cis-Acting Regulatory Sequences: Promoters and Enhancers

As already discussed, transcription in bacteria is regulated by the binding of proteins to cis-acting sequences (e.g., the lac operator) that control the transcription of adjacent genes. Similar cis-acting sequences regulate the expression of eukaryotic genes. These sequences have been identified in mammalian cells largely by the use of gene transfer assays to study the activity of suspected regulatory regions of cloned genes (Figure 6.18). The eukaryotic regulatory sequences are usually ligated to a reporter gene that encodes an easily detectable enzyme. The expression of the reporter gene following its transfer into cultured cells then provides a sensitive assay for the ability of the cloned regulatory sequences to direct transcription. Biologically active regulatory regions can thus be identified, and in vitro mutagenesis can be used to determine the roles of specific sequences within the region.

Figure 6.18

Identification of eukaryotic regulatory sequences. The regulatory sequence of a cloned eukaryotic gene is ligated to a reporter gene that encodes an easily detectable enzyme. The resulting plasmid is then introduced into cultured recipient cells by transfection. (more...)

Genes transcribed by RNA polymerase II have two core promoter elements, the TATA box and the Inr sequence, that serve as specific binding sites for general transcription factors. Other cis-acting sequences serve as binding sites for a wide variety of regulatory factors that control the expression of individual genes. These cis-acting regulatory sequences are frequently, though not always, located upstream of the TATA box. For example, two regulatory sequences that are found in many eukaryotic genes were identified by studies of the promoter of the herpes simplex virus gene that encodes thymidine kinase (Figure 6.19). Both of these sequences are located within 100 base pairs upstream of the TATA box: Their consensus sequences are CCAAT and GGGCGG (called a GC box). Specific proteins that bind to these sequences and stimulate transcription have since been identified.

Figure 6.19

A eukaryotic promoter. The promoter of the thymidine kinase gene of herpes simplex virus contains three sequence elements upstream of the TATA box that are required for efficient transcription: a CCAAT box and two GC boxes (consensus sequence GGGCGG). (more...)

In contrast to the relatively simple organization of CCAAT and GC boxes in the herpes thymidine kinase promoter, many genes in mammalian cells are controlled by regulatory sequences located farther away (sometimes more than 10 kilobases) from the transcription start site. These sequences, called enhancers, were first identified by Walter Schaffner in 1981 during studies of the promoter of another virus, SV40 (Figure 6.20). In addition to a TATA box and a set of six GC boxes, two 72-base-pair repeats located farther upstream are required for efficient transcription from this promoter. These sequences were found to stimulate transcription from other promoters as well as from that of SV40, and, surprisingly, their activity depended on neither their distance nor their orientation with respect to the transcription initiation site (Figure 6.21). They could stimulate transcription when placed either upstream or downstream of the promoter, in either a forward or backward orientation.

Figure 6.20

The SV40 enhancer. The SV40 promoter for early gene expression contains a TATA box and six GC boxes arranged in three sets of repeated sequences. In addition, efficient transcription requires an upstream enhancer consisting of two 72-base-pair (bp) repeats. (more...)

Figure 6.21

Action of enhancers. Without an enhancer, the gene is transcribed at a low basal level (A). Addition of an enhancer, E—for example, the SV40 72-base-pair repeats—stimulates transcription. The enhancer is active not only when placed just (more...)

The ability of enhancers to function even when separated by long distances from transcription initiation sites at first suggested that they work by mechanisms different from those of promoters. However, this has turned out not to be the case: Enhancers, like promoters, function by binding transcription factors that then regulate RNA polymerase. This is possible because of DNA looping, which allows a transcription factor bound to a distant enhancer to interact with RNA polymerase or general transcription factors at the promoter (Figure 6.22). Transcription factors bound to distant enhancers can thus work by the same mechanisms as those bound adjacent to promoters, so there is no fundamental difference between the actions of enhancers and those of cis-acting regulatory sequences adjacent to transcription start sites. Interestingly, although enhancers were first identified in mammalian cells, they have subsequently been found in bacteria—an unusual instance in which studies of eukaryotes served as a model for the simpler prokaryotic systems.

