those with large interstitial deletions (N = 8, 5 with

cognitive data) and those with small interstitial deletions

(N = 6, 4 with cognitive data). The behavioral perfor-

mance of these three groups was compared to each other.

In addition, we have one participant, age 12 years and

8 months who has a small interstitial deletion of only the

TCF4 gene whose data have not been included elsewhere

in our analysis. Again, age equivalent scores were gener-

ated for the three groups because the use of standard peer-

based assessment measures was not possible due to the

participants’ very low cognitive and motor functioning. As

Table 5 indicates, the cognitive functioning of the three

groups is not significantly different. All three groups of

people hemizygous for the TCF4 gene have significantly

delayed cognitive and motor functioning. This suggests

that the size of the hemizygous region has little to no effect

on the developmental impact of the TCF4 gene. In essence,

the effect of TCF4 hemizygosity is so profound that, with

regard to development, children with large regions of

hemizygosity including many other genes are not more

developmentally delayed than children with hemizygosity

for the TCF4 gene alone.

One of the more striking findings was the realization

that, within our large cohort of people with simple terminal

18q deletions (N = 132), the majority of individuals who

died before the age of 18 had a deletion that included the

TCF4 gene. Figure 3 illustrates this point. The figure shows

either the current age or the age of death for the two

Table 3 Comparison of intellectual abilities

18q-, TCF4?/?

Intellectual

abilities

N = 9

CA less than

5 years

N = 3

CA between

6–12 years

N = 4

CA 13 years

and older

N = 2

Overall IQ and overall

range of scores

Full scale IQ 55a 63a 62a 61a

(50–82)

Verbal IQ 57a 66a 59a 63a

(50–81)

Nonverbal IQ 55a 75a 70a 69a

(50–90)

18q-, TCF4?/-

Intellectual abilities

N = 8

CA less than 5 years

N = 3

CA between 6–12 years

N = 4

CA 13 years and older

N = 1

Cognitive abilities AE = 1.3 months of age AE = 7 months of age AE = 6 months of age

Motor abilities AE = 2 months of age AE = 6 months of age AE = 6 months of age

CA chronological age, AE age equivalenta Average standard scores with a mean of 100 and a standard deviation of 15

Table 4 Average vineland

adaptive behavior scales-second

edition parental ratings

a Average standard scores with

a mean of 100 and a standard

deviation of 15

Type Communication Daily living

skills

Socialization Overall adaptive

behavior

TCF4?/? 60.22a 57.44a 61.22a 56.77a

N = 9

Terminal deletion, TCF4?/- 36.38a 36.84a 40.15a 36.46a

N = 11

Hum Genet

123

Author's personal copy

subgroups within our entire cohort; those with one copy of

TCF4 and those with two copies of TCF4. We investigated

the cause of death by reviewing medical records and

interviewing the parents. While a variety of reasons were

given for the cause of death, almost all of the children who

died had a history of multiple pneumonias, primarily

thought to be due to chronic aspiration, as is common in

such severely delayed children. This may have been

complicated by the breathing abnormalities that are com-

mon in the affected individuals.

Discussion

The analysis revealed three key observations. First, the

features unique to those with TCF4 hemizygosity were

abnormal corpus callosum, small penis, accessory nipples,

broad or clubbed fingers, sacral dimple, short neck and

wide spaced nipples. The potentially most clinically sig-

nificant of these physical findings is an abnormally thin or

absent corpus callosum. It would be reasonable to expect

that such an abnormality would be associated with an

inability to walk. However, none of the individuals with

TCF4 hemizygosity in our cohort were able to walk

regardless of the morphology of their corpus callosum.

This should not imply that all individuals with TCF4

deletions are never able to walk. There is anecdotal evi-

dence that some children with TCF4 are able to walk;

however, none of the individuals enrolled in this study had

attained that milestone.

Second, it could be anticipated that individuals with

larger deletions that include TCF4 might be more impaired

physically and mentally than those with smaller deletions.

However, we did not find this to be true. The presence or

absence of TCF4 seems to be more important in predicting

severity than the size of the deletion. We are fortunate to

have in this sample an individual with TCF4 hemizygosity

alone. As Table 5 illustrates, this person has significant

cognitive and motor delay. Given the very small number of

persons in each of our three sub-groups missing the TCF4

gene, it is not possible to determine if the differences in age

equivalent scores among these three deletion types are

statistically significant. Functionally, however, the cogni-

tive and motor delays across all groups are significant and

appear to be lifelong. In fact, in the cohort we evaluated,

TCF4 hemizygosity resulted in a developmental ceiling of

12 months irrespective of chronologic age, which ranged

from 10 to 238 months.

