Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION. Comparisons of related genes in different species show 4.3

Transcription

1 14 A self-splicing intron in an rrna of the large ribosomal subunit. Kenneth Eward/Photo Researchers, Inc. The Interrupted Gene Edited by Donald Forsdyke w NOT FOR SALE chapter OR DISTRIBUTION outline 4.1 Introduction The positions of introns are usually conserved when homologous genes are compared between different organisms. The lengths of the corresponding introns may 4.2 An Interrupted Gene Consists of Exons and Introns Introns are removed by RNA splicing, which occurs in vary greatly, though. Jones & Bartlett cis Learning, individual RNA LLC molecules. Jones & Bartlett Introns Learning, usually do not encode LLC proteins. NOT FOR SALE OR Mutations DISTRIBUTION in exons can affect polypeptide sequence; NOT FOR 4.5 SALE Exon OR Sequences DISTRIBUTION Under Negative Selection mutations in introns can affect RNA processing and hence may influence the sequence and/or production Are Conserved but Introns Vary of a polypeptide. Comparisons of related genes in different species show 4.3 that the sequences of the corresponding exons are Exon and Intron Base Compositions Differ usually conserved but the sequences of the introns are The four rules for Jones DNA base & composition Bartlett are Learning, the first LLC much less similar. and second parity rules, the cluster rule, and the GC rule. Exons and NOT introns FOR can be SALE distinguished OR DISTRIBUTION Introns evolve much on the NOT more FOR rapidly SALE than exons OR DISTRIBUT because of the lack of selective pressure to produce a basis of all rules except the first. polypeptide with a useful sequence. The second parity rule suggests an extrusion of structured stem-loop segments from duplex DNA, which 4.6 Exon Sequences Under Positive Selection Vary would be greater in introns. but Introns Are Conserved Jones The rules & relate Bartlett to genomic Learning, characteristics, LLCor Under Jones positive & Bartlett selection an Learning, individual with LLC an NOT pressures, FOR SALE that constitute OR DISTRIBUTION the genome phenotype. NOT advantageous FOR SALE mutation OR survives DISTRIBUTION (i.e., is able to 4.4 produce more fertile progeny) relative to others Organization of Interrupted Genes May Be without the mutation. Conserved Due to intrinsic genomic pressures, such as that which Introns can be detected when genes are compared with conserves the potential to extrude stem-loops from duplex DNA, introns evolve more slowly than exons that their RNA transcription products by either restriction Jones & Bartlett mapping, Learning, electron microscopy, LLC or sequencing. Jones & Bartlett are under Learning, positive selection LLCpressure

2 Jones & Bartlett chapter Learning, outline, continued LLC 4.7 Genes Show a Wide Distribution of Sizes Due Primarily to Intron Size and Number Variation Most genes are uninterrupted in Saccharomyces cerevisiae but are interrupted in Jones multicellular & eukaryotes. Bartlett Learning, LLC Exons are usually short, typically encoding fewer than 100 amino acids. Introns are short in unicellular/oligocellular eukaryotes but can be many kilobases (kb) in multicellular eukaryotes. The overall length of a gene is determined largely by its Jones introns. & Bartlett Learning, LLC 4.8 Some NOT DNA FOR Sequences SALE Encode OR DISTRIBUTION More Than One Polypeptide The use of alternative initiation or termination codons allows multiple variants of a polypeptide chain. Different polypeptides can be produced from the same Jones & Bartlett sequence Learning, of DNA when the LLC mrna is read in different NOT FOR SALE reading OR frames DISTRIBUTION (as two overlapping genes). Otherwise identical polypeptides, differing by the presence or absence of certain regions, can be generated by differential (alternative) splicing when certain exons are included or excluded. This may take the form of including or excluding individual exons, or of choosing between alternative exons. The exons of some genes appear homologous to the exons of others, suggesting a common exon ancestry Members of a Gene Family Jones Have a Common & Bartlett Learning, LL Organization A set of homologous genes should share common features that preceded their evolutionary separation. All globin genes have a common form of organization with three exons and two introns, suggesting that they are descended from a single ancestral gene. Intron positions in the actin gene family are highly variable, NOT which FOR suggests SALE that OR introns DISTRIBUTION do not separate functional domains There Are Many Forms of Information in DNA Genetic information includes not only that related to characters corresponding to the conventional phenotype, but also Learning, that related to LLC characters Jones & Bartlett (pressures) corresponding to the genome phenotype. In certain contexts, the definition of the gene can be seen as reversed from one gene one protein to one protein one gene. Positional information may be important in development. Sequences transferred horizontally Jones & from Bartlett other Learning, LL 4.9 Some Exons Can Be NOT Equated FOR with SALE Protein OR Functional DISTRIBUTION species to the germline NOT could FOR land in introns SALE or OR DISTRIBUT intergenic DNA and then transfer vertically through Domains the generations. Some of these sequences may be Proteins can consist of independent functional modules, the boundaries of which, in some cases, can be 4.12 Summary involved in intracellular nonself-recognition. equated with those of exons. is operational with respect to phenotype. 