Figure 6.22

DNA looping. Transcription factors bound at distant enhancers are able to interact with general transcription factors at the promoter because the intervening DNA can form loops. There is therefore no fundamental difference between the action of transcription (more...)

The binding of specific transcriptional regulatory proteins to enhancers is responsible for the control of gene expression during development and differentiation, as well as during the response of cells to hormones and growth factors. One of the most thoroughly studied mammalian enhancers controls the transcription of immunoglobulin genes in B lymphocytes. Gene transfer experiments have established that the immunoglobulin enhancer is active in lymphocytes, but not in other types of cells. Thus, this regulatory sequence is at least partly responsible for tissue-specific expression of the immunoglobulin genes in the appropriate differentiated cell type.

An important aspect of enhancers is that they usually contain multiple functional sequence elements that bind different transcriptional regulatory proteins. These proteins work together to regulate gene expression. The immunoglobulin heavy-chain enhancer, for example, spans approximately 200 base pairs and contains at least nine distinct sequence elements that serve as protein-binding sites (Figure 6.23). Mutation of any one of these sequences reduces but does not abolish enhancer activity, indicating that the functions of individual proteins that bind to the enhancer are at least partly redundant. Many of the individual sequence elements of the immunoglobulin enhancer by themselves stimulate transcription in nonlymphoid cells. The restricted activity of the intact enhancer in B lymphocytes therefore does not result from the tissue-specific function of each of its components. Instead, tissue-specific expression results from the combination of the individual sequence elements that make up the complete enhancer. These elements include some cis-acting regulatory sequences that bind transcriptional activators that are expressed specifically in B lymphocytes, as well as other regulatory sequences that bind repressors in nonlymphoid cells. Thus, the immunoglobulin enhancer contains negative regulatory elements that inhibit transcription in inappropriate cell types, as well as positive regulatory elements that activate transcription in B lymphocytes. The overall activity of the enhancer is greater than the sum of its parts, reflecting the combined action of the proteins associated with each of its individual sequence elements.

Figure 6.23

The immunoglobulin enhancer. The immunoglobulin heavy-chain enhancer spans about 200 bases and contains nine functional sequence elements (E, μE1-5, π, μB, and OCT), which together stimulate transcription in B lymphocytes.

Transcriptional Regulatory Proteins

The isolation of a variety of transcriptional regulatory proteins has been based on their specific binding to promoter or enhancer sequences. Protein binding to these DNA sequences is commonly analyzed by two types of experiments. The first, footprinting, was described earlier in connection with the binding of RNA polymerase to prokaryotic promoters (see Figure 6.3). The second approach is the electrophoretic-mobility shift assay, in which a radiolabeled DNA fragment is incubated with a protein preparation and then subjected to electrophoresis through a nondenaturing gel (Figure 6.24). Protein binding is detected as a decrease in the electrophoretic mobility of the DNA fragment, since its migration through the gel is slowed by the bound protein. The combined use of footprinting and electrophoretic-mobility shift assays has led to the correlation of protein-binding sites with the regulatory elements of enhancers and promoters, indicating that these sequences generally constitute the recognition sites of specific DNA-binding proteins.

Figure 6.24

Electrophoretic-mobility shift assay. A sample containing radiolabeled fragments of DNA is divided into two, and one half of the sample is incubated with a protein that binds to a specific DNA sequence. Samples are then analyzed by electrophoresis in (more...)