Third, the analysis of the ages and the age at death of

those with TCF4 hemizygosity and those with deletions of

chromosome 18 not including the TCF4 gene showed that

TCF4 hemizygosity conferred an increased risk of early

death. The cause of death was in most cases related to the

consequences of chronic aspiration.

The long-term goal of the Research Center is to deter-

mine which genes are responsible for which aspects of the

phenotype of 18q-. In this study, we analyzed the effect of

TCF4 hemizygosity to determine which components of the

Table 5 Average age equivalents of the three groups missing the TCF4 gene

Terminal deletion

N = 8

Large interstitial deletion

N = 5

Small interstitial deletion

N = 4

Deletion of only TCF4

N = 1

Cognitive abilities AE = 3.8 months of age AE = 2.4 months of age AE = 5 months of age AE = 11 months of age

Motor abilities AE = 3.4 months of age AE = 2.6 months of age AE = 6.25 months of age AE = 10 months of age

Fig. 3 Current age and age at death. The open circles indicate current

age and the black diamonds indicate the age at death. TCF4?/?; data

from 132 individuals with simple distal 18q deletions. Average

current age of those with two copies of TCF4 is 17 years. The two

individuals in this group who died were a female age 20 years,

6 months and a male 19 years, 9 months. TCF4?/- data from 22

subjects average current age is 11.5 years. Average age at death was

12 years old. The five participants who died included two males and

three females; ages 22 months, 6 years 10 months, 13 years, 20 years

11 months, and 22 years

Hum Genet

123

Author's personal copy

18q- phenotype can be attributed to the loss of one copy

of the gene. We compared the TCF4?/- group with the

TCF4?/? group using data from medical records, parental

questionnaires and in-person physical and behavioral

evaluations. Comparison to the literature was somewhat

hampered by the dearth of detailed phenotypic information

in previously reported cases of Pitt–Hopkins. Much of the

syndrome description is limited to dysmorphology, much

of which is imprecise, (e.g., ‘‘microcephaly’’), and medical

record abstraction. In particular, the behavioral and

developmental assessments are very limited in scope. This

made comparisons to the literature difficult. There is an

urgent need, now that the molecular genetics are known, to

perform a comprehensive multidisciplinary assessment of a

cohort of individuals with hemizygosity of or loss of

functions mutations in the TCF4 gene alone.

It is interesting to note that, in our population, the only

Pitt–Hopkins phenotypic components that were unique to

those with TCF4 hemizygosity were thin or absent corpus

callosum, small penis, accessory nipples, broad or clubbed

fingers, sacral dimple, short neck and wide spaced nip-

ples. Results of our analysis indicated that many of

the features of Pitt–Hopkins were present in both the

TCF4?/- as well as the TCF4?/? populations. Two of the

reported cardinal features of Pitt–Hopkins syndrome are

irregular breathing and seizures. The breathing abnor-

mality phenotype is poorly documented in the literature,

so we were not able to distinguish between documented

central apnea, episodes of heavy breathing and hyper-

ventilation. Interestingly, in our population this feature

was not unique to those with TCF4 hemizygosity. Like-

wise, seizures were not unique to the TCF4?/- group.

This is in part due to the fact that many of the features

such as hypotonia, single palmar crease, microcephaly,

and seizures are genetically heterogeneous and therefore

non-specific findings.

Alterations in the TCF4 gene have been implicated in

schizophrenia, and schizophrenia associated genes have

been linked to myelin-related pathways (Rietkerk et al.

2009). Since a key gene important in the compaction and

function of myelin (MBP) is located near the end of 18q,

and individuals with deletions of a critical region that

includes this gene have dysmyelination of the brain (Cody

et al. 2009), it might be postulated that hemizygosity of

both TCF4 and MBP would exacerbate their individual

effects. However, we did not see significant differences

between the TCF4 hemizygous participants whose 18q

deletion included the myelin basic protein gene (MBP)

(terminal deletions) and those that did not (interstitial

deletions) as shown in Table 5.

