Ribosomal RNA

Ribosomal RNAs (rRNAs) are functional RNA transcripts that constitute one of the main components of cytoplasmic ribosomes. The genes coding for the major form of cytoplasmic rRNA are located in multiple copies on the short arms of the acrocentric chromosomes: 13, 14, 15, 21, and 22. They code for a single large 45S primary transcript that is cleaved into 28S, 18S, and 5.8S rRNA classes, designated by their separation in centrifugation gradients and by several associated proteins. Multiple copies of another gene on chromosome 1 produce 5S rRNA.

Transfer RNA

Transfer RNAs (tRNAs) are small RNA transcripts that bind specific amino acids and transport them to ribosomes for use during protein synthesis. More than 40 subfamilies of tRNA genes are known, dispersed across the genome. The mitochondrial genome uses a separate set of tRNAs.

Messenger RNA

Messenger RNAs (mRNAs) are the RNA transcripts of all genes that encode polypeptides and some other genes that encode unprocessed functional RNA molecules. Most are large. All mRNA transcripts undergo further processing, including excision of large segments of noncoding RNA known as introns, the addition of 7-methylguanosine to the first 5′ nucleotide, forming a CAP structure, cleavage of the 3′ end at a specific point downstream from the end of the coding sequence, and addition of the polyA tail at a site specified in part by the sequence AAUAAA, which is located in the 3′ untranslated region (3′ UTR) of the gene. The polyA tail appears to increase the stability of mRNA. The fully processed mRNA is transported to the cytoplasm, where translation occurs.

Small Nuclear RNA

Small nuclear RNA (snRNA) transcripts are small, uridine-rich RNA transcripts that associate with specific proteins to form ribonucleoprotein particles (RNPs). Some of them function in RNA splicing (removing introns from mRNA). They comprise a large family of genes dispersed across the genome.

Small Nucleolar RNA

Small nucleolar RNAs (snoRNAs) are small RNA transcripts that are present in the nucleolus and have important roles in specific cleavage reactions and base-specific modifications during maturation of ribosomal RNA. About 200 snoRNA genes have been identified.

MicroRNA

MicroRNAs (miRNAs) are another class of small noncoding genes that regulate the expression of protein-encoding genes at the post-transcriptional RNA level [Denli et al., 2004]. The process begins with transcription (synthesis) of primary RNA transcripts that range in size from several hundred to several thousand kb. These transcripts are recognized and cut into precursor miRNAs in the nucleus by a protein known as Dicer, moved to the cytoplasm, and processed into mature miRNAs. The mature miRNAs join the RNA-induced silencing complex (RISC), which recognizes and cleaves (or otherwise silences) a target gene. This process has been demonstrated in many organisms, including mammals, and appears likely to play a key role in regulation of many genes.

Genes

Genes are composed of a continuous length of DNA with definable start and end points, which include the sequence that codes for the RNA or polypeptide product and is thus known as the coding region. It has become clear, however, that the structure of a gene is complex and includes much more than the coding sequence of the protein. All genes include additional sequences on either end of the coding region – designated the 5′ and 3′ UTRs – that do not code for an RNA product or polypeptide. These regions function to regulate transcription and RNA stability. The gene is considered to include the entire sequence represented in the RNA product because some mutations within noncoding regions can impair gene function.

A model of a typical human gene is shown in Figure 30-3. Promoter sequences required for regulation and initiation of RNA transcription (red diamonds in Figure 30-3) are present at the 5′ end of the gene, such as the CAT and TATA boxes whose sequences are tightly conserved among many different genes and species. Downstream from the promoter sequences is a specific sequence that signals the start of transcription. A short way further downstream is an initiator codon, AUG, which codes for methionine. This triplet is the translation start site, which signals the start of the coding sequence for the polypeptide product. The region between the transcription and translation start sites is the 5′ UTR.

Fig. 30-3 The structure of a typical human gene.

The gene includes a primary regulatory region known as the promoter just upstream of the transcription start site that is required for binding of both DNA and RNA polymerases (red diamonds), as well as several types of distant regulatory elements that protect the gene from regulation of other nearby genes (insulator), increase or decrease gene expression (enhancers and silencers), or regulate several genes in the region (locus control region).

The next segment of the gene is the coding region. The coding regions of most genes in prokaryotes and lower eukaryotes are colinear, which means that the coding sequence corresponds exactly to the sequence of amino acids in the polypeptide. By contrast, most higher eukaryotic genes, including human genes, contain additional sequences that lie within the coding region, interrupting the sequence that represents the polypeptide. The regions that code for the final polypeptide (or functional RNA) product are known as exons, whereas the regions that are missing from the final mRNA product are introns. The removal of introns from the final mRNA product is known as splicing, a complex process that is regulated by a large number of proteins and functional RNA transcripts.

The coding sequence ends at one of three specific stop codons: UAA, UAG, or UGA. The last segment of the gene is the 3′ UTR, which contains a polyadenylation signal and presumably a signal to end transcription, although no transcription stop sequence has been identified. The length of a gene may vary, ranging from less than 1 kb to several hundred kb. The longest gene known, which codes for dystrophin, spans more than 2000 kb of genomic sequence, although this is not the largest protein produced in the cell.

Regulatory Regions

Many genes have highly conserved sequences, a longer distance upstream and downstream of the transcribed gene, that are involved in regulating expression, including enhancers, silencers, locus control regions, and insulators (see Figure 30-3). Enhancer elements function to increase gene expression, while silencers reduce gene expression. Locus control regions may regulate expression of several genes within a chromosome region, while insulators prevent co-regulation of more distant genes and gene regions. All of these are sequences that bind proteins called transcription factors, which can be ubiquitous, tissue-specific, and/or temporally expressed. Promoters are located immediately 5′ of the gene and bind to RNA polymerase II, a necessary step for transcription. Other transcription factors bind upstream of the promoter and activate transcription. Enhancers and silencers are often located at a distance from the promoter, and increase or decrease transcription in a tissue-specific or temporal manner. Overall, the transcription of each gene is tightly regulated, with multiple transcription factors involved.

Imprinting

The process by which certain genes in specific chromosomal regions are expressed from only one chromosome, depending on the parental origin of the chromosome, is known as “imprinting.” Although the mechanism is only partly understood, a key component involves allele-specific DNA methylation, found predominantly at the carbon 5 position of about 80 percent of all cytosines that are part of symmetrical cytosine-guanine (CpG) dinucleotides [Jiang et al., 2004; Strachan and Read, 2010; Weksberg et al., 2003].

This process is controlled by regulatory imprinting “centers,” located nearby on the same chromosome as that of the silenced or “imprinted” gene. In effect, then, two alleles of the same gene that are identical in nucleotide sequence but derived from opposite parents are regulated differently in the same nucleus. This process is reversible, so that the silent, imprinted allele can be reactivated and the active allele silenced when passed through the germline of the opposite-sex parent. Most imprinted genes are found in large clusters of greater than 1 Mb (megabase pairs) in length. Imprinted clusters have been identified in chromosomes 6q24, 7p11.2, 11p15.5, 14q32, 15q11–q13, and 20q13.2, and others may exist as well [Cavaille et al., 2002; Gardner et al., 2000; Hall, 1990; Jiang et al., 2004; Weksberg et al., 2003; Wylie et al., 2000]. Imprinted regions share several common characteristics, including differential DNA methylation, allele-specific RNA transcription, antisense transcripts, histone modifications, and differences in timing of replication.

X-Inactivation

In mammalian cells with two (or more) X chromosomes, all but one undergo widespread gene silencing by methylation. This phenomenon, known as X-chromosome inactivation (Xi), causes one of the two X chromosomes in cells of female mammals to become transcriptionally inactive early in embryonic development, a phenomenon known as the Lyon hypothesis [Lyon, 1961, 2002]. In mutant cells with more than two X chromosomes, all but one become inactivated. This has the effect of balancing gene dosage of X-linked genes between male and female cells. The process of Xi is random, so that on average the maternally and paternally derived X chromosomes are each inactivated in approximately 50 percent of cells. Changes in this pattern are seen in female carriers of some X-linked diseases, resulting in skewing of Xi. This alteration can be favorable, with decreased severity of the phenotype, or unfavorable, with increased severity of the phenotype [Dobyns et al., 2004].

