Genetic material consists of double-helical DNA, with each strand composed of the deoxyribonucleotides, A, G, T, C (adenylic acid, guanylic acid, thymidylic acid, and cytidylic acid, respectively) and a sugar phosphate backbone (Fig. 10.1). This backbone runs along the outside of the helix; the base components of the nucleotides face into the interior of the helix. In that central portion, the critical hydrogen bonds between A of one strand and its complement T of the other, or G of one strand and its complement C of the other, are formed. The human genome contains about 3 billion of these base pairs. The base-pairing rules (A with T and G with C) are critical in information transfer, both during DNA replication and in the transcription of the DNA code into RNA. Each nucleotide is linked to its neighbor via a sugar–phosphate linkage involving the 3′ carbon of one sugar attached to a phosphate group. That phosphate is linked in turn to the 5′ carbon of the next nucleotide. Thus, the nucleotide linkages are referred to as 3′ to 5′. This linkage gives a DNA (or RNA) strand its polarity or directionality. This polarity determines the direction in which the DNA is synthesized during replication and read during the transcription of DNA into RNA. Thus, a DNA sequence 3′ to 5′ is decoded during transcription in a 5′ to 3′ direction in RNA.

The great majority of mammalian genes coding for a protein product have split coding regions; that is, the coding regions are discontinuous. The beta-globin gene of adult hemoglobin has three coding regions, termed exons, and two intervening sequences (IVS), called introns. The coding regions are divided between the codons for amino acids 30 and 3l of the l46 amino-acid chain and the codons for amino acids l04 and l05. Although only about 450 nucleotides are necessary to encode the protein (3 nucleotides/amino acid), because of the introns, the gene contains roughly 1,500 nucleotides. Yet, this is a very simple gene. Many genes now have been described with 10 or more introns (the gene for the pro-alpha 1 collagen chain has more than 50 introns), and some genes have a total size of more than 200 kilobases (kb) or 200 thousand base pairs. In fact, dystrophin, the gene affected in Duchenne muscular dystrophy, has over 2 million base pairs. Some large introns contain entire genes within their boundaries. On the other hand, certain genes, such as those encoding histones and interferons, are small and do not contain introns. These genes are unexplained exceptions to the split gene rule.

The mechanisms of expression for protein-coding genes are exemplified by the globins (Fig. 10.2). The entire gene, including the two introns, is transcribed into a precursor RNA. This RNA is immediately modified at both its ends (called 5′ and 3′ ends). The modification at the 5′ end is an addition of a methylated G in an unusual triphosphate linkage. This modification, which is specific for messenger RNA, is called
5′ capping. The modification at the 3′ end is the addition of about l50 A residues (the poly A tail). The 5′ cap is thought to be important in the translation of the RNA into protein, whereas the 3′ poly A tail may have a role in RNA stability. Next, the introns are precisely spliced out of the RNA, thereby approximating the coding region sequences. The precise mechanism, and the enzymes involved in RNA splicing, is now well understood.

RNA splicing also is important in transporting the RNA from the nucleus to the cytoplasm, where it will be translated into protein; nucleotide sequences near the exon–intron junctions are important in normal splicing. Translation then occurs on the fully processed messenger RNA (mRNA) in the cytoplasm to produce the peptide chain, which may need post-translational modification, such as glycosylation or farnysylation, and/or subcellular targeting for proper functioning. Mutations that interfere with any of these steps have the potential to cause human disease.

The genetic code, as found in the mRNA, is critical to translation. There are 64 possible combinations in a code composed of three nucleotides. One combination, AUG (note that U is substituted for T in RNA) is the initiation codon and always places methionine as the initial amino acid in a protein. Methionine usually is not the first amino acid in the finished protein because it often is removed when the nascent protein is in the early stages of production. The code also contains three terminator codons, UAG, UAA, and UGA. These codons cannot be decoded into an amino acid and, therefore, protein synthesis terminates whenever one of these triplets is present in the reading frame. A mutation in the coding region of a gene that produces a terminator codon is called a nonsense mutation. A nucleotide change that produces a mutation by changing the sequence of an amino acid is called a missense mutation. The remaining 60 codons can be decoded so that an amino acid is inserted into the growing or nascent polypeptide chain. The degeneracy of the genetic code allows for six different codons for leucine and serine, and four codons for many other of the total of 20 amino acids. This degeneracy permits some variation in the coding DNA sequence, without necessarily producing a change in the amino acid encoded. The relationship between the nucleotide triplet in DNA, its complement in mRNA, and the amino acid designated by that triplet is demonstrated for the sixth codon of the beta-globin chain (Fig. 10.3). CTC in DNA is decoded as GAG in mRNA and as glutamate in the protein. A single nucleotide substitution (T to A) causes a missense mutation that leads to sickle cell anemia, as discussed later.

