Annual Review of Genomics and Human Genetics - Volume 8, 2007

The invitation to write the prefatory article to this volume of the Annual Review of Genomics and Human Genetics inspired me to collect some thoughts, a few involving ideas that are not new, but perhaps worth resurrecting in light of recent observations made with the data emerging from the Human Genome Diversity Project (HGDP). Data from the many relevant studies based on the HGDP have been made public, as was originally the hope and plan of the project. Here I try to give a short summary of the evolution of modern humans, a unique species in many respects but of special interest to readers of this volume, and a few thoughts on the general rates of evolution that might be relevant to medical genetics and genetic epidemiology. I have made no attempt to give a general bibliography, not even of results from the HGDP, since most authors' conclusions are still unpublished. Citations are limited to very few general concepts and articles discussed in this preface.

Gene duplication is one of the key factors driving genetic innovation, i.e., producing novel genetic variants. Although the contribution of whole-genome and segmental duplications to phenotypic diversity across species is widely appreciated, the phenotypic spectrum and potential pathogenicity of small-scale duplications in individual genomes are less well explored. This review discusses the nature of small-scale duplications and the phenotypes produced by such duplications. Phenotypic variation and disease phenotypes induced by duplications are more diverse and widespread than previously anticipated, and duplications are a major class of disease-related genomic variation. Pathogenic duplications particularly involve dosage-sensitive genes with both similar and dissimilar over- and underexpression phenotypes, and genes encoding proteins with a propensity to aggregate. Phenotypes related to human-specific copy number variation in genes regulating environmental responses and immunity are increasingly recognized. Small genomic duplications containing defense-related genes also contribute to complex common phenotypes.

Each day tens of thousands of DNA single-strand breaks (SSBs) arise in every cell from the attack of deoxyribose and DNA bases by reactive oxygen species and other electrophilic molecules. DNA double-strand breaks (DSBs) also arise, albeit at a much lower frequency, from similar attacks and from the encounter of unrepaired SSBs and possibly other DNA structures by DNA replication forks. DSBs are also created during normal development of the immune system. Defects in the cellular response to DNA strand breaks underpin many human diseases, including disorders associated with cancer predisposition, immune dysfunction, radiosensitivity, and neurodegeneration. Here we provide an overview of the genetic diseases associated with defects in the repair/response to DNA strand breaks.

Reading abilities are acquired only through specific teaching and training. A significant proportion of children fail to achieve these skills despite normal intellectual abilities and an appropriate opportunity to learn. Difficulty in learning to read is attributable to specific dysfunctions of the brain, which so far remain poorly understood. However, it is recognized that the neurological basis for dyslexia, or reading disability, is caused in large part by genetic factors. Linkage studies have successfully identified several regions of the human genome that are likely to harbor susceptibility genes for dyslexia. In the past few years there have been exciting advances with the identification of four candidate genes located within three of these linked chromosome regions: DYX1C1 on chromosome 15, ROBO1 on chromosome 3, and KIAA0319 and DCDC2 on chromosome 6. Functional studies of these genes are offering new insights about the biological mechanisms underlying the development of dyslexia and, in general, of cognition.

RNA interference (RNAi) can mediate the long- or short-term silencing of gene expression at the DNA, RNA, and/or protein level. Although several triggers of RNAi have been identified, the best characterized of these are small interfering RNAs (siRNAs), which can decrease gene expression through mRNA transcript cleavage, and endogenous microRNAs (miRNAs), which primarily inhibit protein translation. An improved understanding of RNAi has provided new, powerful tools for conducting functional studies in a gene-specific manner. In various applications, RNAi has been used to create model systems, to identify novel molecular targets, to study gene function in a genome-wide fashion, and to create new avenues for clinical therapeutics. Here, we review many of the ongoing applications of RNAi in mammalian and human systems, and discuss how advances in our knowledge of the RNAi machinery have enhanced the use of these technologies.

Fragile X syndrome is the most common form of inherited mental retardation. The disorder is mainly caused by the expansion of the trinucleotide sequence CGG located in the 5′ UTR of the FMR1 gene on the X chromosome. The abnormal expansion of this triplet leads to hypermethylation and consequent silencing of the FMR1 gene. Thus, the absence of the encoded protein (FMRP) is the basis for the phenotype. FMRP is a selective RNA-binding protein that associates with polyribosomes and acts as a negative regulator of translation. FMRP appears to play an important role in synaptic plasticity by regulating the synthesis of proteins encoded by certain mRNAs localized in the dendrite. An advancing understanding of the pathophysiology of this disorder has led to promising strategies for pharmacologic interventions.

Artificial selection has created myriad breeds of domestic animals, each characterized by unique phenotypes pertaining to behavior, morphology, physiology, and disease. Most domestic animal populations share features with isolated founder populations, making them well suited for positional cloning. Genome sequences are now available for most domestic species, and with them a panoply of tools including high-density single-nucleotide polymorphism panels. As a result, domestic animal populations are becoming invaluable resources for studying the molecular architecture of complex traits and of adaptation. Here we review recent progress and issues in the positional identification of genes underlying complex traits in domestic animals. As many phenotypes studied in animals are quantitative, we focus on mapping, fine mapping, and cloning of quantitative trait loci.

In humans, genetic factors have long been suspected to contribute to the onset and outcome of tuberculosis. Such effects are difficult to identify owing to their complex inheritance, and to the confounding impact of environmental factors, notably pathogen-associated virulence determinants. Recently, forward genetic approaches in mouse models and in human populations have been used to elucidate a molecular basis for predisposition to mycobacterial diseases. The genetic dissection of host predisposition to infection with Mycobacterium bovis BCG and M. tuberculosis will help to define the key molecules involved in host antituberculous immunity and should provide new insights into this important infectious disease.