Mitosis

Somatic cell division, or mitosis, is an elaborate mechanism that distributes one chromatid of each duplicated chromosome to each of the two daughter cells. The process is continuous but has been divided into the following five stages: prophase, prometaphase, metaphase, anaphase, and telophase (Figure 30-5).

Fig. 30-5 Diagram of mitosis demonstrating two chromosome pairs.

In prophase, the chromatin begins to condense, the nucleolus disappears, and the mitotic spindle begins to form. Prophase is followed by prometaphase, during which the nuclear membrane disappears, allowing the chromosomes to disperse in the cell and attach to the spindle by paired kinetochores located at the centromere. In metaphase, the chromosomes are maximally contracted and arranged at the equatorial plane of the cell. In anaphase, the replicated chromosomes separate at the centromere, allowing the two chromatids to become daughter chromosomes, which move to opposite ends of the cell. In telophase, the chromosomes decondense, the nuclear membrane reforms, and the nucleus returns to the interphase appearance. Shortly afterward, the cytoplasm divides to form two daughter cells. For routine studies, chromosomes are examined during metaphase. For high-resolution studies, they are examined before the point of maximal contraction, during prophase or prometaphase.

Meiosis

Reproductive cell division, or meiosis, is an even more complex mechanism in which two successive cell divisions, known as meiosis 1 and meiosis 2, give rise to the haploid germ cells (Figure 30-6). Meiosis is of critical importance in understanding many of the methods of modern molecular genetics and the pathogenesis of many genetic diseases.

Fig. 30-6 Diagram of meiosis depicting two chromosome pairs.

In meiosis 1, the chromosome number is reduced from the diploid to the haploid number. The key step consists of close pairing of homologous chromosomes during prophase 1, which is further divided into several stages. During leptotene, the chromosomes first become visible, with homologs located close together. During zygotene, the homologs begin to pair closely along their entire length, held together by a thin protein-containing structure known as a synaptonemal complex. During pachytene, synapsis or pairing is completed, and the homologs appear as a bivalent. Pachytene is the stage during which exchange of homologous segments between nonsister chromatids occurs, which is known as recombination or crossing over. The remaining steps are similar to mitosis, except that it is the paired homologs that are pulled apart rather than the centromeres. In meiosis 2, which closely resembles mitosis, the chromatids separate at the centromere to form daughter chromosomes. Ova and sperm have remarkably different timing, but the sequence of meiosis is the same.

Chromosome Structure

In humans, the nuclear DNA is dispersed among 46 separate linear structures or chromosomes, each of which consists of a single, uninterrupted double helix that contains 50–250Mb of DNA, and a group of associated proteins that form the support structure or scaffolding. The scaffolding consists of five basic proteins called histones and several more acidic nonhistone proteins. Two copies of each of four histones – H2A, H2B, H3, and H4 – join to form an octamer. The DNA double helix wraps almost twice around the octamer, which involves about 140 bp. Adjacent octamers are separated by a short spacer segment of 20–60bp that is associated with histone H1. The complex of DNA and core histones is known as a nucleosome (see Figure 30-2).

Strings of nucleosomes are further compacted into a secondary helical structure known as a solenoid. These structures have a diameter of about 30 nm (see Figure 30-2) and contain six nucleosomes per turn. The solenoids are packed into large loops of 10–100 kb of DNA, which are attached to a nonhistone protein scaffolding. These loops pack together loosely to form interphase chromosomes. During early prophase, they pack together more closely to form knoblike thickenings known as chromomeres, which then coalesce further to form the bands observed in prometaphase and metaphase chromosomes when stained with appropriate dyes.

The alternating light and dark bands that characterize all nuclear chromosomes with a variety of staining methods likely reflect the compartmentalization of the genome into isochores, defined as large regions with variation in base composition or variable spacing of scaffold attachment regions. The dark bands observed with Giemsa staining are AT-rich, replicate late in the DNA synthesis phase of the cell cycle, and contain relatively few genes. The light bands observed with Giemsa are GC-rich, replicate early, and contain many genes. Some are greatly enriched for GC and contain high concentrations of genes. Most, although not all, such bands are located near the ends or telomeres of chromosomes and therefore are known as T bands.

Specialized Regions

All nuclear chromosomes have specialized regions that are required for chromosome integrity and function, including centromeres, telomeres, and origins of replication. Centromeres are DNA sequences that act in cis. That is, they act on the chromosome on which they are located and are responsible for the segregation of chromosomes during cell division. Centromeres contain extensive repeats of an approximately 171-bp unit known as alpha-satellite DNA, the sequence of which differs slightly between each chromosome. Fragments of chromosomes that lack a centromere, known as acentric fragments, are lost during cell division.

The two ends of a chromosome are called telomeres and also are required for chromosome stability. In humans, they consist of long arrays of tandem repeats of the sequence TTAGGG, which extend about 5–20 kb. DNA polymerases are unable to replicate the telomeres because of the lack of a template. This problem is resolved by the enzyme telomerase, which contains an RNA component to serve as a template to prime further synthesis on the leading strand. Further extension of the leading strand provides the needed template for the lagging strand.

Origins of replication are specialized sequences where DNA replication begins, and thus are important in maintaining chromosome number and integrity. They consist of autonomously replicating sequence elements that contain a core consensus sequence and some imperfect copies with a length of about 50 nucleotides. A consensus human autonomously replicating sequence has been identified [Strachan and Read, 2010].

Regions of variable staining known as heterochromatin consist of long arrays of repeat sequences as short as 5 bp. These regions are located primarily in the pericentromeric regions of chromosomes 1, 9, and 16, and in distal Yq. The five human acrocentric chromosomes have small satellites attached to the short arm by short stalks or secondary constrictions that contain the rRNA genes.

Chromosome Number

Each human somatic cell contains 46 chromosomes that consist of 22 matched pairs known as autosomes and two sex chromosomes: XX in females and XY in males (Figure 30-7). In contrast, human germ cells contain only 23 chromosomes, consisting of 22 unpaired autosomes and a single sex chromosome. The former is known as the diploid or 2n number, and the latter is known as the haploid or 1n number. The autosomes were numbered according to length, with chromosome 1 the longest and chromosome 22 thought to be the shortest. Although chromosome 21 later proved to be shorter than chromosome 22, the numbers were retained for historical reasons. The two members of each pair of autosomes and the two X chromosomes in females carry the same genes and are known as homologous chromosomes, or homologs. Although they appear similar under the microscope, homologs are not strictly identical. They contain the same genes, but the nucleotide sequence differs at thousands of positions.

Fig. 30-7 Standardized diagram or idiogram of human chromosomes at the 400-band stage.

Chromosome Identification

Individual chromosomes may be seen only when tightly contracted during cell division. Since DNA replication is complete, each chromosome consists of two chromatids that are joined at the primary constriction or centromere. In standard cytogenetic nomenclature, the centromere divides the chromosome into two arms, with the shorter designated the “p” arm and the longer the “q” arm. The tip of each arm is the telomere. Human chromosomes are classified into three types according to the position of the centromere:

Genes and Conserved Noncoding DNA

Somewhat surprisingly, recent estimates predict that the human genome contains less than 30,000 protein-coding genes (possibly closer to 20,000) and an uncertain number of other genes producing functional RNA products. This is far fewer than earlier estimates, and accounts for only about 1.2 percent of nuclear DNA [Lander et al., 2001; Venter et al., 2001]. Another 5 percent of the human genome is more conserved than would be expected from estimates of neutral evolution, which suggests that many of these regions have specific, regulatory functions [Chiaromonte et al., 2003; Waterston et al., 2002]. Studies of these highly conserved regions of DNA have used different thresholds, such as stretches of more than 100 bp with 70–80 percent conservation between mouse and human. Some of these regions have been found to contain important noncoding elements [Dermitzakis et al., 2002, 2003; Frazer et al., 2004; Hardison, 2000]. More stringent analysis demonstrates that the human genome contains 481 sequences of 200 or more bp that are 100 percent conserved among human, mouse, and rat [Bejerano et al., 2004]. These segments were designated “ultra-conserved elements,” and are preferentially located near genes involved in RNA processing or regulation of transcription and development. Similarly, about 5000 sequences of 100 bp or more are conserved among these three species, which emphasizes that noncoding sequences are common and important.