4.1 Introduction Protein-encoding sequences can be interrupted, as can the 59 and 39 sequences (UTRs) The simplest form of a gene is a length of DNA Jones & Bartlett Learning, that directly LLC corresponds to its polypeptide Jones & Bartlett that flank Learning, the protein-encoding LLC sequences product. Bacterial genes are almost NOT always FOR of SALE within OR mrna. DISTRIBUTION The interrupting sequences this type, in which a continuous sequence of are removed from the primary (RNA) 3N bases encodes a polypeptide of N amino transcript (or pre-mrna) during gene acids. However, in eukaryotes ribosomal RNAs expression, generating an mrna that includes (rrnas), transfer RNAs (trnas), and messenger Jones RNAs (mrnas) & Bartlett are Learning, first synthesized LLC the polypeptide product Jones as determined & Bartlett by Learning, the LL a continuous base sequence corresponding to as long NOT precursor FOR SALE transcripts OR DISTRIBUTION that are subsequently shortened (see the chapter titled RNA ing an interrupted protein-encoding gene are genetic code. The sequences NOT FOR of SALE DNA compris- OR DISTRIBUT Splicing and Processing). Thus eukaryotic genes divided into the two categories depicted in are much longer than the functional transcripts FIGURE 4.1: they produce. It is reasonable to assume that Exons are the sequences retained in the Jones & the Bartlett shortening Learning, involved LLC a trimming of additional, perhaps OR DISTRIBUTION regulatory, sequences at the NOT script FOR SALE starts and OR ends DISTRIBUTION with exons that Jones mature & RNA Bartlett product. Learning, A mature LLC tran- NOT FOR SALE 59 and/or 39 ends of transcripts, leaving the correspond to the 59 and 39 ends of the rrna or protein-encoding sequence of the RNA. precursor intact. Introns are the intervening sequences However, as it happens a eukaryotic that are removed when the primary Jones & Bartlett Learning, gene can LLC include additional sequences Jones that & Bartlett RNA Learning, transcript LLC is processed to give the lie both within and outside the region NOT that FOR SALE OR mature DISTRIBUTION RNA product. 82 chapter 4 The Interrupted Gene..

3 The exon sequences are in the same order in Jones the & Bartlett gene and Learning, in the RNA, but LLCan interrupted DNA NOT FOR gene SALE is longer OR DISTRIBUTION than its mature RNA product because of the presence of the introns. RNA synthesis pre-mrna The processing of interrupted genes requires an additional step that is not necessary in uninterrupted genes. The DNA of an interrupted gene is transcribed to an RNA copy introns Splicing removes (a transcript) that is exactly NOT complementary FOR SALE OR to DISTRIBUTION mrna the original DNA sequence. This RNA is only a precursor, though; it cannot yet be used to Protein synthesis produce a polypeptide. First, the introns must be removed from the RNA to give a messenger Protein RNA that consists only of a series of exons. This process is called RNA splicing (see the chapter Length of precursor RNA (not mrna) defines region of gene titled Genes Encode RNAs and Polypeptides) and involves precisely deleting the introns from the Individual coding regions are separated in gene primary transcript and then joining the ends of the RNA on either side of each intron to form FIGURE 4.1 Interrupted genes are expressed via a precursor RNA. Introns Jones are removed & when Bartlett the exons Learning, are spliced together. LLC The mrna a covalently intact molecule (see the chapter has NOT only the FOR sequences SALE of the OR exons. DISTRIBUTION titled RNA Splicing and Processing). The original eukaryotic gene comprises the region in the genome between points Genomic DNA corresponding to the 59 and 39 terminal bases Exon 1 Intron 1 Exon 2 Intron 2 Exon 3 of mature RNA. We know Jones that & transcription Bartlett Learning, A LLC B Jones & Bartlett C Learning, LL starts at the DNA template NOT FOR corresponding SALE OR to DISTRIBUTION the 59 end of the mrna and usually extends A B B C beyond the complement to the 39 end of the mrna Exon 1 Exon 2 Exon 3 mature RNA, which is generated by cleavage A B C of the 39 extension. The gene is also considered to include Jones the regulatory & Bartlett regions Learning, on both sides LLC A B Jones B C & Bartlett Learning, LLC of the NOT gene FOR that are SALE required OR for DISTRIBUTION the initiation FIGURE 4.2 Exons remain NOT in the FOR same SALE order in OR mrna DISTRIBUTION as in DNA, but and (sometimes) termination of transcription. distances along the gene do not correspond to distances along the mrna or polypeptide products. The distance from A B in the gene is smaller than the distance from B C, but the distance from A B in the mrna (and polypeptide) is greater than the distance