Genome Structure and Cognitive Map of Williams Syndrome.

Abstract

* Williams syndrome (WMS) is a most compelling model of human cognition, of human genome organization, and of evolution. Due to a deletion in chromosome band 7q11.23, subjects have cardiovascular, connective tissue, and neurodevelopmental deficits. Given the striking peaks and valleys in neurocognition including deficits in visual-spatial and global processing, preserved language and face processing, hypersociability, and heightened affect, the goal of this work has been to identify the genes that are responsible, the cause of the deletion, and its origin in primate evolution. To do this, we have generated an integrated physical, genetic, and transcriptional map of the WMS and flanking regions using multicolor metaphase and interphase fluorescence in situ hybridization (FISH) of bacterial artificial chromosomes (BACs) and P1 artificial chromosomes (PACs), BAC end sequencing, PCR gene marker and microsatellite, large-scale sequencing, cDNA library, and database analyses. The results indicate the genomic organ ization of the WMS region as two nested duplicated regions flanking a largely single-copy region. There are at least two common deletion breakpoints, one in the centromeric and at least two in the telomeric repeated regions. Clones anchoring the unique to the repeated regions are defined along with three new pseudogene families. Primate studies indicate an evolutionary hot spot for chromosomal inversion in the WMS region. A cognitive phenotypic map of WMS is presented, which combines previous data with five further WMS subjects and three atypical WMS subjects with deletions; two larger (deleted for D7S489L) and one smaller, deleted for genes telomeric to FZD9, through LIMK1 , but not WSCR1 or telomeric. The results establish regions and consequent gene candidates for WMS features including mental retardation, hypersociability, and facial features. The approach provides the basis for defining pathways linking genetic underpinnings with the neuroanatomical, functional, and behavioral consequences that result in human cognition.

INTRODUCTION

Williams syndrome (WMS) is one of the most compelling models of human cognition. Given the emerging grasp of the human genome, study of subjects with WMS provides the opportunity to elucidate the pathways that lead from genes to behavior in the cognitive neurosciences. This understanding may ultimately help to shed light on our evolutionary origins and to elucidate a part of what makes us human.

WMS is a particularly powerful model, because it is characterized by specific deficits coupled with remarkably preserved abilities. It provides the opportunity to probe neurocognitive pathways in humans across different levels, including the cellular, physiological, anatomic, functional, and cognitive, with each of these ultimately related to the underlying changes in gene expression. Armed with this information, one can then begin to infer the interconnections and to understand the molecular basis of human behavior. In this report, the approach to defining the genetic, anatomic, and neurocognitive basis of WMS will be presented along with a physical map and the genomic structure of WMS region. These data will be used to generate a phenotypic map of WMS and to define subsets of genes responsible for a part of the facial features and mental retardation.

Physical Features and Neurocognition in WMS

WMS is a rare genetic disorder that occurs in about one in 20,000 births and is characterized by mental retardation, a hoarse voice, transient neonatal hypercalcemia, and a set of facial and physical features that includes cardiovascular defects, typically congenital supravalvular aortic stenosis (Table 1; Morris, Leonard, & Dilates, 1988; Morris, Loker, Ensing, & Stock, 1993). However, what is most striking about individuals with WMS is their unique cognitive profile. The power of WMS as a model for dissecting cognition lies in the distinct pattern of abilities and deficits. Individuals with WMS perform relatively well on tasks involving language and face processing, but show extreme difficulties on other aspects of spatial processing and auditory processing, particularly at the level of global organization (Bellugi, Wang, & Jernigan, 1994; Bellugi, Klima, & Wang, 1996; Bellugi, Lichtenberger, Mills, Galaburda, & Korenberg, 1999a; Bellugi, Mills, Jernigan, Hickok, & Galaburda, 1999b; Bellugi, Lichtenberger, Jones, Lai, & St. George, this volume). WMS is also associated with hyperacusis, an abnormal sensitivity to sound (Neville et al., 1994), although this is not linked to abnormalities in the peripheral auditory system. Finally, the prime characteristic of WMS individuals is a strong impulse toward social contact and affective expression (Jones et al., this volume; Bellugi & Wang, 1998; Bellugi, Losh, Reilly, & Anderson, 1998; Bellugi et al., 1999a; Reilly, Klima, & Bellugi, 1990) as well as a heightened sensitivity to music (Levitin & Bellugi, 1998).

