|
 The
study of genes and cognition has become an exciting field. However, genes
that significantly affect cognition and behavior have been notoriously
hard to locate within the human genome. Williams syndrome (WS) is a chromosome
deletion disorder with interesting behavioral and cognitive phenotypic
components, and the loss of genes within the WS deletion is responsible
for these phenotypic characteristics. Accordingly, the study of WS gives
us the opportunity to identify, first-hand, genes that influence behavior
and cognition. The identification of such genes will not only help us
understand the molecular basis of WS, but will also expand our knowledge
of how genes shape normal cognition and behavior.
WS
was first reported in 1961 by Williams, who described children with "unusual"
facial features in addition to growth retardation, supravalvular aortic
stenosis (SVAS), and mild mental retardation. Shortly afterward, another
report by Beuren described a similar series of children with these features
plus dental anomalies and friendly personalities. These characteristic
symptoms are present in virtually all individuals with WS, a relatively
rare disorder that is now known to be caused by a microdeletion on the
long arm of chromosome 7 at 7q11.23.
The
incidence of WS is estimated at approximately 1 per 20,000. It usually
occurs in a sporadic manner, although rare cases of autosomal dominant
transmission have been described. WS is associated with a recognizable
facies, including stellate irides, flat nasal bridge, short up-turned
nose with anteverted nostrils, long philtrum, full lips and lower cheeks,
and a small chin. Cardiovascular lesions are very characteristic and are
found in 80% of individuals. These present as a generalized arteriopathy
leading to vascular stenoses, most frequently of the ascending aorta (SVAS)
or peripheral pulmonary arteries, and also to hypertension. SVAS is the
major life-threatening component of the WS phenotype but can be corrected
with surgery. Other symptoms include dental problems such as malocclusion,
small and missing teeth; growth deficiency; hypersensitivity to sounds
such as sirens, vacuum cleaners, and thunder; hypercalcemia, vomiting,
constipation, and colic in infancy; musculoskeletal abnormalities; impaired
visual acuity; and a hoarse, low voice.
Individuals
with WS also tend to have mild mental retardation (average IQ is between
55 and 60). In addition to impaired cognition, patients show hyperreactivity,
sensory integration dysfunction, delayed expressive and receptive language
skills, and multiple developmental motor disabilities affecting balance,
strength, coordination, and motor planning. Approximately 70% of individuals
with WS suffer from attention-deficit/
hyperactivity disorder, and there is a high incidence of anxiety and simple
phobias. A striking aspect of the WS phenotype is the coexistence of an
anxiety disorder with a friendly, socially engaging personality (Pober
and Dykens, 1996).
The
most common cardiovascular lesion found in WS (SVAS) also exists as a
distinct autosomal dominant disease. In 1993, the elastin gene was implicated
in the pathogenesis of SVAS by the identification of a family with SVAS
and a disruption of the elastin (ELN) gene at 7q11. The deletion
of the same gene in two unrelated SVAS families and the subsequent identification
of point mutations in sporadic cases supports the hypothesis that mutations
in ELN are the cause of SVAS. Analysis of the region surrounding
ELN in patients with WS demonstrates that the majority of individuals
harbor a large deletion spanning approximately 1.5 Mb of DNA. This deletion
spans many genes that contribute to the additional clinical symptoms seen
in WS, compared with individuals with isolated SVAS.
Since
the WS deletion was first identified, various groups have built a framework
of genomic clones across the region and used these as the basis for gene
discovery. A total of 17 genes have now been shown to lie within the common
deletion, but none except ELN has been definitively shown to contribute
to any of the symptoms. The genes code for proteins that span a large
range of cellular functions, including some whose function remains unclear.
Efforts to link individual genes with specific parts of the WS clinical
picture have followed two paths: the study of individuals with atypical
deletions of the region and the study of the protein products themselves.
In
the vast majority of individuals with WS, the deletion breakpoints cluster
within a small stretch of DNA, resulting in the same-sized deletion. This
homogenous-sized deletion is thought to occur because of unequal crossover
during meiosis, a mechanism that has been shown to be responsible for
other deletion disorders, including DiGeorge syndrome, Smith-Magenis syndrome,
and some forms of neurofibromatosis type I. As a result of the presence
of a "common"deletion, genotype-phenotype correlation in WS
has been problematic. Groups throughout the world have been searching
for the rare individuals who harbor smaller deletions of the region, and
they have succeeded in identifying only a few. These individuals can be
separated into two distinct subgroups: individuals who have classic WS
but have shorter deletions and individuals who do not have classic WS
but have deletions involving the WS region.
Two
individuals with WS have been reported whose deletions are smaller than
the common one (Botta et al., 1999). Both children were from Italy and,
although young, they appeared to exhibit most of the main features of
WS (typical facies, cardiovascular abnormalities, cognitive impairment,
hyperactivity). The elder of the two, who was 6 years old upon examination,
showed the characteristic cognitive and behavioral profile, with more
pronounced deficits in visual-spatial skills and a friendly but anxious
personality. These children had a similar-sized deletion of the WS region
that shared a telomeric breakpoint with the common deletion but had a
unique centromeric breakpoint. At least six genes that were usually deleted
in WS were not deleted in these children, suggesting that these genes
do not contribute to the major features of WS.