from B C. 4.2 An Interrupted Gene Consists of Exons EXON 1 EXON 2 EXON 3 INTRON 1 INTRON 2 INTRON 1 INTRON 2 (as NOT determined FOR by SALE recombination OR DISTRIBUTION analysis) do and Introns not correspond to the distances between sites in the processed mrna. The length of a gene is Key concepts defined by the length of the primary mrna transcript instead of the length of the mature mrna. Introns are removed by RNA splicing, which occurs in cis in individual RNA molecules. Jones & Bartlett Learning, All exons LLC of a gene are on one RNA molecule, Jones & Bartlett Learning, LL Mutations in exons can affect polypeptide and their splicing together is an intramolecular sequence; mutations in NOT introns FOR can affect SALE RNA OR DISTRIBUTION processing and hence may influence the sequence reaction. There is usually no joining of exons and/or production of a polypeptide. carried by different RNA molecules, so there is rarely cross-splicing of sequences. (However, How does the existence of introns change our in a process known as trans-splicing, sequences view of Jones the gene? & Bartlett During splicing, Learning, the exons LLC from different mrnas Jones are ligated & Bartlett together into Learning, LLC are always NOT FOR joined SALE together OR in DISTRIBUTION the same order a single molecule for NOT translation.) FOR SALE OR DISTRIBUTION they are found in the original DNA, so the correspondence between the gene and polypeptide of a polypeptide must occur in exons. What Mutations that directly affect the sequence sequences is maintained. FIGURE 4.2 shows that are the effects of mutations in the introns? the order of exons in a gene remains the same The introns are not part of the mature mrna, Jones & as Bartlett the order of Learning, exons the LLC processed mrna, so mutations Jones in & Bartlett them cannot Learning, directly affect LLC NOT FOR but SALE the distances OR DISTRIBUTION between sites in the gene the NOT polypeptide FOR sequence. SALE OR However, DISTRIBUTION they may An Interrupted Gene Consists of Exons and Introns 83

4 affect the processing of the mrna production equal amounts of A and T, and equal amounts Jones & Bartlett Learning, by inhibiting LLC Jones & Bartlett of C and G, Learning, in each single LLC the splicing of exons. A mutation strand of the duplex. of this sort acts only on the allele that NOT carries FOR it. SALE Like the OR first DISTRIBUTION parity rule, this extends to oligonucleotide sequences: For example, in a very Mutations that affect splicing are usually deleterious. The majority are single-base substitutions at the junctions between introns and bers of AC and TG dinucleotides. The reasons long strand there are approximately equal num- exons. They may cause an exon to be left out of for the existence of this rule are not clear, but sequencing of many genomes Jones has & Bartlett shown it to Learning, be LL the product, cause an intron to be included, or make NOT splicing FOR occur SALE at a different OR DISTRIBUTION site. The most nearly universally NOT true. The FOR second SALE parity OR rule DISTRIBUT common outcome is a termination codon that applies more closely to introns than to exons, shortens the polypeptide sequence. Thus, intron partly due to a further rule that purines tend mutations may affect not only the production to cluster on one DNA strand and pyrimidines tend to cluster on the other. This cluster rule as Jones & of Bartlett a polypeptide, Learning, but also LLC its sequence. About applied Jones to exons & Bartlett is that the Learning, purines, A LLC 15% of the point mutations that cause human and G, tended NOT to FOR be clustered SALE in OR one DISTRIBUTION diseases disrupt splicing. DNA strand of the DNA duplex (usually the nontemplate strand) Some eukaryotic genes are not interrupted and these are complemented by clusters of the and, like prokaryotic genes, correspond directly pyrimidines, T and C, in the template strand. with the polypeptide product. In the yeast The fact that in single-stranded DNA an Jones & Bartlett Learning, Saccharomyces LLC cerevisiae, most genes are Jones uninterrupted. In multicellular eukaryotes & Bartlett oligonucleotide Learning, is accompanied LLC in series by NOT most FOR SALE equal quantities OR DISTRIBUTION of its reverse complementary genes are interrupted, and the introns are usually much longer than exons, so that genes are oligonucleotide suggests that duplex DNA has the potential to extrude folded stem-loop structures, the stems of which can display base parity considerably larger than their coding regions. and the loops of which can display some degree of base clustering. Indeed, Jones the potential & Bartlett for such Learning, LL 4.3 NOT Exon FOR and SALE Intron OR DISTRIBUTION Base secondary structure NOT is found FOR to SALE be greater OR DISTRIBUT in Compositions Differ introns than exons, especially in exons under positive selection pressure (see the section titled Key concepts Exon Sequences Under Positive Selection Vary but The four rules for DNA base composition are Introns Are Conserved later in this chapter). Jones & Bartlett the first and Learning, second parity LLC rules, the cluster rule, Finally, Jones there & Bartlett is the GC Learning, rule, which LLC is that and the GC rule. Exons and introns can be distinguished on the basis of all rules except the first. the NOT overall FOR proportion SALE OR of G1C DISTRIBUTION in a genome (GC content) tends to be a species-specific character The second parity rule suggests an extrusion of (although individual genes within that genome structured stem-loop segments from duplex DNA, tend to have distinctive values). The GC content tends to be greater in exons than in introns. which would be greater in introns. The rules relate to genomic characteristics, or Jones & Bartlett Learning, Chargaff s four rules are seen to relate to characters or pressures that are intrinsic to the genome, pressures, LLCthat constitute the genome phenotype. contributing to what was termed the genome In the 1940s Erwin Chargaff initiated studies of phenotype (see the section titled There Are Many DNA base composition that led to four rules, Forms of Information in DNA later in this chapter). beginning with the first parity rule for duplex DNA (see the chapter titled Genes Are DNA). This rule applies Jones to most & regions Bartlett of DNA, Learning, including LLC both 4.4 Organization of exons NOT and FOR introns. SALE Base A OR in one DISTRIBUTION strand of the duplex is matched by a complementary base (T) Interrupted Genes in the other strand, and base G in one strand of May Be Conserved the duplex is matched by a complementary base (C) in the other strand. By extension, the rule Key concepts Jones & applies Bartlett not only Learning, to single bases, LLC but also to dinucleotides, OR trinucleotides, DISTRIBUTION and oligonucleotides. NOT with their FOR RNA transcription SALE OR products DISTRIBUTION by either restric- Introns can be detected when genes are compared NOT FOR SALE Thus, GT pairs with its reverse complement AC, tion mapping, electron microscopy, or sequencing. and ATG pairs with its reverse complement CAT. The positions of introns are usually conserved In addition to the well known first parity rule, when homologous genes are compared between later work by Chargaff led him to propose a second parity LLC rule. The little-known second Jones parity & Bartlett Learning, LLC different organisms. The lengths of the corresponding introns may vary greatly. Jones & Bartlett Learning, rule is that, to a close approximation, there are Introns usually do not encode proteins. 84 chapter 4 The Interrupted Gene..

5 When a gene is uninterrupted, the restriction protists, and one metazoan (a sea anemone), Jones map & Bartlett of its DNA Learning, corresponds LLC with the map of and in chloroplast genes. Genes with introns NOT FOR its SALE mrna. OR When DISTRIBUTION a gene possesses an intron, have NOT been FOR found SALE in every OR DISTRIBUTION class of eukaryotes, the map at each end of the gene corresponds to the map at each end of the message sequence. Within the gene, however, the maps diverge archaea, bacteria, and bacteriophages, although they are extremely rare in prokaryotic genomes. because additional regions that are found in Some interrupted genes have only one the gene are not represented in the mature or a few introns. The globin genes provide a mrna. Each such region NOT corresponds FOR SALE to OR an DISTRIBUTION much studied example (see the section NOT titled FOR SALE OR DISTRIBUT intron. The example in FIGURE 4.3 compares the restriction maps of a b-globin gene and its mrna. There are two introns, each of which Members of a Gene Family Have a Common Organization later in this chapter). The two general classes of globin gene, a and b, share a common contains a series of restriction sites that are organization. They originated from an ancient absent from the cdna. The pattern of restriction sites NOT in FOR the exons SALE is the OR same DISTRIBUTION in both the paralogous genes, NOT or paralogs. FOR SALE The consis- OR DISTRIBUTION gene duplication event and are described as cdna and the gene. The finer comparison of the base sequences of a gene and its mrna permits precise identification of introns. An intron tent structure of mammalian globin genes is evident from the generic globin gene presented in FIGURE 4.4. usually has no open reading frame. An intact Introns are found at homologous positions (relative to the coding sequence) in all reading frame is created in an mrna sequence NOT FOR by SALE the removal OR DISTRIBUTION of the introns from the primary known NOT active FOR globin SALE genes, OR including DISTRIBUTION those of transcript. The structures of eukaryotic genes show extensive variation. Some genes are uninterrupted mammals, birds, and frogs. Although intron lengths vary, the first intron is always fairly short and the second is usually longer. Most and their sequences are colinear with of the variation in the lengths of different those of the corresponding mrnas. Most globin genes results from length variation multicellular eukaryotic NOT genes FOR are interrupted, SALE OR DISTRIBUTION in the second intron. For example, NOT the second FOR SALE OR DISTRIBUT but the introns vary enormously in both number and size. Genes encoding polypeptides, rrna, or intron in the mouse a-globin gene