One of the prototypes of eukaryotic transcription factors was initially identified by Robert Tjian and his colleagues during studies of the transcription of SV40 DNA. This factor (called Sp1, for specificity protein 1) was found to stimulate transcription from the SV40 promoter, but not from several other promoters, in cell-free extracts. Then, stimulation of transcription by Sp1 was found to depend on the presence of the GC boxes in the SV40 promoter: If these sequences were deleted, stimulation by Sp1 was abolished. Moreover, footprinting experiments established that Sp1 binds specifically to the GC box sequences. Taken together, these results indicate that the GC box represents a specific binding site for a transcriptional activator—Sp1. Similar experiments have established that many other transcriptional regulatory sequences, including the CCAAT sequence and the various sequence elements of the immunoglobulin enhancer, also represent recognition sites for sequence-specific DNA-binding proteins (Table 6.2).

Table 6.2

Examples of Transcription Factors and Their DNA-Binding Sites.

The specific binding of Sp1 to the GC box not only established the action of Sp1 as a sequence-specific transcription factor; it also suggested a general approach to the purification of transcription factors. The isolation of these proteins initially presented a formidable challenge because they are present in very small quantities (e.g., only 0.001% of total cell protein) that are difficult to purify by conventional biochemical techniques. This problem was overcome in the purification of Sp1 by DNA-affinity chromatography (Figure 6.25). Multiple copies of oligonucleotides corresponding to the GC box sequence were bound to a solid support, and cell extracts were passed through the oligonucleotide column. Because Sp1 bound to the GC box with high affinity, it was specifically retained on the column while other proteins were not. Highly purified Sp1 could thus be obtained and used for further studies, including partial determination of its amino acid sequence, which in turn led to cloning of the gene for Sp1.

Figure 6.25

Purification of Sp1 by DNA-affinity chromatography. A double-stranded oligonucleotide containing repeated GC box sequences is bound to agarose beads, which are poured into a column. A mixture of cell proteins containing Sp1 is then applied to the column; (more...)

The general approach of DNA-affinity chromatography, first optimized for the purification of Sp1, has been used successfully to isolate a wide variety of sequence-specific DNA-binding proteins from eukaryotic cells. Protein purification has been followed by gene cloning and nucleotide sequencing, leading to the accumulation of a great deal of information on the structure and function of these critical regulatory proteins.

Structure and Function of Transcriptional Activators

Because transcription factors are central to the regulation of gene expression, understanding the mechanisms of their action is a major area of ongoing research in cell and molecular biology. The most thoroughly studied of these proteins are transcriptional activators, which, like Sp1, bind to regulatory DNA sequences and stimulate transcription. In general, these factors have been found to consist of two domains: One region of the protein specifically binds DNA; the other activates transcription by interacting with other components of the transcriptional machinery (Figure 6.26). Transcriptional activators appear to be modular proteins, in the sense that the DNA binding and activation domains of different factors can frequently be interchanged using recombinant DNA techniques. Such manipulations result in hybrid transcription factors, which activate transcription by binding to promoter or enhancer sequences determined by the specificity of their DNA-binding domains. It therefore appears that the basic function of the DNA-binding domain is to anchor the transcription factor to the proper site on DNA; the activation domain then independently stimulates transcription by interacting with other proteins.

Figure 6.26

Structure of transcriptional activators. Transcriptional activators consist of two independent domains. The DNA-binding domain recognizes a specific DNA sequence, and the activation domain interacts with other components of the transcriptional machinery. (more...)

Many different transcription factors have now been identified in eukaryotic cells, as might be expected, given the intricacies of tissue-specific and inducible gene expression in complex multicellular organisms. Molecular characterization has revealed that the DNA-binding domains of many of these proteins are related to one another (Figure 6.27). Zinc finger domains contain repeats of cysteine and histidine residues that bind zinc ions and fold into looped structures (“fingers”) that bind DNA. These domains were initially identified in the polymerase III transcription factor TFIIIA but are also common among transcription factors that regulate polymerase II promoters, including Sp1. Other examples of transcription factors that contain zinc finger domains are the steroid hormone receptors, which regulate gene transcription in response to hormones such as estrogen and testosterone.