Repetitive DNA

Repetitive DNA in the human genome consists of several classes of DNA whose nucleotide sequence is repeated, either exactly or with minor variations, hundreds to millions of times. Some classes are clustered, whereas others are dispersed throughout the genome. Clustered, repeated sequences constitute 10–15 percent of the genome and are collectively called satellite DNA because of their separation from other DNA on density centrifugation. Satellite DNA consists of head-to-tail or tandem arrayed repeat sequences that can extend for several thousand kb. Dispersed, repeat sequences constitute 6–10 percent of the genome and belong to several different classes. Minisatellite or variable number of tandem repeat (VNTR) sequences are dispersed, intermediate-length (15–65 bp) repeats that usually span only several kb. The Alu family of DNA repeats includes about 500,000 related sequences that are each about 300 bp in length and together make up about 3 percent of the genome. The L1 family of repeats includes about 10,000 related sequences that extend up to 6 kb in length and make up another 3 percent of the genome. Although the origin of these sequences is not known, no functions have been identified, and it appears likely that they simply exploit cellular processes to propagate themselves. Several classes have been useful as polymorphic DNA markers.

Low Copy Repeats

Segmental duplications, also known as low copy repeats (LCRs), are DNA sequences of 10–250 kb, present in multiple copies with greater than 95 percent sequence identity, that make up approximately 5 percent of the human genome [Babcock et al., 2003; Bailey et al., 2002; Cheung et al., 2001; Stankiewicz and Lupski, 2002]. LCRs are dynamic regions of the genome because specific repeats tend to cluster within the same genomic regions, where they mediate unequal nonhomologous recombination events, producing segmental deletions and duplications that are collectively designated “copy number variants” (CNVs). Several of these have been associated with well-known developmental disorders in humans, such as Williams’ syndrome in 7q11.23, Angelman’s syndrome and Prader–Willi syndrome in 15q12, hereditary neuropathy with predisposition to pressure palsies and Charcot–Marie–Tooth neuropathy type 1A in 17p12, Smith–Magenis syndrome in 17p11.2, and DiGeorge’s syndrome in 22q11.2 [Babcock et al., 2003]. Many new CNV-associated devlopmental brain disorders have been described over the past few years.

Minisatellites

One of the most useful classes of polymorphisms in the genome is that of the minisatellite or VNTR DNA sequences. These are intermediate-length (15–65 bp) DNA sequences that are repeated one to several dozen times in tandem and usually span several kb in total length. They are highly polymorphic, and their extreme polymorphic nature, coupled with the complexity of multilocus minisatellites, makes them valuable for DNA fingerprinting applications, such as forensic, paternity, and zygosity testing and linkage mapping. They also are inherently unstable and susceptible to mutation at a higher rate than observed for other sequences of DNA.

Microsatellites

Microsatellites, also known as satellite DNA or short tandem repeats, are segments of DNA 2–5 nucleotides in length (dinucleotide, trinucleotide, tetranucleotide, or pentanucleotide repeats) that are scattered throughout the genome in noncoding regions between genes or within genes (in introns). They often are used as markers for linkage analysis because of the naturally occurring high variability in repeat number between individuals. These regions are inherently unstable and susceptible to mutations.

The most common microsatellite family consists of 50,000–100,000 cytosine-adenine (CA) repeats, which consist of short tandem repeats of the dinucleotide CA on one strand and guanine-thymine on the complementary strand. They thus take the form (CA)n/(GT)n, with n in the range of 6–30 [Weber and May, 1989]. The number of repeats within a (CA)n block varies greatly among different members of a species, producing a set of alleles that always differ in size by multiples of two bases. About 70 percent of the human population is heterozygous at any given (CA)n repeat locus, making these highly polymorphic. The human genome contains about 50,000–100,000 interspersed (CA)n blocks, which is enough to place 1 block every 30–60 kb, if evenly spaced.

For both VNTR and CA repeat sequences, the combination of high frequency in the genome and a high rate of polymorphism has made them very useful for genetic mapping and association studies. Some microsatellite repeats, most often trinucleotide repeats, present within coding regions of genes or, less often, the 5′ or 3′ UTR, can expand to an abnormal length and are the basis of triplet repeat diseases such as Huntington’s disease, some forms of spinocerebellar ataxia, and fragile X syndrome.

Single-Nucleotide Polymorphisms

Single-nucleotide polymorphisms (SNPs, pronounced “snips”) are DNA sequence variations that occur when a single nucleotide (A, T, C, or G) in the genome sequence is changed. For example, a SNP might change the DNA sequence TCACG to TTACG. The most common sequence change involves replacement of cytosine (C) with thymidine (T), which accounts for about two-thirds of all SNPs. As with other types of sequence variation, a SNP must occur in at least 1 percent of the population to be classified as a polymorphism. SNPs occur in both unique-sequence (coding and noncoding) and repetitive DNA, and are responsible for about 90 percent of human genetic variation. On average, SNPs are found approximately every 100–300 bases along the entire human genome. Although most SNPs likely have no function, some are known to influence disease predisposition or responses to drugs, and thus are proving to be very valuable in studying the causes of common human diseases. The current inventory of known SNPs can be found in the Human SNP database (dbSNP) on the NCBI Entrez website (see Table 30-1).

Restriction enzymes are DNA-cutting enzymes or endonucleases derived from bacteria that cut DNA at specific short sequences found at locations across the entire genome. SNPs can alter the sequences recognized by restriction enzymes, thus adding or removing a cutting site. This is the biological basis for restriction enzyme fragment length polymorphisms (RFLPs). Depending on the location of restriction enzyme sites, specific DNA fragment lengths are obtained on digestion with restriction endonucleases. The presence of a SNP at one of these restriction enzyme sites will affect cleavage and produce two DNA fragments of different sizes that is the RFLP. RFLPs also can be produced by any change that alters the size of the DNA fragment on which the restriction site is located, such as deletions or duplications. RFLPs are a measure of naturally occurring variations or polymorphisms of normal DNA, and are inherited according to mendelian principles. RFLPs have been useful for gene mapping.

DNA Sequence Analysis

Mutation Scanning

Mutation scanning refers to methods used to determine the presence of a sequence change in a region of DNA (such as an exon of a gene). These methods need to be followed up by DNA sequencing, however, to determine the exact nature of the sequence change. Mutation scanning generally is less labor-intensive, faster, and more cost-effective to perform than DNA sequencing and may be the method of choice when a large gene needs to be analyzed for the presence of mutations. The exons of the entire gene are subjected to mutation scanning, and only those exons that demonstrate the presence of a sequence change need to be sequenced to determine the precise nature of the sequence change.

Various methods of mutation scanning exist that differ in their sensitivities of mutation detection [Cotton, 1997; Eng and Vijg, 1997; Grompe, 1993]. The general basis of most mutation scanning methods is the abnormal migration of a DNA fragment that contains a sequence change from a normal “wild-type” sequence. All are PCR-based methods – the DNA region to be studied is amplified by PCR before the different mutation scanning methods are performed.

Allele-Specific Oligonucleotide Hybridization

For allele-specific oligonucleotide (ASO) hybridization, separate ASO probes complementary to either the normal or a specific mutant allele are synthesized. The probes consist of the point mutation or its normal counterpart and 9–12 bp flanking it on either side, for a total length of 19–25 bp. The probes then are hybridized to the DNA source under stringent conditions. Because the probes are short, they are highly sensitive to sequence changes at even a single nucleotide. A probe complementary to the normal allele will hybridize to the normal allele but not to the mutant allele, and vice versa. Thus, DNA from a person homozygous for the normal gene will hybridize to the ASO probe complementary to the normal DNA sequence, but not to the probe complementary to the point mutation. DNA from persons homozygous for the mutation will hybridize to the ASO probe complementary to the mutation, but not to the normal probe. Only DNA from heterozygotes will hybridize to both probes. It is important to remember that DNA from persons who are heterozygous for different mutations in the same gene – compound heterozygotes – also will hybridize to the normal probe. Several different formats have been developed to detect ASO probes, ranging from radioactivity to chemiluminescence to fluorescence, and can be performed in both solid and liquid phases.