is only 150 bp of the total 850 bp of the gene, whereas the homologous intron in the mouse trna may all have introns. Introns also are major b-globin gene is 585 bp of the total found in mitochondrial genes of plants, fungi, 1382 bp. The difference in length of the genes is much greater than NOT that FOR of their SALE mrnas OR DISTRIBUTION (a-globin mrna bases; Genomic DNA Sites cleaved by restriction enzymes b-globin mrna bases). Exon 1 Intron 1 Exon 2 Intron 2 Exon 3 The example of dihydrofolate reductase (DHFR), a somewhat & Bartlett larger gene, Learning, is shown LLCin NOT FOR FIGURE SALE 4.5. OR The DISTRIBUTION mammalian Jones DHFR gene is organized into cdna six exons that correspond to a cdna map corresponds to 2000-base mrna. The gene itself exon1 + exon2 + exon3 of genomic map is long because the introns are FIGURE 4.3 Comparison of the restriction maps of cdna and genomic DNA very long. In three mammal Jones species the exons are essentially NOT FOR the SALE OR DISTRIBUT & Bartlett Learning, LL for mouse b-globin shows that NOT the gene FOR has SALE two introns OR that DISTRIBUTION are not present in the cdna. The exons can be aligned exactly between cdna and the gene. same and the relative positions of Intron length Exon length Exon 1 Intron 1 Exon 2 Intron 2 Exon Contains 5 UTR + coding 1 30 Amino acids Coding 105 end + 3 UTR FIGURE 4.4 All functional globin genes have an interrupted structure with three exons. The lengths indicated in the figure apply to the mammalian b-globin genes Organization of Interrupted Genes May Be Conserved 85

6 5 globin Major 5 Exon Exon Exon Exon 1 NOT 1 2 FOR 3 SALE 4 5 OR DISTRIBUTION 6 Exons kb FIGURE 4.5 Mammalian genes for DHFR have the same relative organization of rather short exons and very long introns, but vary extensively in the lengths of introns. globin Minor Exon 2 Exon 3 the introns are unaltered, but the lengths of 3 individual introns vary extensively, resulting FIGURE 4.6 The sequences of the mouse b maj - and b min - Jones & Bartlett Learning, a variation LLC in the length of the gene Jones from & Bartlett Learning, LLC globin genes are closely related in coding regions but differ 25 to 31 kb. NOT FOR SALE in the flanking OR DISTRIBUTION UTRs and the long intron. Data provided by The globin and DHFR genes are examples Philip Leder, Harvard Medical School. of a general phenomenon: genes that share a common ancestry have similar organizations with conservation of the positions (of at least some) of both copies, with substitutions restricted the introns. the exons by the need to encode a functional polypeptide. As we will see in the chapter titled Genome 4.5 Exon Sequences Under Evolution, where we consider the evolution of Negative Selection the genome, exons can be considered basic building blocks that may be assembled in various combinations. It is possible for a gene to Jones & Bartlett Are Learning, Conserved LLCbut have NOT some FOR exons SALE related OR to DISTRIBUTION Introns Vary those of another gene, with the remaining exons unrelated. Key concepts Usually, in such cases, the introns are not Comparisons of related genes in different species related at all. Such homologies between genes show that the sequences of the corresponding may result from duplication and translocation Jones & Bartlett Learning, exons are LLC usually conserved but the sequences Jones of & Bartlett of individual Learning, exons. LLC the introns are much less similar. NOT FOR SALE The OR homology DISTRIBUTION between two genes can be Introns evolve much more rapidly than exons plotted in the form of a dot matrix comparison, as in FIGURE 4.6. A dot is placed in each because of the lack of selective pressure to produce a polypeptide with a useful sequence. position that is identical in both genes. The Is a single-copy structural gene completely dots form a solid line on the diagonal of the unique Jones among & other Bartlett genes Learning, in its genome? LLC matrix if the two sequences Jones & are Bartlett completely Learning, LL The NOT answer FOR depends SALE on OR how DISTRIBUTION completely identical. If they NOT are not FOR identical, SALE the OR line DISTRIBUT unique is defined. Considered as a whole, the is broken by gaps that lack homology and gene is unique, but its exons may be related is displaced laterally or vertically by nucleotide deletions or insertions in one or the to those of other genes. As a general rule, when two genes are related, the relationship other sequence. between their exons is closer than the relationship OR between DISTRIBUTION their introns. In an extreme compared NOT FOR in SALE this way, OR a DISTRIBUTION line of homology When the two mouse b-globin genes are NOT FOR SALE case, the exons of two genes may encode the extends through the three exons and the small same polypeptide sequence while the introns intron. The line disappears in the flanking UTRs are different. This situation can result from and in the large intron. This is a typical pattern the duplication of a common ancestral gene in related genes; the coding sequences and f ollowed by unique base substitutions in areas of introns adjacent to exons retain their 86 chapter 4 The Interrupted Gene..