Figure 6.27

Families of DNA-binding domains. (A) Zinc finger domains consist of loops in which an α helix and a β sheet coordinately bind a zinc ion. (B) Helix-turn-helix domains consist of three (or in some cases four) helical regions. One helix (more...)

The helix-turn-helix motif was first recognized in prokaryotic DNA-binding proteins, including the E. coli catabolite activator protein (CAP). In these proteins, one helix makes most of the contacts with DNA, while the other helices lie across the complex to stabilize the interaction. In eukaryotic cells, helix-turn-helix proteins include the homeodomain proteins, which play critical roles in the regulation of gene expression during embryonic development. The genes encoding these proteins were first discovered as developmental mutants in Drosophila. Some of the earliest recognized Drosophila mutants (termed homeotic mutants in 1894) resulted in the development of flies in which one body part was transformed into another. For example, in the homeotic mutant called Antennapedia, legs rather than antennae grow out of the head of the fly (Figure 6.28). Genetic analysis of these mutants, pioneered by Ed Lewis in the 1940s, has shown that Drosophila contains nine homeotic genes, each of which specifies the identity of a different body segment. Molecular cloning and analysis of these genes then indicated that they contain conserved sequences of 180 base pairs (called homeoboxes) that encode the DNA-binding domains (homeodomains) of transcription factors. A wide variety of additional homeodomain proteins have since been identified in fungi, plants, and other animals, including humans. Vertebrate homeobox genes are strikingly similar to their Drosophila counterparts in both structure and function, demonstrating the highly conserved roles of these transcription factors in animal development.

Figure 6.28

The Antennapedia mutation. Antennapedia mutant flies have legs growing out of their heads in place of antennae. (A) Head of a normal fly. (B) Head of an Antennapedia mutant. (Courtesy of F. Rudolf Turner, Indiana University.)

Two other families of DNA-binding proteins, leucine zipper and helix-loop-helix proteins, contain DNA-binding domains formed by dimerization of two polypeptide chains. The leucine zipper contains four or five leucine residues spaced at intervals of seven amino acids, resulting in their hydrophobic side chains being exposed at one side of a helical region. This region serves as the dimerization domain for the two protein subunits, which are held together by hydrophobic interactions between the leucine side chains. Immediately following the leucine zipper is a region rich in positively charged amino acids (lysine and arginine) that binds DNA. The helix-loop-helix proteins are similar in structure, except that their dimerization domains are each formed by two helical regions separated by a loop. An important feature of both leucine zipper and helix-loop-helix transcription factors is that different members of these families can dimerize with each other. Thus, the combination of distinct protein subunits can form an expanded array of factors that can differ both in DNA sequence recognition and in transcription-stimulating activities. Both leucine zipper and helix-loop-helix proteins play important roles in regulating tissue-specific and inducible gene expression, and the formation of dimers between different members of these families is a critical aspect of the control of their function.

The activation domains of transcription factors are not as well characterized as their DNA-binding domains. Some, called acidic activation domains, are rich in negatively charged residues (aspartate and glutamate); others are rich in proline or glutamine residues. These activation domains are thought to stimulate transcription by interacting with general transcription factors, such as TFIIB or TFIID, thereby facilitating the assembly of a transcription complex on the promoter. For example, the activation domains of several transcription factors (including Sp1) have been shown to interact with TFIID by binding to TBP-associated factors (TAFs) (Figure 6.29). An important feature of these interactions is that different activators can bind to different general transcription factors or TAFs, providing a mechanism by which the combined action of multiple factors can synergistically stimulate transcription—a key feature of transcriptional regulation in eukaryotic cells.

Figure 6.29

Synergistic action of transcriptional activators. Different transcriptional activators can interact with the general transcription factor TFIID by binding to different TAFs.