Single-Base Extension

Oligonucleotide primers of approximately 20 bp are designed to hybridize just upstream or downstream of the nucleotide to be genotyped, and an extension reaction – more specifically, a single-base extension (SBE) reaction – is performed in the presence of ddNTPs. In the case of a normal sequence, a ddNTP corresponding to the complementary normal nucleotide is added to the primer; in the case of a mutation, a ddNTP corresponding to the complementary mutant nucleotide is added to the primer. The ddNTPs generally are labeled with different fluorochromes, allowing their detection. In a person homozygous for a normal gene, extension will occur only with the ddNTP corresponding to the normal allele, whereas in a person homozygous for a mutated gene, extension will occur only with the ddNTP corresponding to the mutated allele. A heterozygous person will have both normal and mutated extension products. Several different formats have been developed to detect SBE products and can be performed in both solid and liquid phases.

DNA Arrays

In DNA arrays, hundreds to thousands of DNA targets are arranged (arrayed) on a solid medium such as a glass slide or microchip. DNA arrays are of two main types – genotyping arrays and sequencing arrays. Genotyping arrays consist of hundreds to thousands of different ASO probes or SBE primers arrayed on a chip that allows the genotyping of a large number of different loci simultaneously. DNA sequencing arrays consist of a series of hundreds of thousands of approximately 20-bp oligonucleotides, spanning the length of a gene, that are arranged with equal spacing, or “tiled,” across the microchip. For each nucleotide to be sequenced, four oligonucleotides are present that differ only at the central position and have an A, C, G, or T. Digested DNA is fluorescently labeled and hybridized to the tiled oligonucleotide chip under stringent conditions. The DNA hybridizes to those oligonucleotides that correspond to the correct sequence and the fluorescent signal read from which the DNA sequence is deciphered. DNA sequencing and genotyping arrays are available for several genes and are still largely used for research, although some clinical applications have been developed. Expression arrays are different and refer to cDNA arrays to which RNA is hybridized and are used to determine the expression profile of hundreds of genes simultaneously. This technique is currently used only in the research setting.

Restriction Enzyme Analysis

Some mutations result in the destruction or creation of a restriction enzyme site in the DNA sequence. PCR amplification of the region of interest, followed by digestion of the genomic DNA with the appropriate restriction enzyme and gel electrophoresis, can be performed to determine the presence or absence of a mutation that affects a restriction enzyme site. Destruction or creation of a restriction enzyme site will result in a different-sized DNA fragment compared with a normal control fragment. The sickle cell anemia mutation and the hereditary hemochromatosis mutation both affect restriction enzyme sites.

Triploidy and Tetraploidy

Occasionally, fetuses with three or four times the normal haploid number of chromosomes have been observed. These abnormal chromosome complements are called triploidy (3n) and tetraploidy (4n). The few children who are liveborn survive only briefly after birth, unless the abnormality is mosaic (i.e., involves only a proportion of their cells). Failure of a maturational division in either egg or sperm results in triploidy, whereas failure of completion of an early division of the zygote causes tetraploidy.

Aneuploidy

Aneuploidy is defined as any chromosome complement that deviates from a multiple of the haploid number. In most cases, it consists of either monosomy, which is defined as loss of an entire chromosome, or trisomy, which refers to gain of an entire chromosome. Aneuploidy is the most common and clinically significant type of chromosome disorder, occurring in 3–4 percent of all recognized pregnancies.

Both monosomy and trisomy of autosomes are lethal during early pregnancy in a large majority of affected fetuses. Autosomal monosomy is uniformly lethal, except for a few reports of liveborn children with monosomy 21. Monosomy X is prenatally lethal in most affected fetuses, but many survive and will have the phenotype of Turner’s syndrome. The effects of trisomy vary, depending on the chromosome involved. Trisomy 16 is the most frequent autosomal trisomy at conception but is uniformly lethal before birth. The most common type of trisomy in liveborn infants is trisomy 21, which is the chromosome abnormality observed in 95 percent of children with Down syndrome. The only other autosomal trisomies observed at appreciable frequencies are trisomy 13 and trisomy 18, although trisomy 8 may be observed in mosaic form.

The most common mechanism is nondisjunction, which is the failure of a pair of chromosomes to separate correctly during one of the two stages of meiosis, usually meiosis 1. The consequences of nondisjunction during meiosis 1 and 2 are somewhat different. If the error occurs during meiosis 1, the unbalanced gamete with 24 chromosomes contains both the maternal and the paternal members of the pair. If the error occurs during meiosis 2, the unbalanced gamete will contain either the maternally or the paternally derived chromosome, but not both.

Mechanisms

Some progress has been made recently in current understanding of the mechanisms causing or at least predisposing to structural chromosome rearrangements. Many appear to occur randomly, with no evidence of recurrent breakpoints. For example, no consistent breakpoints have been found for the interstitial deletions and reciprocal translocations involving chromosome band 17p13.3 [Cardoso et al., 2003]. In many other locations, the small duplicated regions known as LCRs can mediate several different types of rearrangements. These were first identified as the cause of common deletion or microdeletion syndromes, such as deletion (del) 2q13 with juvenile nephronophthisis and Joubert’s syndrome-related disorder [Parisi et al., 2004; Saunier et al., 2000], 7q11.23 in Williams’ syndrome [Osborne et al., 2001; Urban et al., 1996], del 15q11.2–q13 in Angelman’s syndrome and Prader–Willi syndrome [Amos-Landgraf et al., 1999], del17p12 in hereditary neuropathy with predisposition to pressure palsies [Chance et al., 1994], del17p11.2 in Smith–Magenis syndrome [Chen et al., 1997], and del22q11.2 in DiGeorge’s syndrome/velocardiofacial syndrome [McDermid and Morrow, 2002]. The same LCRs are associated with duplications of the same regions, including duplication 15q11.2–q13 [Mohandas et al., 1999], 17p12 [Pentao et al., 1992], and 17p11.2 [Potocki et al., 2000]. A simple diagram of this mechanism is shown in Figure 30-11.

Fig. 30-11 Drawing of a recombination event mediated by low copy repeats.

The black boxes represent low copy repeats and the long four-armed linear structures represent chromosomes with two chromatids each during meiosis 1. In the drawing on the left, the outermost chromatids have paired correctly, while the inner two chromatids have paired incorrectly (long arrow pointing to site of crossover). In the drawing on the left, the outermost chromatids appear normal. But in the innermost pair, the crossover event within the mispaired region results in one daughter chromosome with a duplication and another with a deletion. The dark squares represent the low copy repeats.

Many of the LCRs are inverted with respect to each other, which can lead to very complex combinations of deletions and duplications. This can occur for LCRs on the same chromosome, as seen with Williams’ syndrome and some X chromosome rearrangements [Giglio et al., 2000; Osborne et al., 2001], or between homologous chromosomes. The latter appears to be more common when two matching LCRs on homologous chromosomes are inverted with respect to each other, a novel type of polymorphism [Giglio et al., 2001]. Similar mechanisms also can predispose to structural rearrangements between completely different chromosomes involving either an LCR or a gene cluster, such as olfactory receptor clusters [Giglio et al., 2002; Spiteri et al., 2003].

Balanced and Unbalanced Chromosomal Rearrangements

As noted previously, structural chromosome rearrangements that result in no net loss or gain of genomic sequence are balanced, whereas those that do result in a net loss or gain of material are unbalanced. Persons with balanced rearrangements usually are normal, unless one of the chromosomal breaks disrupts an important gene. More recent studies, however, have found that chromosome rearrangements that appear balanced often have submicroscopic loss or gain of material and so are actually unbalanced [Astbury et al., 2004b], and some chromosome rearrangements are more complex than standard chromosome analysis suggests [Astbury et al., 2004a].

Deletions and Duplications

A deletion consists of loss or gain of a chromosome segment. Deletions may be either interstitial or terminal, with terminal deletions including the telomere (see Figure 30-8 and Figure 30-12D), whereas most duplications are interstitial. Most interstitial deletions and duplications result from unequal crossing over in LCRs (see Figure 30-11). Terminal deletions are more likely to result from simple chromosome breakage, although some apparently terminal deletions prove to be interstitial deletions in which one breakpoint happens to be close to the telomere. Any carrier of a deletion is hemizygous for the information on the corresponding segment of the normal homolog. Thus, small but cytogenetically visible deletions involving critical genes occasionally produce single-gene phenotypes, such as lissencephaly or retinoblastoma.

Duplicated segments usually are adjacent to each other and may be in the same orientation (direct dup), or inverted (inverted dup) with respect to one another. In general, the phenotypic effects of duplications are less severe than the effects of deletion of a similar segment. Small duplications also can result in single-gene phenotypes, such as Charcot–Marie–Tooth neuropathy [Chance et al., 1994], although this is recognized less often than with deletions.