7 similarity, but there is greater divergence in longer & Bartlett introns and Learning, in the regions LLCon either side of Jones NOT FOR the SALE coding OR sequence. DISTRIBUTION The overall degree of divergence between two homologous exons in related genes corresponds to the differences between the polypeptides. It is mostly a result of base substitutions. In the translated regions, changes in exon sequences are NOT constrained FOR SALE by selection against mutations that alter or destroy the OR DISTRIBUTION function of the polypeptide. In other words, the exon sequences are conserved by the negative selection of individuals in which the sequences have changed (have not been conserved) to result NOT in a phenotype FOR SALE that OR is less DISTRIBUTION able to survive and produce fertile progeny. Many of the preserved changes do not affect codon meanings because they change a codon into another for the same amino acid (i.e., they are synonymous substitutions). In this case, NOT FOR the SALE polypeptide OR DISTRIBUTION will not change and negative selection will not operate on the phenotype conferred by the polypeptide. Similarly, there are higher rates of change in untranslated 4.6 Jones Exon & Sequences Bartlett Learning, Under LLC Positive Selection Vary but Introns Are Conserved Key concepts Under positive selection an individual with Jones an & Bartlett Learning, LL advantageous mutation survives (i.e., is able to produce more fertile progeny) relative to NOT others FOR SALE OR DISTRIBUT without the mutation. Due to intrinsic genomic pressures, such as that which conserves the potential to extrude stem-loops from duplex DNA, introns evolve more slowly than exons that are under positive Jones selection & Bartlett pressure. Learning, LLC A mutation that confers a more advantageous phenotype to an organism, relative to individuals in the same population without the mutation, may result in the preferential survival (positive Jones selection) & of Bartlett that organism. Learning, Pathogenic LLC bacteria NOT are FOR killed SALE by an OR antibiotic, DISTRIBUTION but a bacterium with a mutation that confers antibiotic resistance survives (i.e., is positively selected). Mutations conferring venom- resistance to prey of venomous snakes can result in the positive selection of that prey relative to its fellows Jones that & Bartlett Learning, LL regions of the gene (specifically, those that are transcribed to the 59 UTR [leader] and 39 UTR succumb to the poison (i.e., are negatively [trailer] of the mrna). selected). Likewise, a snake that, when confronted by a venom-resistant prey population, In homologous introns, the pattern of divergence involves both changes in length has a mutation that enhances the power of its (due to deletions and insertions) and base venom, will be positively selected. This can trig- substitutions. Introns evolve much more rapidly than exons when the exons are under ger an attack defense Jones cycle an & Bartlett arm s race Learning, LLC negative NOT selection FOR SALE pressure. OR When DISTRIBUTION a gene is between two protagonist NOT species. FOR SALE OR DISTRIBUTION compared among different species, there are In such situations the pattern of exon conservation and intron variation seen in genes instances where its exons are homologous but its introns have diverged so much that very lit- under negative selection can be reversed because exons evolve faster than introns. Thus, a plot tle homology is retained. Although mutations Jones & in certain Bartlett intron Learning, sequences LLC (branch site, splicing SALE junctions, OR DISTRIBUTION and perhaps other sequences NOT FOR influencing splicing) will be subject to selection, most intron mutations are expected to be selectively neutral. similar Jones to Figure & 4.6 Bartlett will have Learning, lines introns LLC and NOT gaps in FOR exons. SALE Another OR way DISTRIBUTION of showing this is to plot base substitutions along the length of a gene. FIGURE 4.7 shows a plot of the substitutions observed when two snake venom alkaline phosphatase genes are compared. The protein- In general, mutations occur at the same rate in both exons and introns, Jones but exon & Bartlett mutations Learning, encoding LLC parts of exons (2, 3, and the first Jones half of & Bartlett Learning, LL are eliminated more NOT effectively FOR by SALE selection. OR DISTRIBUTION exon 4) have many base substitutions NOT (i.e., they FOR SALE OR DISTRIBUT However, because of the low level of functional are varying), whereas the three introns have constraints, introns may more freely accumulate point substitutions and other changes. What is being conserved in introns? First, relatively few (i.e., they are conserved). Indeed, it is sometimes possible to locate exons intron sequences needed for RNA splicing the 59 and 39 splice sites and Jones the branch & Bartlett site are Learning, LLC in uncharted sequences by virtue of their conservation NOT relative FOR SALE to introns OR (see DISTRIBUTION the chapter conserved (see the NOT chapter FOR titled SALE RNA Splicing OR DISTRIBUTION The Content of the Genome). From this description and Processing). In addition to these, base order it is all too easy to conclude that introns do not has been adapted to promote the potential of have a sequence-specific function. Genes under the duplex DNA in the region to extrude stem positive selection, however, cast a different light loop structures (fold potential). Thus, a plot of on the problem. base order dependent fold potential along the 4.6 Exon Sequences Under Positive Selection Vary but Introns Are Conserved 87..