Eukaryotic Repressors

Gene expression in eukaryotic cells is regulated by repressors as well as by transcriptional activators. Like their prokaryotic counterparts, eukaryotic repressors bind to specific DNA sequences and inhibit transcription. In some cases, eukaryotic repressors simply interfere with the binding of other transcription factors to DNA (Figure 6.30A). For example, the binding of a repressor near the transcription start site can block the interaction of RNA polymerase or general transcription factors with the promoter, which is similar to the action of repressors in bacteria. Other repressors compete with activators for binding to specific regulatory sequences. Some such repressors contain the same DNA-binding domain as the activator but lack its activation domain. As a result, their binding to a promoter or enhancer blocks the binding of the activator, thereby inhibiting transcription.

Figure 6.30

Action of eukaryotic repressors. (A) Some repressors block the binding of activators to regulatory sequences. (B) Other repressors have active repression domains that inhibit transcription by interactions with general transcription factors.

In contrast to repressors that simply interfere with activator binding, many repressors (called active repressors) contain specific functional domains that inhibit transcription via protein-protein interactions (Figure 6.30B). The first such active repressor was described in 1990 during studies of a gene called Krüppel, which is involved in embryonic development in Drosophila. Molecular analysis of the Krüppel protein demonstrated that it contains a discrete repression domain, which is linked to a zinc finger DNA-binding domain. The Krüppel repression domain could be interchanged with distinct DNA-binding domains of other transcription factors. These hybrid molecules also repressed transcription, indicating that the Krüppel repression domain inhibits transcription via protein-protein interactions, irrespective of its site of binding to DNA.

Many active repressors have since been found to play key roles in the regulation of transcription in animal cells, in many cases serving as critical regulators of cell growth and differentiation. As with transcriptional activators, several distinct types of repression domains have been identified. For example, the repression domain of Krüppel is rich in alanine residues, whereas other repression domains are rich in proline or acidic residues. The functional targets of repressors are also diverse. Some repressors inhibit transcription by interacting with general transcription factors, such as TFIID; others are thought to interact with specific activator proteins.

The regulation of transcription by repressors as well as by activators considerably extends the range of mechanisms that control the expression of eukaryotic genes. One important role of repressors may be to inhibit the expression of tissue-specific genes in inappropriate cell types. For example, as noted earlier, a repressor-binding site in the immunoglobulin enhancer is thought to contribute to its tissue-specific expression by suppressing transcription in nonlymphoid cell types. Other repressors play key roles in the control of cell proliferation and differentiation in response to hormones and growth factors (see Chapters 13 and 14).

Relationship of Chromatin Structure to Transcription

In the preceding discussion, the transcription of eukaryotic genes was considered as if they were present within the nucleus as naked DNA. However, this is not the case. The DNA of all eukaryotic cells is tightly bound to histones, forming chromatin. The basic structural unit of chromatin is the nucleosome, which consists of 146 base pairs of DNA wrapped around two molecules each of histones H2A, H2B, H3, and H4, with one molecule of histone H1 bound to the DNA as it enters the nucleosome core particle (see Figure 4.9). The chromatin is then further condensed by being coiled into higher-order structures organized into large loops of DNA. This packaging of eukaryotic DNA in chromatin clearly has important consequences in terms of its availability as a template for transcription, so chromatin structure is a critical aspect of gene expression in eukaryotic cells. Indeed, both activators and repressors regulate transcription in eukaryotes not only by interacting with general transcription factors and other components of the transcriptional machinery, but also by inducing changes in the structure of chromatin.

The relationship between chromatin structure and transcription is evident at several levels. First, actively transcribed genes are found in decondensed chromatin, corresponding to the extended 10-nm chromatin fibers discussed in Chapter 4 (see Figure 4.10). For example, microscopic visualization of the polytene chromosomes of Drosophila indicates that regions of the genome that are actively engaged in RNA synthesis correspond to decondensed chromosome regions (Figure 6.31). Similarly, actively transcribed genes in vertebrate cells are present in a decondensed fraction of chromatin that is more accessible to transcription factors than is the rest of the genome.