Inversions

Inversions are segments within a chromosome that are inverted with respect to the normal orientation; they result from crossovers within existing duplicated segments (LCRs) or from two breaks within a single chromosome, followed by inversion of the intervening segment and repair of the breaks. When the inverted segment includes the centrosome, the rearrangement is described as a pericentric inversion (see Figure 30-12A), whereas rearrangements in which the inverted segment does not include the centrosome are designated as paracentric inversions. Both types of inversions may result in production of unbalanced gametes because of the effects of recombination within the inverted segment. With pericentric inversions, a loop is formed between the inverted chromosome and its homolog during meiosis 1 (Figure 30-13). Recombination is somewhat, but not completely, suppressed within inversion loops, so crossovers are common in larger loops.

Fig. 30-13 The effect of recombination within the loop of a pericentric inversion.

A, A normal chromosome is depicted on the left, with loci 1 to 7 in order and the centromere located between loci 3 and 4. A pericentric inversion is depicted on the right, with the segment containing loci 3–5 and the centromere inverted. B, Pairing of the normal and inverted chromosomes during meiosis 1, with a crossover occurring within the inversion loop in the middle two chromatids. C, The four types of gametes produced after completion of meiosis include a normal chromosome, a derivative chromosome with duplication of the distal short arm and deletion of the distal long arm (dup p), the reverse derivative chromosome with duplication of the distal long arm and deletion of the distal short arm (dup q), and a balanced pericentric inversion.

Recombination within a pericentric inversion loop produces derivative chromosomes in which segments distal to the breaks are either duplicated or deleted (see Figure 30-13C). The effects on the phenotype are inversely proportional to the size of the inversion. Thus, the distal segments typically are large with small pericentric inversions, and most unbalanced offspring are spontaneously aborted. Liveborn children with birth defects are more likely with larger inversions that lead to relatively small distal segments.

Crossovers within a paracentric inversion loop, which commonly result from crossovers within LCRs, result in acrocentric or dicentric chromosomes. Acrocentric chromosomes are quickly lost during subsequent rounds of cell division. Dicentric chromosomes inactivate one of the two centromeres and are retained. For both, the loss or gain of chromosomal material and genes is so great that almost all affected embryos are spontaneously aborted, unless a large part of the derivative chromosome happens to break off and become lost. This mechanism has been proved and may be more common than appreciated [Giglio et al., 2002].

Reciprocal Translocations

Reciprocal translocations consist of breaks in nonhomologous chromosomes, with a reciprocal exchange of the broken segments (see Figure 30-12B). Usually, only two chromosomes are involved, but complex translocations involving three or more chromosomes have been described and likely are more common than standard chromosome analysis has suggested [Astbury et al., 2004a]. Population studies have detected reciprocal or robertsonian translocations in about 1 in 500 newborns.

Reciprocal translocations often result in the production of unbalanced gametes. During meiosis 1, the derivative chromosomes and their normal homologs form a quadriradial shape that may separate into pairs in one of three ways: alternate, adjacent 1, and adjacent 2 segregation. Alternate segregation produces balanced gametes that have either normal chromosomes or both derivatives, which are therefore balanced. Adjacent 1 segregation produces unbalanced gametes in which homologous centromeres separate into different daughter cells. It results in duplication of the distal segment of one derivative chromosome and deletion of the distal tip of the other. In most translocation carriers, alternate and adjacent 1 segregation account for a large majority of the gametes (Figure 30-14). Adjacent 2 segregation also produces unbalanced gametes. In this uncommon mechanism, homologous centromeres pass to the same daughter cell. The resulting nondisjunction produces 3:1 and even a 4:0 segregation.

Fig. 30-14 Diagram of the alternative types of segregation and gametes produced in the carrier of a reciprocal translocation between the short arms of chromosomes 3 (light purple) and 6 (dark purple).

The top line represents the parental chromosome pairs, the middle line represents the four types of gametes produced by the father, and the bottom line represents four possible chromosome combinations that may be observed in offspring. Alternate segregation (depicted on the right) produces offspring with either normal chromosomes or the balanced translocation. Adjacent 1 segregation (depicted on the left) produces offspring with unbalanced karyotypes. Children with the derivative 6 karyotype (der[6]) have deletion of the distal segment of 6p and duplication of the distal segment of 3p. Children with the derivative 3 karyotype (der[3]) have deletion of 3p and duplication of 6p. In both alternate and adjacent 1 segregation, homologous centromeres pass to different daughter cells. In adjacent 2 segregation (not depicted), homologous centromeres pass to the same daughter cells, leading to even greater chromosomal imbalance.

Reciprocal translocations often are detected in normal adults evaluated because of repeat fetal loss, or after the birth of a child with multiple congenital anomalies caused by transmission of the translocation in unbalanced form. Apparently balanced reciprocal translocations sometimes are found in children with birth defects or abnormal development. In these instances, the abnormal phenotype usually results from either submicroscopic loss of genetic material or disruption of a gene at one of the breakpoints [Astbury et al., 2004b].

Robertsonian Translocations

Robertsonian translocations involve two acrocentric chromosomes that fuse in or near the centromere region, with loss of the short arms (see Figure 30-12C). Because the short arms contain repetitive DNA elements, especially rRNA, no phenotypic effects result. Carriers of a robertsonian translocation on chromosome 21 have a high risk of producing a child with translocation Down syndrome.

Insertions

Insertions occur when a small segment of a chromosome is removed and inserted into a different region on the same or another chromosome. If the segment is inserted with the same orientation with respect to the centromere, it is known as a direct insertion. If it is inserted with the reverse orientation, it is called an inverted insertion. Insertions are rare because three separate chromosomal breaks are required. Segregation during meiosis can produce either abnormal offspring, with duplication or deletion of the inserted segment, or normal offspring and balanced carriers.

Rings

Rings – ring chromosomes – are formed when a chromosome undergoes two breaks, usually one in each arm, and the broken ends are rejoined (see Figure 30-12E). The two segments distal to the breaks are lost, resulting in deletion of both telomeres and adjacent regions of both the short and the long arms of the chromosome. Rings may not segregate properly during mitosis and meiosis, especially if a crossover occurs. Crossover often results in breakage followed by fusion, which may produce larger or smaller rings.

Isochromosomes and Dicentrics

Isochromosomes are chromosomes in which one arm is missing and the other is duplicated as the result of misdivision of the centromere during meiosis 2. Isochromosomes also can result from translocation of an entire arm to its homolog with a breakpoint adjacent to the centromere. The most commonly observed isochromosome involves Xq. Dicentrics are rare chromosomes in which two segments, each containing a centromere, fuse end to end. They tend to break during mitosis because of the double centromeres.

Genome and Chromosome Mutations

Abnormalities of chromosome number, including triploidy, tetraploidy, and aneuploidy, are classified as genome mutations. Similarly, structural rearrangements are classified as chromosome mutations. Both of these were reviewed in the preceding section. Genome and chromosome mutations are rarely perpetuated to the next generation because most result in spontaneous abortions. Thus, the frequencies cited are probably underestimates.

Gene Mutations

Gene mutations differ from genome and chromosome mutations because the segment of DNA involved is much smaller and the mechanisms are different. The most common types are basepair substitutions and small deletions or insertions that can be caused by an error during DNA replication or by base changes induced by extrinsic agents referred to as mutagens. Because genome and chromosome mutations usually are lethal, most significant heritable mutations are gene mutations.

DNA Replication Errors

DNA replication normally is an accurate process. The DNA polymerases (see Table 30-2) insert an incorrect base only once in every 10 million bp. A series of DNA repair enzymes exist that are able to recognize and replace noncomplementary bases, correcting more than 99.9 percent of errors. The overall mutation rate is therefore only 10⁻¹⁰ per basepair per cell division. The human genome consists of about 6 × 10⁹ bp, so this mutation rate results in less than 1 bp mutation per cell division. Nevertheless, an estimated 10¹⁵ cell divisions occur during the lifetime of an adult human. Thus, thousands of new mutations occur at virtually every position in the genome. Not surprisingly, inherited defects in DNA replication and repair enzymes lead to a striking increase in the frequencies of all types of mutations.