There
have been nine identified in the other group of atypical individuals,
with aspects of the WS phenotype and deletions involving the WS region.
The deletions remove from 2 to 15 of the commonly deleted genes, but none
of the individuals has a classic WS phenotype. In an effort to try to
dissect the parts of the WS behavioral or cognitive profile, psychologists
have developed tests designed to pick up cognitive deficits that are characteristic
of WS, such as poor visuospatial skills (Mervis et al., 1999). It is postulated
that by reducing the complexity of the cognitive profile, a genotype–phenotype
relationship may be established for some aspects of WS.
The
picture derived from these smaller deletions is far from clear, however.
In 1996 two large families were identified with deletions involving only
ELN and a neighboring gene, LIM kinase 1 (LIMK1) (Frangiskakis
et al., 1996). The majority of individuals from these kindreds had SVAS,
as would be predicted because ELN was disrupted, but also showed
an impairment in visuospatial skills that was comparable with that seen
in WS. This finding led the authors to conclude that the LIMK1 protein
was intimately involved in proper visuospatial cognition. The role of
LIMK1 in neurons, as a key molecule in the cycle of building and
dismantling the actin cytoskeleton, supported this conclusion. However,
three additional individuals with deletions encompassing LIMK1
were identified in 1999, but none of them showed any visuospatial impairment
(by the same testing methods), refuting the LIMK1 hypothesis (Tassabehji
et al., 1999). The contribution of LIMK1 to WS is, therefore, still
under debate. It is clearly an excellent functional and positional candidate
for playing an important role in neuronal development and function, but
we await the results of further experiments, including an animal model,
to define its role in WS.
Taking
the deletion data as a whole, a preliminary map of the genotype–phenotype
relationships seems to be emerging (Fig.
1). This map suggests that the genes toward the telomeric end of the
deletion may play a larger role in the development of many of the classic
WS features, and particularly the cognitive profile. Firm conclusions,
however, cannot be based on the small number of reported individuals with
atypical deletions, particularly when there is the possibility of somatic
mosaicism, an occurrence that has recently been reported in DiGeorge syndrome.
Further
insight into the contribution of particular genes to the cognitive profile
might be gained by studying the deleted genes themselves, in an attempt
to identify functional candidates rather than positional candidates. Of
the 17 commonly deleted genes, the majority are expressed at some level
in the central nervous system. A few are expressed at high levels, or
exclusively in the brain, which makes them more attractive candidates,
but this does not exclude the remaining genes. Cytoplasmic linker 2 (CYLN2)
is found only in the brain, where it is thought to link specific organelles
within neurons to the cytoskeleton (De Zeeuw et al., 1997). How a reduction
in this protein could affect brain function is still unclear. Syntaxin
1A (STX1A) is another protein that is found almost exclusively in neurons,
and its function is better understood (Osborne et al., 1997). STX1A is
a key component of a protein complex that mediates the release of neurotransmitters
across the synapse, thus conveying chemical signals from one neuron to
another. Studies in model organisms such as flies and worms have shown
that the amount of this protein is critical to its proper function, suggesting
that a 50% reduction, as seen in WS, could cause a clinically relevant
phenotype.
Predictions
cannot be made about the function of many of the remaining genes, but
several are thought to be transcriptional regulators. Most are these are
widely expressed, but it is quite possible that reducing the level of
a particular transcription factor could have different effects in different
tissues, ranging from inconsequential to severe. We must also consider
the potentially additive effects of deleting many genes at once. This
may be particularly pertinent to a pair of genes adjacent to the telomeric
boundary of the deletion, whose protein products are similar in both structure
and function, which means that they may be at least partially functionally
redundant. Reducing the amount of either gene individually may not produce
a noticeable phenotype, but in combination the effect is considerable.
These genes, general transcription factor 2I (GTF2I) (Cheriyath
and Roy, 2000) and GTF2I repeat domain containing protein 1 (GTF2IRD1)
(Bayarsaihan and Ruddle, 2000), code for general transcription factors
that mediate the activation of transcription of a wide variety of genes
with the help of other spatially or temporally regulated transcription
factors.
Research
into the basic molecular function of each of the candidate genes will
provide valuable information, but perhaps the greatest insight may come
from the generation and study of animal models. Mouse models have become
powerful tools for investigating both the molecular and physiological
basis of human genetic disease. The mouse genome can be easily manipulated
to produce either single-gene knockouts or larger alterations that remove
several genes at once. This technology gives us the opportunity to engineer
WS deletions of choice, instead of relying on the rare atypical deletions
that we can identify in humans. In addition, inbred laboratory strains
of mice have homogenous genetic backgrounds, eliminating the influence
of genes outside the WS deletion region on development of the phenotype.
Many sophisticated behavioral analyses can be performed on mice in order
to assess their cognitive abilities, so we should be able to model at
least some of the WS cognitive and behavioral profile.
In
summary, there is still not a clear picture of the genetic basis for the
WS cognitive phenotype. No single gene can yet be excluded from a role,
however minor, in WS. It is likely that the cognitive impairment is the
result of the deletion of several genes, although specific components
of the impairment, such as visuospatial deficits, may be attributable
to a single gene. The correlation of specific genes from within the WS
deletion with cognitive impairment may bring to light some interesting
and unsuspected culprits that will give us entry points into novel biological
pathways.
top of page
|