Figure 6.31

Decondensed chromosome regions in Drosophila. A light micrograph showing decondensed regions of polytene chromosomes (arrows), which are active in RNA synthesis. (Courtesy of Joseph Gall, Carnegie Institute.)

Decondensation of chromatin, however, is not sufficient to make the DNA an accessible template for transcription. Even in decondensed chromatin, actively transcribed genes remain bound to histones and packaged in nucleosomes, so transcription factors and RNA polymerase are still faced with the problem of interacting with chromatin rather than with naked DNA. The tight winding of DNA around the nucleosome core particle is a major obstacle to transcription, affecting both the ability of transcription factors to bind DNA and the ability of RNA polymerase to transcribe through a chromatin template. This inhibitory effect of nucleosomes is relieved by acetylation of histones and by the binding of two nonhistone chromosomal proteins (called HMG-14 and HMG-17) to nucleosomes of actively transcribed genes. (HMG stands for high-mobility group proteins; these proteins migrate rapidly during gel electrophoresis.) Additional proteins called nucleosome remodeling factors facilitate the binding of transcription factors to chromatin by altering nucleosome structure.

Acetylation of histones has been correlated with transcriptionally active chromatin in a wide variety of cell types (Figure 6.32). The core histones (H2A, H2B, H3 and H4) have two domains: a histone fold domain, which is involved in interactions with other histones and in wrapping DNA around the nucleosome core particle, and an amino-terminal tail domain, which extends outside of the nucleosome. The amino-terminal tail domains are rich in lysine and can be modified by acetylation at specific lysine residues. Acetylation reduces the net positive charge of the histones, and may weaken their binding to DNA as well as altering their interactions with other proteins. Importantly, recent experiments have provided direct evidence that histone acetylation facilitates the binding of transcription factors to nucleosomal DNA, indicating that histone acetylation increases the accessibility of chromatin to DNA-binding proteins. In addition, direct links between histone acetylation and transcriptional regulation have come from experiments showing that transcriptional activators and repressors are associated with histone acetyltransferases and deacetylases, respectively. This association was first revealed by cloning a gene encoding a histone acetyltransferase from Tetrahymena. Unexpectedly, the sequence of this histone acetyltransferase revealed that it was closely related to a previously known yeast transcriptional coactivator called Gcn5p, which stimulates transcription in association with several different sequence-specific transcriptional activators. Further experiments revealed that Gcn5p itself has histone acetyltransferase activity, suggesting that transcriptional activation results directly from histone acetylation. These results have been extended by the finding that several mammalian transcriptional coactivators are also histone acetyltransferases, as is a general transcription factor (TAFII250, a component of TFIID). Conversely, histone deacetylases (which remove the acetyl groups from histone tails) are associated with transcriptional repressors in both yeast and mammalian cells. Histone acetylation is thus regulated by both transcriptional activators and repressors, indicating that it plays a key role in eukaryotic gene expression.

Figure 6.32

Histone acetylation. (A) The core histones have histone-fold domains, which interact with other histones and with DNA in the nucleosome, and N-terminal tails, which extend outside of the nucleosome. The N-terminal tails of the core histones (e.g., H3) (more...)

Nucleosome remodeling factors are protein complexes that facilitate the binding of transcription factors by altering nucleosome structure (Figure 6.33). The mechanism of action of nucleosome remodeling factors is not yet clear, but they appear to increase the accessibility of nucleosomal DNA to other proteins (such as transcription factors) without removing the histones. One possibility is that they catalyze the sliding of histone octamers along the DNA molecule, thereby repositioning nucleosomes to facilitate transcription factor binding. The mechanisms by which nucleosome remodeling factors are targeted to actively transcribed genes also remain to be established, although some studies suggest that they can be brought to enhancer or promoter sites in association with transcriptional activators or as components of the RNA polymerase II holoenzyme (see Figure 6.14).

Figure 6.33

Nucleosome remodeling factors. Nucleosome remodeling factors facilitate the binding of transcription factors to chromatin by repositioning nucleosomes on the DNA.