Most of the mutations occur in somatic cells, where they may cause cancer or a genetic disease affecting only part of the body, such as segmental neurofibromatosis. Fewer mutations occur in germ cells. During oogenesis, female germ cells undergo mitosis approximately 22 times and begin meiosis only once during fetal life. The cells are suspended in meiosis from fetal life till shortly before ovulation during the reproductive years. Spermatogenesis consists of approximately 30 mitoses from conception until puberty and approximately 20–5 per year thereafter. Thus, the opportunity for mutations is expected to be far greater for sperm than for ova. This phenomenon has been confirmed in several genetic disorders, such as neurofibromatosis type 1, achondroplasia, and hemophilia A. It has been estimated that as many as 1 in 10 sperm in healthy males may carry a new deleterious mutation. Most are recessive or lethal, and therefore not apparent in liveborn children.

Mutation Rate

The mutation rate for any given gene or other DNA segment depends on both its size and its location. The location may be important because certain areas of the genome are known to be “hot spots” for recombination. The average mutation rate is about 1 × 10⁻⁶ mutation per locus per generation, but the rate varies by more than a thousand-fold for different genes. For example, the number of new mutations per 10⁶ gametes is 40–100 for Duchenne muscular dystrophy and neurofibromatosis 1, but only 2–5 for aniridia and hemophilia B. These statistics include only mutations causing genetic diseases. The rate of change of protein polymorphisms suggests a rate as high as 6 × 10⁻⁶ per locus per generation. (A locus is the position of a gene on a chromosome.)

Nucleotide Substitutions

Point mutations or single-base substitutions represent one of the most common types of mutation. Most are related to an error in DNA synthesis by the enzyme DNA polymerase that was not corrected by DNA repair enzymes. Some combinations of nucleotides are mutation-prone, however. More than 30 percent of point mutations found in some genetic diseases are the result of cytosine to thymine transitions, which are caused by methylation of cytosine residues to 5-methylcytosine, especially cytosine residues occurring as the first base in a 5′-CG-3′ dinucleotide pair. The latter then undergoes spontaneous deamination to thymidine. Thus, the 5′-CG-3′ doublet represents a “hot spot” for mutation in the human genome.

Deletions, Duplications, and Insertions

The remainder of mutations consist of loss or gain of nucleotide bases somewhere in the genome; a variety of mechanisms for such changes are known. A deletion consists of any loss of DNA sequence, whereas a duplication consists of a second copy of a DNA sequence that is usually located immediately adjacent to the first copy. An insertion consists of a DNA sequence that has been removed or copied from one location and moved to a nonhomologous region elsewhere on the same chromosome or to a different chromosome. Deletions, duplications, and insertions that involve one or a few basepairs can be detected only by nucleotide sequencing. Larger deletions and duplications may be detected by several of the methods described earlier, including direct sequencing, Southern blot analysis, FISH, and CGH.

Missense Mutations

A point mutation within the coding region of a gene can alter the genetic code by changing the nucleotide triplet and cause the replacement of one amino acid by another in the gene product, thus altering function of the gene product. Such mutations are called missense mutations and do not change the reading frame of the DNA sequence. The best-known example is the A to T substitution in the sixth codon of the β-globin gene, which causes sickle cell anemia, by substituting valine for glutamic acid in the β-globin protein chain. Not all mutations within coding regions of a gene result in a missense mutation, however. All but 2 of the 20 amino acids are specified by more than one codon, most often differing in the third or “wobble” position of the triplet. The gene product will be identical if the new triplet codes for the same amino acid.

Nonsense (Chain Termination) Mutations

Mutations that generate one of the three stop codons result in premature termination of translation, whereas those that alter a stop codon allow translation to continue until the next stop codon is reached. Those mutations that result in a premature stop codon are called nonsense mutations. In general, these mutations have no effect on transcription (DNA to RNA), but the shortened polypeptide may have lost critical functional domains of the protein, or the mRNA may be so unstable that it is rapidly degraded in the cell. The latter process is known as nonsense-mediated mRNA decay, a process by which mRNA species containing premature termination codons are recognized and degraded before translation, although this typically spares truncation mutations in the last coding exon [Frischmeyer and Dietz, 1999]. Both base substitutions and nucleotide loss or gain mutations may result in nonsense mutations.

RNA Splicing Mutations

The sequence surrounding intron splice sites is highly conserved, and mutations of key nucleotides frequently prevent or reduce efficiency of splicing. The key nucleotide sequences at most splice junctions are shown in Figure 30-4.

Splicing mutations may either inactivate existing splice sites or create new ones. In the first type, the mutation alters the splice-donor, branch, or splice-acceptor site, resulting in failure to splice the intron correctly at that site. This failure results in a large insertion of nucleotides that normally are not translated into the processed mRNA. This insertion is almost certain to introduce a stop codon within the next hundred or so codons, because 3 of the possible 64 triplet combinations are stop codons. In the second type, mutations within the intron create alternative splice-donor or acceptor sites that compete with the normal splice sites during mRNA processing. Thus, a proportion of the mature mRNA will contain incorrectly spliced intron sequences. Both base substitutions and nucleotide loss or gain mutations may result in splicing mutations.

One example of the first type is a G to C transition in the first position of the intron at the donor splice site in the hexosaminidase A gene found in many Ashkenazi Jewish patients with Tay–Sachs disease [Nussbaum et al., 2007]. In this example, the bases in the exon are capitalized, whereas those in the intron are not, and the mutation is underlined.

Frameshift Mutations

Small nucleotide loss or gain mutations may alter the reading frame of the mRNA product from the point of the mutation on, which results in a completely different amino acid sequence at the carboxyl end of the protein product, or premature chain termination if a stop codon is encountered in the new reading frame. Any loss or gain mutation that involves a multiple of three bases maintains the reading frame, whereas a mutation that does not involve a multiple of three nucleotides changes the reading frame. Larger deletions that include one or more introns also may cause a frameshift mutation, because exon/intron splice sites may occur at any point in the reading frame, thereby splitting codons. If the exon just downstream from the deletion normally begins at a different position in the triplet than the deleted intron, the reading frame will be changed. By contrast, base substitutions do not cause frameshift mutations. Deletions and insertions cause dysfunction of the gene more often than point mutations because of the possibility of a frameshift.

One of the best-known frameshift mutations is a single-base deletion in the ABO blood group locus that results in the nonfunctional O allele. The deletion alters the reading frame at codon 86 until a premature stop codon is reached 30 codons later. The stop codon is normally out of frame and is therefore not read.

With some intermediate-size deletions of approximately 1 kb to 1 Mb, one or more exons of a gene may be duplicated or deleted. About two-thirds of these will change the reading frame and result in a frameshift mutation. Those that maintain the reading frame produce truncated products that may or may not retain function. These relatively small deletions and duplications are too small to be seen with chromosome analysis or FISH and cannot be found by sequencing, which is not sensitive to dosage (recall that all autosomal genes and X-linked genes in females have two copies of each gene). These may be rare or common mechanisms of mutation. For example, small deletions and duplications are the most common mutational types for Duchenne muscular dystrophy and Becker muscular dystrophy, both caused by mutations of the dystrophin or DMD gene. These are simple to detect for X-linked diseases in males, because only one gene copy is present. When any autosome of the X chromosome in a female is involved, the mutation can be detected by other methods, such as quantitative PCR assay.

Transcriptional Control Mutations

Mutations involving promoter sequences in the 5′ UTR or other regulatory sequences in the 3′ UTR of a gene may result in a significant decrease in the amount of mature, processed mRNA produced. Both base substitutions and nucleotide loss or gain mutations may result in transcriptional control abnormalities.

Autosomal-Dominant Inheritance

The most important attributes of autosomal-dominant inheritance are expression of the trait in heterozygotes and male-to-male transmission (see Figure 30-16). The autosomal-dominant pattern may be recognized because:

Autosomal-dominant inheritance is readily identified in most families but may be difficult to discern in others. When the disease occurs as a result of a new mutation, no relatives are affected. With reduced penetrance, low expressivity, and late age at onset, other affected family members may go unrecognized. Among the best examples for each of these characteristics are myotonic dystrophy and Huntington’s disease. Finally, incorrect information regarding family relationships, such as false paternity, may complicate interpretation of the pedigree.