Any discussion of mutations and their consequences requires a brief introduction to normal variation. Much has been learned about normal variation at the protein or enzyme level. Scientists are learning that variations in the DNA are even more extensive than we imagined from protein data. A number of different types of variation have been found. Originally, common normal variation in DNA was termed DNA polymorphism, and it could be detected in the laboratory as a restriction fragment length polymorphism (RFLP). This variation usually is the result of single nucleotide substitutions and is detected if the variation affects a restriction endonuclease site. The insertion or deletion of a large number of nucleotides also can be detected as an RFLP. One RFLP that affects a site that is cleaved by the restriction enzyme HincII near the beta-globin gene is shown (Fig. 10.4). When genomic DNA is digested with HincII, electrophoresed, and hybridized with a radioactive probe containing the beta-globin gene, chromosomes that contain this site yield a 3.7-kb fragment, and those that lack the site demonstrate an 8-kb fragment. Because this polymorphism is very frequent, nearly 50% of individuals in the population are heterozygous and demonstrate both fragments by this analysis.

Southern blotting (Box 10.1) is a technique used to find a number of DNA polymorphisms in a gene cluster and in and around a large number of other genes. Thus, normal variation in DNA is extensive. The 50,000 nucleotides of the paternal beta-globin gene cluster contain l00 or more nucleotide differences from the 50,000 maternal nucleotides of this cluster. In total, the haploid DNA derived from one parent contains 3 to 10 million differences (single nucleotide substitutions) from the genetic material derived from one’s other parent.

Even more important forms of DNA polymorphisms are variable number of tandem repeats (VNTRs) and simple sequence repeats (SSRs). VNTRs are repeats of 15 to 50 nucleotides, and the repeat number at a locus may vary widely (Fig. 10.5). SSRs are variations in the number of repeats of very short sequences, usually two to four nucleotides. Often, a VNTR locus may contain many possible alleles, perhaps 100 or more, if a population group is studied. SSR loci may contain 5 to 10 alleles. The variation in VNTRs and SSRs contrasts with that of most common RFLPs, which are dimorphisms. Because of the large number of alleles at many VNTR loci, a particular genotype at four or five such loci may be present in only one person in several million individuals in the population. This normal DNA variation has had substantial practical value in the forensic investigation of criminal cases. DNA analysis has played a major role in the conviction or exclusion of many individuals accused of rape or other crimes.

DNA polymorphisms also have been used as markers to trace the inheritance of particular regions of the genome in families. When a large number of such markers are studied in families affected with single-gene diseases, such as Huntington disease, it has been possible to find a single marker that is co-inherited with the disease gene in the affected families. After this marker is mapped to some particular chromosome segment, the disease gene is ipso facto mapped to the same location. In this way, the Huntington disease (HD) locus was mapped to the short arm of chromosome 4, the neurofibromatosis locus (NF-1) to the centromeric region of chromosome 17, and the cystic fibrosis locus (CF) to the long arm of chromosome 7, to name a few successful examples of this approach.

The genes responsible for these diseases have been isolated by this method of finding the gene’s location in the genome before knowing the function of the encoded product. One of the great promises of the Human Genome Project is that this type of reverse genetic analysis, called positional cloning, will be very easy to execute. Once a gene is mapped to a region of the chromosome, and the DNA sequence of that region is known, finding genes in that region can be done quickly using a combination of computational and experimental methods. For example, if a metabolic disorder known to be associated with an electron transport chain defect is localized to a specific chromosomal region, a researcher might examine the DNA sequence in that region carefully for gene products that are predicted to have a role in energy metabolism and then directly sequence those genes in the DNA from affected patients to determine causation. This sequence of events has been performed a number of times with great rapidity since the genome sequence has become available.

In the late 1980s, the idea of determining the exact DNA sequence of the entire human genome was discussed. By the late 1990s, the project was well underway. Two remarkable reports appeared in February 2001, describing the sequencing and analysis of most of the human genome, completed several years earlier than anticipated. The availability of this sequence or blueprint for life is certain to revolutionize the impact of genetics in pediatrics and in medicine in general. For that reason, we review selected aspects of the human genome sequencing project. The strategies for mapping and sequencing the human genome are described in Box 10.2, Figure 1.