Perhaps surprisingly, the packaging of DNA in nucleosomes does not present an impassable barrier to transcriptional elongation by RNA polymerase, which is able to transcribe through a nucleosome core by disrupting histone-DNA contacts. The ability of RNA polymerase to transcribe chromatin templates is facilitated by acetylation of histones and by the association of the nonhistone chromosomal proteins HMG-14 and HMG-17 with the nucleosomes of actively transcribed genes. The binding sites of these proteins on nucleosomes overlap the binding site of histone H1, and HMG-14 and HMG-17 appear to stimulate transcription by altering the interaction of histone H1 with nucleosomes to maintain a decondensed chromatin structure that facilitates transcription through a nucleosome template. As with nucleosome remodeling factors, the signals that target HMG-14 and HMG-17 to actively transcribed genes remain to be elucidated by future research.

DNA Methylation

The methylation of DNA is another general mechanism by which control of transcription in vertebrates is linked to chromatin structure. Cytosine residues in vertebrate DNA can be modified by the addition of methyl groups at the 5-carbon position (Figure 6.34). DNA is methylated specifically at the C's that precede G's in the DNA chain (CpG dinucleotides). This methylation is correlated with reduced transcriptional activity of genes that contain high frequencies of CpG dinucleotides in the vicinity of their promoters. Methylation inhibits transcription of these genes via the action of a protein, MeCP2, that specifically binds to methylated DNA and represses transcription. Interestingly, MeCP2 functions as a complex with histone deacetylase, linking DNA methylation to alterations in histone acetylation and nucleosome structure.

Figure 6.34

DNA methylation. A methyl group is added to the 5-carbon position of cytosine residues in DNA.

Although DNA methylation is capable of inhibiting transcription, its general significance in gene regulation is unclear. In many cases, methylation of inactive genes is thought to be a consequence, rather than the primary cause, of their lack of transcriptional activity. However, an important regulatory role of DNA methylation has been established in the phenomenon known as genomic imprinting, which controls the expression of some genes involved in the development of mammalian embryos. In most cases, both the paternal and maternal alleles of a gene are expressed in diploid cells. However, there are a few imprinted genes (over two dozen have been described in mice and humans) whose expression depends on whether they are inherited from the mother or from the father. In some cases, only the paternal allele of an imprinted gene is expressed, and the maternal allele is transcriptionally inactive. For other imprinted genes, the maternal allele is expressed and the paternal allele is inactive.

Although the biological role of genomic imprinting is uncertain, DNA methylation appears to distinguish between the paternal and maternal alleles of imprinted genes. A good example is the gene H19, which is transcribed only from the maternal copy (Figure 6.35). The H19 gene is specifically methylated during the development of male, but not female, germ cells. The union of sperm and egg at fertilization therefore yields an embryo containing a methylated paternal allele and an unmethylated maternal allele of the gene. These differences in methylation are maintained following DNA replication by an enzyme that specifically methylates CpG sequences of a daughter strand that is hydrogen-bonded to a methylated parental strand (Figure 6.36). The paternal H19 allele therefore remains methylated, and transcriptionally inactive, in embryonic cells and somatic tissues. However, the paternal H19 allele becomes demethylated in the germ line, allowing a new pattern of methylation to be established for transmittal to the next generation.

Figure 6.35

Genomic imprinting. The H19 gene is specifically methylated during development of male germ cells. Therefore, sperm contain a methylated H19 allele and eggs contain an unmethylated allele. Following fertilization, the methylated paternal allele remains (more...)

Figure 6.36

Maintenance of methylation patterns. In parental DNA, both strands are methylated at complementary CpG sequences. Following replication, only the parental strand of each daughter molecule is methylated. The newly synthesized daughter strands are then (more...)

Box

Key Experiment: Isolation of a Eukaryotic Transcription Factor.

Box

Molecular Medicine: The Pit-1 Transcription Factor and Growth Hormone Deficiency.