These data help us to make recommendations for this

unique group of individuals with 18q deletions. The cause

of death data highlights the need for aggressive detection

and intervention to prevent aspiration. Our data also

highlight the deleterious and chronic impact that hemizy-

gosity of the TCF4 gene has on cognitive and behavioral

development. In our study, all individuals hemizygous for

the TCF4 gene regardless of their age were similar to

typically developing babies less than 12 months old. It is

critical that parents plan for the long-term 24-h care their

children will need and refocus their developmental

expectations. It is important to provide a nurturing envi-

ronment that is rich in sensory stimulation and is one where

the child can feel comfortable and secure.

Of note, we are now faced with the challenge of

devising a meaningful nomenclature for 18q- that conveys

both the information about the genotype and as well as its

implications for phenotype. In this study, we used the

mouse nomenclature as a guide, since our goal is to merely

indicate gene copy number. We are indicating the diploid

state with regard to the TCF4 gene as TCF4?/? and the

hemizygous state as TCF4?/-. However, as we are able

to classify more genes as being either haplosufficient or

haploinsufficient, it will become a challenge for the

molecular cytogeneticist to write a clinically meaningful

karyotype. We do not envision a diagnostic code for

someone with a segmental deletion to include a long list of

all the hemizygous genes, but rather an edited list of those

genes that are haploinsufficient, i.e. have clinical signifi-

cance. Ultimately such a genotype would imply a particular

phenotype and thereby direct a plan of medical surveillance

and therapy.

Lastly, these data reinforce the need to eliminate the

word ‘‘syndrome’’ when referring to 18q- for two reasons.

First, rather than being defined by a constellation of fea-

tures, this condition is defined by a genotype, as implied by

the fact that it is named after the type of chromosome

aberration. Second, this particular chromosome abnormal-

ity is uniquely heterogeneous. No two unrelated individuals

have the same exact region of hemizygosity. Therefore, we

could identify numerous pairs of individuals who both have

18q- yet have no hemizygous genes in common (Heard

et al. 2009). The term ‘‘syndrome’’, which implies a col-

lection of similar phenotypic findings attributable to the

same genetic cause, is thus not appropriate. Rather, the data

presented here begin to define the molecular basis of 18q-,

highlighting the role of genomic heterogeneity in creating

phenotypic heterogeneity. Thus, the word ‘‘syndrome’’ is

no longer appropriate.

Acknowledgments The authors would like to first thank the fami-

lies that participated in this study for their willingness to share their

knowledge and for answering numerous questionnaires and emails.

This work was funded by the MacDonald family, The Chromosome

18 Registry & Research Society, the Institute for the Integration of

Medicine and Science (UL 1RR025767; National Center for Research

Resources) and CHRISTUS Santa Rosa Children’s Hospital.

Hum Genet

123

Author's personal copy

References

Amiel J, Rio M, de Pontual L, Redon R, Malan V, Boddaert N, Plouin

P, Carter NP, Lyonnet S, Munnich A, Colleaux L (2007)