Most persons affected by a disorder of autosomal-dominant inheritance are heterozygous, but rarely a homozygous person is encountered. Generally, the phenotype in homozygous persons is significantly more severe than in heterozygous persons. For example, one child born to parents who each had hereditary motor and sensory neuropathy type I had a much more severe neuropathy consistent with Dejerine–Sottas disease, or hereditary motor and sensory neuropathy type III [Killian and Kloepfer, 1979]. The best-known exception to this rule is Huntington’s disease, in which persons heterozygous for this trait cannot be distinguished from homozygotes.

Autosomal-Recessive Inheritance

The most important attributes of autosomal-recessive inheritance are expression in homozygotes and equal gender distribution (see Figure 30-16). This pattern may be recognized by the following four characteristics:

In Western societies, a child with a disorder of autosomal-recessive inheritance may be the only affected person in the family, owing to small family size and a tendency for parents of affected children to have fewer children after the birth of a child with a genetic disease. This practice does not hold true in many other cultures, especially those with inbred populations.

When the frequency of a rare recessive allele is relatively high within a family or population, the disease may appear in more than one generation. This pattern is known as pseudodominant inheritance. Some genes on the X chromosome have functional homologs on the Y chromosome, and traits or diseases associated with these genes will behave in the same manner as for autosomal loci. This pattern is known as pseudoautosomal inheritance.

The risk of bearing a child with an autosomal-recessive disease or trait is increased when the parents are consanguineous or related by descent. More formally, the probability that a homozygote has received both alleles of a pair from an identical ancestral source is known as the coefficient of inbreeding. It also is the proportion of loci at which a person is homozygous by descent. For example, any child born to first cousins is homozygous at 1/16 of all loci. Although the relative risk of abnormal offspring is higher for first cousins than for unrelated parents, it is still low, at about 5 percent.

X-Linked Inheritance

The inheritance of diseases and traits associated with genes located on the X chromosome differs markedly from autosomal forms of inheritance because females have two X chromosomes, whereas males have only one. Thus, when mutations of an X-linked gene occur in males, no genetic “backup” is available. The situation in heterozygous (carrier) females is more complicated, owing to the phenomenon of X chromosome inactivation (Xi), which ensures that dosage for X-linked genes is the same in male and in female cells. Because Xi is random in most females, the maternally derived and paternally derived X chromosomes usually are active in about half of the cells in a female organism. Thus, mutation of one gene should cause no more than a 50 percent loss of function of the protein or other gene product. With this background, several mechanisms have been described that lead to disease expression in female carriers of X-linked mutations. First, some genes are dosage-sensitive, so that 50 percent expression is not enough for normal function. Next, by chance or because of cell selection (usually favoring the normal allele), some females have skewing of X-inactivation, such that one X chromosome is inactivated in a high proportion (80–100 percent) of cells. Unfavorable skewing will cause or worsen disease, whereas favorable skewing will prevent or reduce disease expression. Finally, skewing of X-inactivation also may result from mutations of the genes that actually control inactivation, especially the XIST gene, which is responsible for inactivation of one of the two X chromosomes. Not surprisingly, affected females usually have a less severe phenotype than that seen in affected males.

The important characteristics of X-linked inheritance result from differential segregation of the X and Y chromosomes in males and females, and the differences in gene dosage. The most consistent characteristics include more severe phenotype in males than in females, transmission of disease through carrier females who are unaffected or less affected than males, and lack of male-to-male transmission (see Figure 30-16). The last is explained by transmission of the Y chromosome from fathers to sons, whereas the disease genes are located on the X chromosome. X-linked disorders traditionally have been divided into dominant and recessive subtypes, just as for autosomal single-gene disorders. This distinction was first made in fruit flies under experimental conditions but has never worked very well for human disorders. In a recent survey of more than 30 X-linked diseases, a remarkably wide range of penetrance was found, with many disorders intermediate between so-called X-linked dominant and recessive patterns [Dobyns et al., 2004]. On the basis of these and other arguments, use of these subtypes should be discontinued. The rules for X-linked inheritance have been modified to reflect this change (Box 30-1).

Box 30-1 Rules for X-Linked Inheritance in Humans

Rules Related to Segregation of the X and Y Chromosomes

Hemizygous males transmit X chromosomes to daughters and Y chromosomes to sons

Male-to-male transmission of X–linked disorders cannot occur

Sons of hemizygous males never inherit the disorder

Daughters of affected males all are heterozygous (carriers or affected)

All affected males in a family are related through heterozygous females

Heterozygous females transmit X chromosomes to both sons and daughters

Fifty percent of sons of heterozygous females will be hemizygous males

Fifty percent of daughters of heterozygous females also will be heterozygous females

Rules Related to Penetrance, Expressivity, and Sex Ratio

Penetrance and Expressivity

In general, penetrance is higher and expressivity more severe in hemizygous males than in heterozygous females

Sex Ratio of Affected Persons

Males are predominantly affected when female penetrance is low

Female to male ratio is close to 2:1 when female penetrance and male survival are high

Females are predominantly affected, with prenatal lethality in males

In many instances, children with X-linked diseases present primarily or frequently to pediatric neurologists, including those with adrenoleukodystrophy, the Duchenne and Becker forms of muscular dystrophy, fragile X syndrome, many X-linked mental retardation syndromes, two forms of X-linked lissencephaly, and many others. These diseases all result in more severe phenotypes in males than in females. Another class of X-linked disorders, “X-linked, male lethal,” have proved to be particularly important in pediatric neurology. These diseases are observed almost exclusively in females and are thought to cause prenatal lethality in males or to cause a much more severe phenotype that is not recognized as the same disorder. Examples are Aicardi’s, Goltz’s, and Rett’s syndromes and orofaciodigital syndrome type I.

Genomic Imprinting

In most single-gene disorders, the expression of a trait or disease is expected to be the same, regardless of whether the gene was inherited from the mother or the father. Significant differences in expression based on the gender of the transmitting parent, however, have been observed in several disorders. This phenomenon is known as imprinting and reflects differences in the state of the maternal and paternal contributions to the genome, especially differential methylation of the maternally and paternally derived chromosomes. An imprinted gene can be imprinted or differentially methylated (i.e., differentially silenced) in all cells of the body, or only in selected tissues, such as brain.

This differential silencing means that imprinted genes will be expressed from the maternally derived gene or from the paternally derived gene, but not from both. So if the functioning copy of an imprinted gene is lost owing to a deletion or mutation, the affected person is left with no functioning gene, and a disease phenotype will result. The underlying mechanisms are under study but still are not well understood [Hall, 1990; Jiang et al., 2004]. Imprinting disorders result from deletions or other types of mutations of genes within imprinted regions (or in the imprinting control regions). The most common of these result in infantile developmental disorders relevant for pediatric neurologists, such as Beckwith–Wiedemann syndrome due to defects of imprinted genes on 11p15.5 [Weksberg et al., 2003], and Angelman’s and Prader–Willi syndromes due to defects of imprinted genes on 15q11.2–q13 [Amos-Landgraf et al., 2006]. These disorders are reviewed elsewhere in this book.

Uniparental Disomy

Defects of imprinted genes also may result from a rare mutation type known as uniparental disomy (UPD). UPD is defined as the presence of a diploid cell line containing two chromosome homologs inherited from the same parent. It is believed to result from nondisjunction, which produces trisomy for a particular chromosome. Trisomy is followed by loss of one of the three homologs, reducing the chromosome number back to normal. This mutation is known as uniparental isodisomy when the two homologs are identical, and uniparental heterodisomy when they are different (as a result of crossing over, different regions of the affected chromosomes usually are involved).

When UPD involves a chromosome with an imprinted region, problems occur. For example, a child may inherit two paternally derived chromosomes, in which case no maternally derived chromosome will be present. Any genes that are normally expressed only from the maternally derived gene will not be expressed at all. This is one cause of Angelman’s syndrome. The same type of problem occurs in reverse if the child inherits two maternally inherited chromosomes. No paternally derived genes will be present, and a disease will occur. This is one cause of Prader–Willi syndrome [Amos-Landgraf et al., 2006].

Very rarely, UPD can result in a genetic disease by causing homozygosity of a recessive disease gene. Because the two chromosomes are identical, they are homologous at all loci. If the involved chromosome contains any recessive disease loci, the person with the mutation will be homozygous and therefore affected.

