PE R S PE C TI V E on genes, speech, and language
On Genes, Speech, and Language
Simon E. Fisher, D.Phil.
Related article, page 1694
L earning to talk is one of the
most important milestones in
human development, but we still
clusionary manner (the presence
of speech or language problems
that cannot be explained by an
fine motor sequences that under-
lie speech. This large variety of
communication problems is re-
have only a limited understand- obvious medical condition) and flected by diagnostic schemes that
ing of the way in which the pro- so encompass a wide variety of contain several distinct categories
cess occurs. It normally takes just phenotypes.1 For example, lin- (expressive, mixed, phonologic,
a few years to go from babbling guistic deficits can be confined apraxic–dyspraxic, and so on). In
newborn to fluent communica- to expressive language or can ex- practice, there frequently are co-
tor. During this period, the child tend to receptive abilities, al- existing disorders involving vari-
learns to produce a rich array of though pure receptive impairment ous types of impairment, and the
speech sounds through intricate is seldom seen. When it comes to boundaries between these disor-
control of articulatory muscles, speech output, affected children ders can be fluid; a person may
assembles a vocabulary compris- may fail to produce sounds that move from one category to an-
ing thousands of words, and de- would be expected on the basis of other at various stages of life.
duces the complicated structural age and dialect, which may be Although the causes of these
rules that permit construction associated with difficulties in the kinds of disorders remain large-
of meaningful sentences. All of planning and execution of the ly elusive, familial clustering and
this (and more) is achieved with
little conscious effort.
The acquisition of language
usually proceeds along robust 21
lines without any need for explic- 15.3
it tuition, in stark contrast to oth- 14
er complex learned abilities, like 13
reading, writing, and mathemat- 12
ics. However, a small minority of 11.1 Centromere
children are unable to acquire 11.22
speech and language proficiency, 11.23
despite growing up in language- 21.1
rich environments and showing 21.3
adequate performance in other 22 q
areas, such as hearing and non- 31.1
verbal cognition. For some of
these children, early difficulties
with communication resolve with 32
age, but for others the problems 34
continue into adulthood. Since 36
modern society depends heavily
on language and literacy skills,
persistent impairment is often Chromosome 7 Gene Dosage and Speech and Language Disorder.
accompanied by wider problems An ideogram of chromosome 7 shows cytogenetic banding (with two regions of the
in educational, social, and emo- long q arm highlighted), a 600-kb region containing the FOXP2 gene, and a 1.5-Mb
interval that is commonly deleted in the Williams–Beuren syndrome (WBS). Point
tional development in later life. mutations and chromosomal abnormalities involving FOXP2 are one cause of develop-
Developmental communication mental verbal dyspraxia. Duplication of the WBS region may also be associated with a
disorders are diagnosed in an ex- delay in speech.
n engl j med 353;16 www.nejm.org october 20, 2005 1655
PE R S PE C T IV E on genes, speech, and language
twin-based heritability studies been hypothesized that reduced the consequences of deletion and
provide strong evidence of ge- amounts of functional FOXP2 duplication of this same interval
netic influences.1 At the same protein yield subtle anomalies appear to be strikingly different,
time, it is clear that the observed in neural organization that im- at least with respect to speech
phenotypic heterogeneity is un- pair speech and language acqui- articulation and expressive lan-
derpinned by a mixture of ge- sition. guage. Third, the present report
netic effects, which range from A report in this issue of the concerns only a single case, but
common risk alleles acting in a Journal (pages 1694–1701) presents the investigators speculate that
multifactorial framework to rare evidence that alteration in dosage duplication of the region could
instances of highly penetrant of genes elsewhere on the long occur at a frequency similar to
point mutations behaving in a arm of chromosome 7 may be that of the common WBS dele-
classic mendelian fashion.2 The another cause of speech and lan- tion, since it is mediated by a
dissection of such a complicated guage disturbances. Previously, related mechanism of interchro-
state of affairs calls on geneticists studies of a rare cytogenetic ab- mosomal recombination, involv-
to use multiple complementary normality known as supernumer- ing flanking low-copy-number
strategies of both a traditional ary ring chromosome 7 indicated repeats. As such, 7q11.23 dupli-
and a novel nature. By adopting that large duplications of 7q are cations may contribute to the in-
a variety of approaches, linkage associated with severely retarded cidence of unexplained deficits in
studies of prevalent types of speech development.5 In the cur- speech and language in human
speech and language disorders rent study, Somerville and col- populations.