Mutations in TCF4, encoding a class I basic helix-loop-helix

transcription factor, are responsible for Pitt–Hopkins syndrome,

a severe epileptic encephalopathy associated with autonomic

dysfunction. Am J Hum Genet 80(5):988–993

Andrieux J, Lepretre F, Cuisset AG, Delobel B, Manouvrier-Hanu S,

Holder-Espinasee M (2008) Deletion 18q21.2q21.32 involving

TCF4 in a boy diagnosed by CGH-array. Eur J Med Genet

51:172–177

Bayley N (2006) Bayley scales of infant and toddler development, 3rd

edn. Harcourt Assessment Inc., San Antonio

Brockschmidt A, Todt U, Ryu S, Hoischen A, Landwehr C, Birnbaum

S, Frenck W, Radlwimmer B, Lichter P, Engels H, Driever W,

Kubisch C, Weber RG (2007) Severe mental retardation with

breathing abnormalities (Pitt–Hopkins syndrome) is caused by

haploinsufficiency of the neuronal bHLH transcription factor

TCF4. Hum Mol Genet 16(12):1488–1494

Cody JD, Heard PL, Crandall AC, Carter EM, Li J, Hardies LJ,

Lancaster J, Perry B, Stratton RF, Sebold C, Schaub RL, Soileau

B, Hill A, Hasi M, Fox PT, Hale DE (2009) Narrowing critical

regions and determining penetrance for selected 18q- pheno-

types. Am J Med Genet 149A(7):1421–1430

de Pontual L, Mathieu Y, Golzio C, Rio M, Malan V, Boddaert N,

Soufflet C, Picard C, Durandy A, Dobbie A, Heron D, Isidor B,

Motte J, Newburry-Ecob R, Pasquier L, Tardieu M, Viot G,

Jaubert F, Munnich A, Colleaux L, Vekemans M, Etchevers H,

Lyonnet S, Amiel J (2009) Mutational, functional, and expres-

sion studies of the TCF4 gene in Pitt–Hopkins syndrome. Hum

Mutat 30:669–676

Elliot CD (2007) Differential ability scales, 2nd edn. Harcourt

Assessment Inc., San Antonio

Friedman JI, Vrijenhoek T, Markx S, Janssen IM, Vliet WA, Faas

BHW, Knoers NV, Kahn W, Kahn RS, Edelman L, Davis KL,

Silverman JM, Brunner HG, Kessel AG, Wijmenga C, Ophoff

RA, Veltman JA (2008) CNTNAP2 gene dosage variation is

associated with schizophrenia and epilepsy. Mol Psychiatry

13:261–266

Gilliam JE (1995) Gilliam autism rating scale, 1st edn. Pro-Ed Inc.,

Austin

Gilliam JE (2006) Gilliam autism rating scale, 2nd edn. Pro-Ed Inc.,

Austin

Giurgea I, Missirian Ch, Cacciagli P, Whalen S, Fredriksen T, Gaillon

T, Rankin J, Mathieu-Dramard M, Morin G, Martin-Coignard D,

Dubourg C, Chabrol B, Arfi J, Giuliano F, Lambert JC, Philip N,

Sarda P, Villard L, Goossens M, Moncla A (2008) TCF4deletions in Pitt–Hopkins syndrome. Hum Mutat 1026(29):

E242–E251

Guillemot F (2007) Spatial and temporal specification of neural

fates by transcription factor codes. Development 134:3771–

3780

Heard PL, Carter EM, Crandall AC, Sebold C, Hale DE, Cody JD

(2009) High resolution genomic analysis of 18q- using oligo-

microarray comparative genomic hybridization (aCGH). Am J

Med Genet Part A 149A(7):1431–1437

Kalscheuer VM, Feenstra I, Van Ravenswaaij-Arts CMA, Smeets

DFCM, Menzel C, Ullmann R, Musnate L, Ropers HH (2008)

Disruption of the TCF4 gene in a girl with mental retardation but

without the classical Pitt–Hopkins syndrome. Am J Med Genet

Part A 146A:2053–2059

Kirov G, Rujescu D, Ingason A, Collier DA, O’Donovan MC, Owen

MJ (2009) Neurexin 1(NRXN1) deletions in schizophrenia.

Schizophr Bull 35:851–854

O’Donnell L, Soileau B, Heard P, Carter E, Sebold C, Gelfond J, Hale

DE, Cody JD (2010) Genetic determinants of autism in

individuals with deletions of 18q. Hum Genet 128:155–164

Peippo MM, Simola KO, Valanne LK, Larsen AT, Kahkonen M,

Auranen MP, Ignatius J (2006) Pitt–Hopkins syndrome in two

patients and further definition of the phenotype. Clin Dysmor-

phol 15:47–54

Pitt D, Hopkins I (1978) A syndrome of mental retardation, wide

mouth and intermittent overbreathing. Aust Paediatr J 14:182–

184

Reynolds CR, Kamphaus RW (2004) Behavior assessment system for

children, 2nd edn. AGS Publishing, Circle Pines

Rietkerk T, Boks MPM, Sommer IEC, Jong DS, Kahn RS, Ophoff RA

(2009) Network analysis of positional candidate genes of

schizophrenia highlights myelin-related pathways. Mol Psychi-

atry 14:353–355

Rose MF, Een J, Ahmad KA, Chao HT, Klisch TJ, Flora A, Greer JJ,

Zoghbi HY (2009) Math1 is essential for the development of

hindbrain neurons critical for perinatal breathing. Neuron

64:341–354

Rosenfeld JA, Leppig K, Ballif BC, Thiese H, Erdie-Lalena C, Bawle

E, Sastry S, Spence E, Bandholz A, Surti U, Zonana J, Keller K,

Meschino W, Bejjani BA, Torchia BS, Shaffer LG (2009)