The neurobiological profile of WMS is being revealed through the studies of brain function, structure, and cytoarchitechtonics. In these domains, specific event-related potentials (ERPs) have been defined as markers for aspects of face and language processing in WMS (Mills et al., this volume; Bellugi et al., 1999b; Neville, Mills, & Bellugi, 1994). Neuroanatomical studies employing MRI and histomorphometric approaches have revealed consistent morphological features of decreased overall cerebral cortical volume; spared limbic structures of the temporal lobe including the amygdala, hippocampus, and parahippocampal gyrus; larger neocerebellar vs. paleocerebellar lobules (Jernigan & Bellugi, 1994); and preservation in volume of Heschl's gyrus, an area in the primary auditory cortex (Reiss et al., this volume; Bellugi et al., 1999b; Galaburda, Wang, Bellugi, & Rossen, 1994; Galaburda & Bellugi, this volume). These features have been related to both the functional abnormalities and their possible embryological or igins. Using the approach described, it is these neurocognitive and neurobiological features that will ultimately be combined with the genetic structure of WMS individuals to define the genes and pathways responsible.

Genetics of WMS

WMS is generally associated with a 1-2 Mb deletion of chromosome band 7q11.23 that includes the genes for elastin (Robinson et al., 1996; Gilbert-Dussardier et al., 1995; Nickerson, Greenberg, Keating, McCaskill, & Schaffer, 1995; Ewart et al., 1993), which is responsible for the congenital heart disease (Ewart, Jin, Atkinson, Morris, & Keating, 1994; Curran et al., 1993), LIM-kinase 1, which may contribute in part to the spatial deficit (Frangiskakis et al., 1996), and a growing number of other genes mapping in the region (illustrated in Figure 5). However, although WMS is clearly caused by the direct and downstream effects of genes located within the commonly deleted region, specific genes responsible for the major neurocognitive and physical features of WMS remain unknown. Gene candidates and transcripts mapping in the region include: WSCR1-5 (Osborne et al., 1996); RFC2, the replication factor C subunit 2 (Peoples, Perez-Jurado, Wang, Kaplan, & Francke, 1996); FZD9, the human frizzled homolog of the Droso phila wnt receptor (Wang et al., 1997); STX1A, the syntaxin 1A gene (Osborne et al., 1997b); GTF2I, general transcription factor 21 (Perez-Jurado et al., 1998); WS-[beta]TRP (WS-beta transducin repeats protein, ubiquitously expressed); WS-bHLH (a novel gene of the basic helix-loop-helix leucine zipper family of transcription factors that bind to E boxes, predominantly expressed in liver and kidney); BCL7B, a novel ubiquitously expressed gene (Meng et al., 1998); WSTF, WMS transcription factor (Lu, Meng, Morris, & Keating, 1998); FKBP6 (FK-506 binding protein, ubiquitously expressed); CYLN2 (cytoplasmic linker-2 gene encoding the protein CLIP-115) (Hoogenraad et al., 1998); CPE-R (Clostridium perfringens enterotoxin receptor); and RVP1 (rat ventral prostate protein 1) (Paperna, Peoples, Wang, & Francke, 1998). In addition, the gene for NCF1 (neutrophil cytosolic factor 1) is located close to but is not deleted in WMS (Francke et al., 1990). It is not known which of these genes contributes to the phenotypic fea tures of WMS. Using analyses of WMS subjects with atypical deletions, this report will illustrate the assignment of parts of the cognitive phenotype to the subsets of these genes.

Of importance for understanding both the cause and the phenotypic variability of WMS is the existence of genomic duplications that flank the largely unique region containing the genes described above. Both expressed genes and pseudogenes, as well as the breakpoints in the common WMS deletion, are thought to be located in these duplicated regions (Meng et al., 1998; Perez-Jurado et al., 1998; Osborne et al., 1997a; Korenberg et al., 1996; Robinson et al., 1996). However, the number of expressed genes in these duplicated regions and their contribution to the WMS phenotype is unknown. The current report will illustrate the structure of the flanking duplications and suggest a possible relationship to the WMS breakpoints.

In the current work, we will describe the construction of a working physical map of the WMS region including the flanking duplications. Combining data from fluorescence in situ hybridization (FISH), BAC end sequencing and large-scale sequence analyses, we will illustrate that the genomic organization of the WMS region includes two nested sets of duplications, the inner of which is involved in generating the common WMS deletion. A second gene for NCF1 will be shown to exist within the region deleted in some subjects with WMS. The genes for CPE-R and RVP1 have been assigned to single BACs. Three further gene fragments will be described and used to define the structure of the duplicated regions: ATRA (autoimmune thyroid antigen), a pseudogene for prohibitin that is located within an intron of the gene for GTF2I, and a novel gene that is duplicated but not deleted in WMS. Finally, using FISH and polymorphic marker analyses, we will describe the breakpoints in 78 individuals with WMS, two with larger deletions an d one with a smaller deletion, and use this information to generate a phenotypic map of WMS, defining regions and genes likely to contribute independently to the mental retardation and to the facial features.

Data that describe progress in four areas will be presented below:

(a) The development of a physical map of the WMS region.

(b) Chromosomal breakpoints in WMS deletions.

(c) Evolution and human variation in the WMS region.

(d) The development of a phenotypic cognitive map of WMS.

RESULTS

The Development of a Physical Map of the WMS Region: Chromosome Band 7q11.23

Using an approach …