Mosaicism

Mosaic is a term used to refer to an individual organism or tissue that contains two or more cell lines that differ in DNA sequence, although they are derived from a single zygote. All organisms begin with a specific DNA sequence in the cell of origin or zygote. As cell division proceeds, some mutations occur that produce small differences among different cell lines. The presence of two or more cell lines differing in their DNA sequence but derived from a single zygote is known as mosaicism. Mosaicism is clinically important in many disorders and probably explains some unusual diseases in which only part of the body appears to be affected with a birth defect or genetic disease. A good example is segmental neurofibromatosis.

This phenomenon can involve any tissue or group of tissues in the body. When mosaicism is found in lymphocytes, fibroblasts, or other somatic cells of the body, it is designated somatic mosaicism. The mosaic individual typically demonstrates at least mild signs of disease. When mosaicism is found only in germ cells (egg, sperm), it is known as gonadal mosaicism. The mosaic individual usually is identified as the parent of two or more children with a genetic disease, despite having no signs of the disease clinically or on mutation analysis. Mosaicism may begin with a somatic mutation in the germline of the affected person, which then persists in the clonal descendants of that cell, including a proportion of the ova or sperm. When the mutation exists only in the germline, the parent has no signs of the disease but may conceive multiple affected children. This phenomenon has been seen frequently in Duchenne muscular dystrophy due to mutations of the DMD gene, the autosomal-dominant form of osteogenesis imperfecta associated with mutations of the COL1A1 or COL1A2 gene, and X-linked lissencephaly due to mutations of the DCX gene. The distinction between somatic and germline mosaicism is most likely artificial, however, because standard evaluations examine very few tissues.

Complex Inheritance

The most important attributes of complex or multifactorial inheritance are lack of a clear pattern of inheritance in single families, although more than one relative may be affected, and a relatively low risk for first-degree relatives (approximately the square root of the population risk), typically in the range of 1–5 percent. This form of inheritance results from variation at two or more loci with two or more alleles each, often with a prominent environmental influence. This pattern also is referred to as polygenic (we prefer oligogenic) or multifactorial inheritance. Examples of traits inherited in this pattern are head circumference, autism, neural tube defects, and common forms of epilepsy. Some are continuous traits that can be measured, such as head circumference, whereas others fall into non-overlapping groups, such as autism and epilepsy. For continuous traits, such as head circumference, children are likely to be intermediate between their parents, or closer to the mean than either parent (so-called regression to the mean).

Mitochondrial Inheritance

Mitochondria are the cellular organelles that are primarily responsible for cellular respiration and production of adenosine triphosphate, both essential for cellular energy management. Each mitochondrion contains multiple copies of a small 16.5-kb circular chromosome that codes for 13 proteins and many rRNA and tRNA genes that differ in sequence from the nuclear rRNA and tRNA genes. The proteins all are components of the respiratory chain, and the remainder of the respiratory pathway enzymes are encoded by nuclear genes.

The mitochondria in any one person are derived almost exclusively from the mother through the ovum. Each ovum contains hundreds of mitochondria, and each mitochondrion contains many copies of the circular mitochondrial chromosome. Sperm contain a few mitochondria, most of which are degraded rapidly by the proteasome-dependent protein degradation pathway of the ubiquitin system within the ovum after fertilization [Sutovsky et al., 2003]. A small paternal contribution of mitochondria, however, has been demonstrated in several species, such as sheep [Zhao et al., 2004]. The same likely is true for humans, as suggested by one example of paternal inheritance of a mitochondrial disorder [Schwartz and Vissing, 2002]. This must be very rare, however, owing to the small proportion of paternal compared with maternal mitochondria, and as indicated by studies in humans with mitochondrial diseases [Filosto et al., 2003].

In patients with mutations of mitochondrial genes, a variable proportion of the mitochondrial chromosomes carry the mutation. Thus, diseases caused by mutations in mitochondrial DNA exhibit strict maternal inheritance (with very rare exceptions) and usually will exhibit phenotypic variation within a family owing to variation in the proportion of mutant and normal mitochondria between individuals [Zeviani et al., 1989].

Maternal inheritance may be recognized because:

These criteria have been met for several diseases associated with mitochondrial DNA mutations, including Kearns–Sayre syndrome, Leber’s hereditary optic neuropathy, MELAS (mitochondrial encephalomyopathy, lactic acidosis, and strokelike episodes), and MERRF (myoclonic epilepsy with ragged-red fibers).

Standard of Care

All physicians have a professional responsibility to make certain that genetic counseling has been provided in appropriate situations and to ensure that the counseling meets current standards of practice [Directors and Directors, 1995; Parker, 2010]. Failure to provide this information may have tragic results. Perhaps the best example for pediatric neurologists is the birth of a second or even a third male with Duchenne muscular dystrophy in a family. Courts have upheld the principle of physician responsibility to provide accurate counseling on several occasions. For example, the parents of a child with Down syndrome claimed negligence because the mother had not been referred for prenatal diagnosis. In another case, parents who were tested for Tay–Sachs disease carrier status both were told that they were not carriers. They later had an affected child and filed a claim. Cases of this type are known as wrongful life claims.

Responsibility to Relatives

The responsibility to provide genetic counseling does not formally extend beyond the consultand, or person seeking genetic advice. People with genetic disorders are entitled to the same confidentiality as for persons with any other type of disease. Nevertheless, the need for confidentiality does not mean that no effort should be made to inform relatives of a common risk. Whenever the genetic evaluation suggests that other family members or their future children may be at risk, the consultand should be encouraged to contact those persons or ask the physician (or designee) to contact them and advise them to seek genetic evaluation.

A majority of people act responsibly in this regard, but exceptions do arise that may present an ethical dilemma for health-care practitioners providing the counseling.

Relevance for Pediatric Neurology

The obligation to provide accurate genetic counseling is particularly important for pediatric neurologists. Many neurologic and neuromuscular diseases in children have a genetic basis or a genetic component, including some of the most common problems seen in clinics. For disorders such as Duchenne muscular dystrophy, neurofibromatosis type 1, and tuberous sclerosis, the genetic basis and need for genetic counseling are well known. For others, the genetic basis or contribution is not widely recognized. Febrile seizures, benign rolandic epilepsy, and some types of primary generalized epilepsy may affect many persons in a family and appear to have autosomal-dominant inheritance, although penetrance is not complete. Mental retardation, microcephaly, and cerebral palsy have a significant genetic component, with recurrence risks of 5–10 percent. Most brain malformations are sporadic, but familial recurrence of almost every known type of brain malformation has been described.

Because of the possibility of recurrence in relatives, pediatric neurologists should take a genetic history and advise patients and parents of their genetic risks. Referrals from pediatric neurology clinics to genetics clinics should occur frequently. Even for single-gene disorders with a known risk of recurrence, referral usually is needed to provide an accurate presentation of methods of prevention, such as prenatal diagnosis when this is available, artificial insemination by donor, contraception, sterilization, and adoption. If the parents decide to terminate a pregnancy, continued professional support is an appropriate and important part of genetic counseling.

Genetic Risk

One of the most crucial steps in offering accurate genetic counseling is to estimate the risk of recurrence of a genetic disorder in other family members. Estimation is not difficult for most diseases with a single-gene pattern of inheritance, but even these may be complex because of late age at onset or incomplete penetrance. For many other disorders, empirical recurrence risk estimates are used. These risk figures are derived from previous experience with the same disorder. Although such figures generally are accurate, exceptions occur because of causal heterogeneity and lack of knowledge regarding many rare disorders. The recurrence risks for some of the more common diseases seen in pediatric neurology clinics were reviewed by Baraitser [Baraitser, 1997]. In any given family, the recurrence risk may be different; consultation with a geneticist or genetic counselor may be helpful in providing this information.

Prenatal Diagnosis

Prenatal diagnosis can now be performed for hundreds of genetic disorders, including many of those discussed in this chapter. The major methods used include chromosome analysis, enzyme assays and other biochemical tests, molecular genetic tests, and direct examination of the fetus by ultrasonography. The last is a far more rigorous procedure than routine prenatal ultrasonography and usually is referred to as high-resolution, level 2, or genetic ultrasonography. The purposes of prenatal diagnosis are to provide:

The results of prenatal tests are normal in more than 98 percent of pregnancies evaluated, and parents are reassured that the infant will be unaffected by the condition in question. Of course, the infant remains at risk for other disorders, just as do children born to any other parents. In a small proportion of cases, the fetus is indeed found to have a serious defect. Because effective prenatal therapy is not possible for most disorders, the parents then have the option of terminating the pregnancy.