Genotype-phenotype analysis of TCF4 mutations causing Pitt–

Hopkins syndrome shows increased seizure activity with

missense mutations. Genet Med 11:797–805

Singh HA (1993) Mental retardation, macrostomia and hyperpnoea

syndrome. J Paediatr Child Health 29:156–157

Sparrow SS, Cicchetti DV, Balla DA (2005) Vineland adaptive

behavior scales, 2nd edn. AGS Publishing, Circle Pines

Stefansson H, Ophoff RA, Steinberg S, Andreassen OA, Cichon S,

Rujescu D, Werge T, Pietilainen OPH, Mors O, Mortensen PB,

Sigurdsson E, Gustaffson O, Nyegaard M, Tuulio-Henrikson A,

Ingason A, Hansen T, Suvisari J, Lonnqvist J, Paunio T,

Børglum AD, Hartmann A, Fink-Jensen A, Nordentoft M,

Hougaard D, Norgaard-Pedersen B, Bottcher Y, Olesen J, Breuer

R, Moller HJ, Giegling I, Rasmussen HB, Timm S, Mattheisen

M, Bitter I, Rethelyi JM, Magnusdottir BB, Sigmundsson T,

Olason P, Masson G, Gulcher JR, Haraldsson M, Fossdal R,

Thorgeirsson TE, Thorsteinsdottir U, Ruggeri M, Tosato S,

Franke B, Strengman E, Kiemeney LA, Genetic Risk and

Outcome in Psychosis (GROUP), Melle I, Djurovic S, Abramova

L, Kaleda V, Sanjuan J, de Frutos R, Bramon E, Vassos E, Fraser

G, Ettinger U, Picchioni M, Walker N, Toulopoulou T, Need

AC, Ge D, Yoon JL, Shianna KV, Freimer NB, Cantor RM,

Murray R, Kong A, Golimbet V, Carracedo A, Arango C, Costas

J, Jonsson EG, Terenius L, Agartz I, Petursson H, Nothen MM,

Rietschel M, Matthews PM, Muglia P, Peltonen L, St Clair D,

Goldstein DB, Stefansson K, Collier DA (2009) Common

variants conferring risk of schizophrenia. Nature 460(7256):744–

747

Taddeucci G, Bonuccelli A, Mantellassi I, Orsini A, Tarantino E

(2010) Pitt-Hopkins syndrome: report of a case with a TCF4gene mutation. Italian J Pediatr 36:12

Takano K, Lyons M, Moyes C, Jones J, Schwartz CE (2010) Two

percent of patients suspected of having Angelman syndrome

have TCF4 mutations. Clin Genet 78:282–288

Van Balkom ID, Quartel S, Hennekam RC (1998) Mental retardation,

‘‘coarse’’ face, and hyperbreathing: confirmation of the Pitt–

Hopkins syndrome. Am J Med Genet 75:273–276

Wechsler D (1997) Wechsler adult intelligence scales, 3rd edn. The

Psychological Corporation, San Antonio

Zweier C, Peipoo MM, Hoyer J, Sousa S, Bottani A, Clayton-Smith J,

Reardon W, Saraiva J, Cabral A, GOhring I, Devriendt K, Ravel

de Th, Bijlsma EK, Hennekam RCM, Orrico A, Cohen M,

Hum Genet

123

Author's personal copy

Dreweke A, Reis A, NUrnberg P, Rauch A A et al (2007)

Haploinsufficiency of TCF4 causes mental retardation with

intermittent hyperventilation (Pitt–Hopkins Syndrome). Am J

Hum Genet 80:994–1001

Zweier C, Sticht H, Bijlsma EK, Clayton-Smith J, Bonnen SE, Fryer

A, Greally MT, Hoffman L, Hollander NS, Jongmans M, Kant

SG, King MD, Lynch SA, McKee S, Midro AT, Park SM,

Ricotti V, Tarantino E, Wessels M (2008) Further delineation of

Pitt-Hopkins syndrome: phenotypic and genotypic description of

16 novel patients. J Med Genet 45:738–744

Zweier C, Jong EK, Zweier M, Orrico A, Ousager LB, Collins AL,

Bijlsma EK, Oortveld MAW, Ekici A, Reis A, Schneck A,

Rauch A (2009) CNTNAP2 and NRXN1 are mutated in

Autosomal-Recessive Pitt-Hopkins-like mental retardation and

determine the level of a common synaptic protein in Drosophila.

Am J Med Genet 85:655–666

Hum Genet

123

Author's personal copy