8 Fold Potential UTR Substitutions Jones & 5 Bartlett Learning, LLC Fold Potential Jones & Bartlett Learning, 10 LLC Length of Sequence (bases) FIGURE 4.7 The sequences of snake venom phospholipase genes differ in coding regions, but are closely related in introns and flanking regions. Fold potential (here the contribution of base order to the potential to extrude stem loop structures) is low (more positive) in the protein-encoding exons and high (more negative) in introns. Jones The positions & Bartlett of the four Learning, exons are shown LLCas numbered boxes. Modified from D. R. Forsdyke, Conservation of Stem-Loop Potential in Introns of Snake Venom Phospholipase A2 Genes: NOT An Application FOR of SALE FORS-D OR Analysis, DISTRIBUTION Mol. Biol. Evol., vol. 12(6), pp , by permission of Oxford University Press. 3 UTR Substitutions 20 length Jones of the gene & Bartlett shows that Learning, fold potential LLC (measured NOT FOR in negative SALE OR units) DISTRIBUTION is high (more negative) in introns, and low (more positive) in exons (Figure 4.7). This reciprocal relationship between substitution frequency and the contribution of base order to fold potential is a Jones & characteristic Bartlett Learning, of DNA sequences LLC under positive selection. Indeed, the Jones low (more & positive) Bartlett value Learning, LL of fold potential in NOT an exon FOR provides SALE evaluation OR DISTRIBUT of the extent to which it has been under positive selection, without the need to compare two sequences (the classical way of determining if selection is positive or negative). 4.7 Genes Show a Wide Distribution of Sizes Due Primarily to Intron Size S. cerevisiae 80 Jones & Bartlett and Learning, Number LLCVariation 60 Key concepts 40 Most genes are uninterrupted in Saccharomyces 20 cerevisiae but are interrupted in multicellular eukaryotes. 30 Jones D. melanogaster & Bartlett Learning, LLC Exons are usually short, Jones typically & encoding Bartlett fewer Learning, LL than 100 amino acids % Introns are short in unicellular/oligocellular 10 eukaryotes but can be many kb in multicellular eukaryotes. 15 Mammals The overall length of a gene is determined largely by its introns. 10 6% 5 FIGURE NOT 4.8 FOR compares SALE the OR organization DISTRIBUTION of genes in a yeast, an insect, and mammals. In the yeast <40 >60 Saccharomyces cerevisiae, the majority of genes Number of exons (.96%) are uninterrupted, and those that have FIGURE 4.8 Most genes are uninterrupted in yeast, but most genes are interrupted exons generally have three or fewer. There are Jones in flies & and Bartlett mammals. Learning, (Uninterrupted LLC genes have only one exon and are totaled Jones in & Bartlett virtually no Learning, S. cerevisiae LLC genes with more than NOT FOR the leftmost SALE column OR in DISTRIBUTION red.) NOT FOR SALE four exons. OR DISTRIBUTION Percent 96% 88 chapter 4 The Interrupted Gene..

9 In insects and mammals the situation is 6 Jones reversed. & Bartlett Only Learning, a few genes LLC have uninterrupted 5 NOT FOR coding SALE sequences OR DISTRIBUTION (6% in mammals). Insect 4 genes tend to have a small number of exons, Yeast 3 typically fewer than 10. Mammalian genes are 2 split into more pieces and some have more 1 than 60 exons. Approximately 50% of mammalian genes have more than 10 introns. If we Jones & Bartlett Learning, 6 LLC examine the effect of NOT intron FOR number SALE variation OR DISTRIBUTION 5 on the total size of genes, we see in FIGURE Fly that there is a striking difference between yeast 3 and multicellular eukaryotes. The average yeast 2 gene is 1.4 kb long, and very few are longer than 1 5 kb. The predominance of interrupted genes in multicellular NOT FOR eukaryotes, SALE however, OR DISTRIBUTION 6 means that 5 the gene can be much larger than the sum total of the exon lengths. Only a small percentage of 4 Human genes in flies or mammals are shorter than 2 3 kb, and most have lengths between 5 kb and kb. The average human gene is 27 kb long. 1 NOT FOR The SALE dystrophin OR DISTRIBUTION gene, with a length of 2000 kb, is the longest known human gene. Exon length in nucleotides The switch from largely uninterrupted to largely interrupted genes seems to have occurred with the evolution of multicellular eukaryotes. In fungi other than S. cerevisiae, the majority of genes NOT are interrupted, FOR SALE but they OR DISTRIBUTION have a relatively small number of exons (,6) and are fairly short (,5 kb). In the fruit fly, gene sizes have a bimodal distribution many FIGURE 4.10 Exons encoding polypeptides are usually short. FIGURE 4.10 shows that exons encoding stretches of protein tend to be fairly small. In multicellular eukaryotes, the average exon codes for 50 amino acids, and the general distribution is consistent with Jones the & hypothesis Bartlett that Learning, LLC Percent S. cerevisiae are short but some are quite long. With this increase in the length of the gene due to the increased NOT number FOR SALE of introns, OR DISTRIBUTION the correlation genes have evolved NOT by the FOR gradual SALE addition OR of DISTRIBUTION between genome size