have implicated several regions of leagues have identified a boy with The duplicated region in this
the genome, most notably on childhood apraxia of speech ac- case contains as many as 27
chromosomes 3, 13, 16, and 19. companied by dramatic disrup- genes, and it is unclear which of
The putative risk genes underly- tion of expressive language, a these genes are relevant to the
ing these linkages have yet to be disorder that appears to be disorder. Somerville et al. dem-
identified, but progress is en- caused by a 1.5-Mb duplication onstrate increased expression of
couraging. of 7q11.23. several of the duplicated genes
In the case of a rare domi- These findings are intriguing in lymphoblastoid cells, but re-
nant mendelian form of impair- on a number of counts. First, they lating this sort of finding to neu-
ment, it has been possible to go add support to the hypothesis rodevelopment could be difficult,
even further and home in on a that the development of neural particularly since neural tissue
specific gene, known as FOXP2, circuitry involved in speech and may tolerate changes in dosage
located in chromosomal band language acquisition can be high- of some genes but not others.
7q31 (see diagram). People who ly sensitive to gene copy num- Moreover, the genes whose in-
carry heterozygous disruption of ber. Although FOXP2-related dis- creased dosage leads to language
this gene have problems sequenc- order has thus far been linked disruption in the child carrying
ing the precise movements of to reduced functional dosage, the the duplication may not neces-
tongue, lips, jaw, and palate that present study shows that increased sarily be the same as the genes
contribute to intelligible speech levels of transcription of one or that yield an uneven linguistic
(known as verbal dyspraxia or more genes may also be an im- profile in WBS. A proper under-
childhood apraxia of speech).3,4 portant mechanism accounting standing of genotype–phenotype
They also have difficulties with for cases of speech and language relationships will require exten-
learning and production of non- disorder. Second, the duplicated sive future study and may de-
speech sequences involving the interval of 7q11.23 corresponds pend on discovery of partial du-
orofacial musculature (orofacial directly with the region most plications of the common WBS
dyspraxia) and have a broad pro- commonly deleted in the Wil- region, as well as investigation
file of linguistic deficits in ex- liams–Beuren syndrome (WBS), of 7q11.23 gene polymorphism
pressive and receptive domains a neurodevelopmental disorder in in cohorts of children with speech
— problems that affect both which language tends to be an and language disorders, particu-
oral and written language. It has area of relative strength. Thus, larly those with speech apraxia.
1656 n engl j med 353;16 www.nejm.org october 20, 2005
PE R S PE C TI V E on genes, speech, and language
We are just starting to unrav- bination with findings from 2. Fisher SE, Lai CSL, Monaco AP.
Deciphering the genetic basis of speech and
el the genetic causes of develop- other disciplines, such as neuro- language disorders. Annu Rev Neurosci
mental communication disorders imaging and developmental psy- 2003;26:57-80.
and explore how they relate to chology, hold promise for shed- 3. Vargha-Khadem F, Gadian DG, Copp A,
Mishkin M. FOXP2 and the neuroanatomy of
language problems in other neu- ding light on the mechanisms by speech and language. Nat Rev Neurosci
rodevelopmental syndromes, such which speech and language ac- 2005;6:131-8.
as autism and dyslexia. Success quisition can go awry. 4. MacDermot KD, Bonora E, Sykes N, et al.
Identification of FOXP2 truncation as a novel
in this area raises the possibility cause of developmental speech and lan-
Dr. Fisher is a Royal Society Research Fellow
of using molecular diagnostics at the Wellcome Trust Centre for Human Ge- guage deficits. Am J Hum Genet
for early identification of children netics, University of Oxford, Oxford, England. 2005;76:1074-80.
5. Lichtenbelt KD, Hochstenbach R, van
who are at increased risk, thus Dam WM, Eleveld MJ, Poot M, Beemer FA.
enabling environmental interven- 1. Bishop DVM. Genetic and environmental Supernumerary ring chromosome 7 mosa-
risks for specific language impairment in icism: case report, investigation of the gene
tions to begin at a younger age. children. Philos Trans R Soc Lond B Biol Sci content, and delineation of the phenotype.
Overall, genetic studies, in com- 2001;356:369-80. Am J Med Genet A 2005;132:93-100.
n engl j med 353;16 www.nejm.org october 20, 2005 1657