Multi-sequence alignments of large genomic regions are at the core of many computational genome-annotation approaches aimed at identifying coding regions, RNA genes, regulatory regions, and other functional features. Such alignments also underlie many genome-evolution studies. Here we review recent computational advances in the area of multi-sequence alignment, focusing on methods suitable for aligning whole vertebrate genomes. We introduce the key algorithmic ideas in use today, and identify publicly available resources for computing, accessing, and visualizing genomic alignments. Finally, we describe the latest alignment-based approaches to identify and characterize various types of functional sequences. Key areas of research are identified and directions for future improvements are suggested.

Over the past five years, the importance of a diverse class of 18–24 nucleotide RNA molecules, known as microRNAs (miRNAs) has increasingly been recognized. These highly conserved RNAs regulate the stability and translational efficiency of complementary target messenger RNAs. The human genome is now predicted to encode nearly 1,000 miRNAs that likely regulate at least one third of all human transcripts. Despite rapid progress in miRNA discovery, the physiologic functions of only a small number have been definitively established. In this review, we discuss the principles of miRNA function that have emerged from the studies performed thus far in vertebrates. We also discuss known and potential roles for miRNAs in human disease states and discuss the influence of human genetic variation on miRNA-mediated regulation.

Eukaryotic genomes contain vast amounts of repetitive DNA derived from transposable elements (TEs). Large-scale sequencing of these genomes has produced an unprecedented wealth of information about the origin, diversity, and genomic impact of what was once thought to be “junk DNA.” This has also led to the identification of two new classes of DNA transposons, Helitrons and Polintons, as well as several new superfamilies and thousands of new families. TEs are evolutionary precursors of many genes, including RAG1, which plays a role in the vertebrate immune system. They are also the driving force in the evolution of epigenetic regulation and have a long-term impact on genomic stability and evolution. Remnants of TEs appear to be overrepresented in transcription regulatory modules and other regions conserved among distantly related species, which may have implications for our understanding of their impact on speciation.

Congenital disorders of glycosylation (CDG) are a large family of genetic diseases resulting from defects in the synthesis of glycans and in the attachment of glycans to other compounds. These disorders cause a wide range of human diseases, with examples emanating from all medical subspecialties. Since our 2001 review on CDG (36), this field has seen substantial growth: The number of N-glycosylation defects has doubled (from 6 to 12), five new O-glycosylation defects have been added to the two previously known ones, three combined N- and O-glycosylation defects have been identified, the first lipid glycosylation defects have been discovered, and a new domain, that of the hyperglycosylation defects, has been introduced. A number of CDG are due to defects in enzymes with a putative glycosyltransferase function. There is also a growing group of patients with unidentified defects (CDG-x), some with typical clinical presentations and others with presentations not seen before in CDG. This review focuses on the clinical, biochemical, and genetic characteristics of CDG and on advances expected in their future study and clinical management.

Noncoding RNA genes produce a functional RNA product rather than a translated protein. More than 1500 homologs of known “classical” RNA genes can be annotated in the human genome sequence, and automatic homology-based methods predict up to 5000 related sequences. Methods to predict novel RNA genes on a whole-genome scale are immature at present, but their use hints at tens of thousands of such genes in the human genome. Messenger RNA-like transcripts with no protein-coding potential are routinely discovered by high-throughput transcriptome analyses. Meanwhile, various experimental studies have suggested that the vast majority of the human genome is transcribed, although the proportion of the detected RNAs that is functional remains unknown.

A postgenome challenge is to understand how the code in DNA is converted into the biological processes underlying various cell fates. By establishing the appropriate technical tools, we are moving from an era in which such questions have been asked by studying individual genes to one in which large domains, whole chromosomes, and the entire human genome can be investigated. These developments will allow us to study in parallel the transcriptional program and components of the epigenetic program (nuclear position, timing of replication, chromatin structure and modification, DNA methylation) to determine the hierarchy and order of events required to switch genes on and off during differentiation and development.

In the past, to study Mendelian diseases, segregating families have been carefully ascertained for segregation analysis, followed by collecting extended multiplex families for linkage analysis. This would then be followed by association studies, using independent case-control samples and/or additional family data. Recently, for complex diseases, the initial sampling has been for a genome-wide linkage analysis, often using independent sib-pairs or nuclear families, to identify candidate regions for follow-up with association studies, again using case-control samples and/or additional family data. We now have the ability to conduct genome-wide association studies using 100,000–500,000 diallelic genetic markers. For such studies we focus especially on efficient two-stage association sampling designs, which can retain nearly optimal statistical power at about half the genotyping cost. Similarly, beginning an association study by genotyping pooled samples may also be a viable option if the cost of accurately pooling DNA samples outweighs genotyping costs. Finally, we note that the sampling of family data for linkage analysis is not a practice that should be automatically discontinued.

Large-scale genomic databases are becoming increasingly common. These databases, and the underlying biobanks, pose several substantial legal and ethical problems. Neither the usual methods for protecting subject confidentiality, nor even anonymity, are likely to protect subjects’ identities in richly detailed databases. Indeed, in these settings, anonymity is itself ethically suspect. New methods of consent will need to be created to replace the blanket consent common to such endeavors, with a consent procedure that gives subjects some real control over what they might consider inappropriate use of their information and biological material. Through their use, these biobanks are also likely to yield information that will be of some clinical significance to the subjects, information that they should have access to. Failure to adjust to these new challenges is not only legally and ethically inappropriate, but puts at risk the political support on which biomedical research depends.