and organism complexity exon units that encode short, functionally independent protein domains (see the Genome Evo- becomes weak. lution chapter). There is no significant difference in the average size of exons in different multicellular Jones eukaryotes, & Bartlett although Learning, the size range LLC is smaller NOT FOR in vertebrates SALE OR for DISTRIBUTION which there are few exons longer than 200 bp. In yeast, there are some longer exons that represent uninterrupted genes for which the coding sequence is intact. There is a tendency for exons containing Jones D. melanogaster & Bartlett Learning, untranslated LLC59 and 39 regions to be longer Jones than & Bartlett Learning, LL those that encode proteins. FIGURE 4.11 shows that introns vary widely in size among multicellular eukaryotes. (Note that the scale of the x-axis differs from that of Mammals Figure 4.10.) In worms and flies, the average intron is not longer than Jones the exons. & Bartlett There are Learning, LLC no very long introns NOT in worms, FOR SALE but flies OR contain many. In vertebrates, the size distribution DISTRIBUTION < 0.5 < 1 < 2 < 5 < 10 < 25 < 50 < 100 > 100 is much wider, extending from approximately Size of gene (kb) the same length as the exons (,200 bp) up to 60 kb in extreme cases. (Some fish, such as fugu, FIGURE 4.9 Yeast genes are short, but genes in flies and mammals have a dispersed bimodal distribution extending to very long sizes. % exons have compressed Jones & genomes Bartlett with Learning, shorter introns LLC and NOT intergenic FOR regions SALE than OR mammals DISTRIBUTION have.) 4.7 Genes Show a Wide Distribution of Sizes Due Primarily to Intron Size and Number Variation 89..

10 20 Very long genes are the result of very long introns, not the result of encoding longer products. OR DISTRIBUTION There is no correlation between 15 NOT FOR SALE Worm total gene size and total exon size in multicellular eukaryotes, nor is there a good correlation 10 between gene size and number of exons. The 5 size of a gene is therefore determined prima rily by the lengths of its individual introns. In NOT 20 FOR SALE OR DISTRIBUTION mammals and insects, NOT the FOR average SALE gene OR DISTRIBUT is 15 approximately 53 that of the total length of Fly its exons. 10 Jones & Bartlett 5 Learning, LLC 4.8 Some DNA Sequences 20 % introns Human Intron length in kb FIGURE 4.11 Introns range from very short to very long. Full-length protein Part-length protein START Triplet codons Alternative START FIGURE 4.12 Two proteins can be generated from a single gene by starting (or terminating) expression at different points. Bases START Codons used for protein 1 Encode More Than One Polypeptide Key concepts The use of alternative initiation or termination codons allows multiple variants of a polypeptide chain. Different polypeptides can be produced from the same sequence of DNA when the mrna is read in different reading frames (as two overlapping genes). Otherwise identical NOT polypeptides, FOR differing SALE by OR the DISTRIBUT presence or absence of certain regions, can be generated by differential (alternative) splicing when certain exons are included or excluded. This may take the form of including or excluding individual exons, or of choosing between alternative exons. Jones & Bartlett Learning, LLC Many structural genes consist of a sequence that encodes a single polypeptide, although the gene may include noncoding regions at both ends and introns within the coding region. However, there are some cases in which a single sequence of DNA encodes more than one polypeptide. In one simple example, a single DNA STOP sequence may have two alternative start codons in the same reading frame (see FIGURE 4.12). Thus, under different conditions one or the other of the start codons may be used, allowing the production of either NOT a short FOR form SALE of the OR polypeptide or a full-length form, where the short DISTRIBUT form is the last portion of the full-length form. An actual overlapping gene occurs when the same sequence of DNA encodes two Jones nonhomologous & Bartlett proteins Learning, because LLC it uses more NOT than FOR one SALE reading OR frame. DISTRIBUTION Usually, a coding DNA sequence is read in only one of the three potential reading frames. In some viral and mitochondrial genes, however, there is some overlap between two adjacent genes that Codons used for protein 2 START Jones & FIGURE Bartlett 4.13 Learning, Two genes may LLC overlap by reading the same DNA sequence Jones in & Bartlett are read Learning, different reading LLC frames, as illustrated OR in FIGURE DISTRIBUTION The length of overlap NOT FOR SALE different frames. OR DISTRIBUTION NOT FOR SALE is 90 chapter 4 The Interrupted Gene..

11 W X Z One mrna has the exon Exons W X Z Exon 1 Intron Exon 2 Intron Exon 3 RNA synthesis One mrna has the exon W X Z FIGURE 4.14 Alternative splicing generates the a and b variants of troponin T. 5 5 Only introns spliced out 3 mrna 3 C usually short, Jones so that & Bartlett most of the Learning, DNA sequence LLC encodes NOT a unique FOR SALE polypeptide OR sequence. DISTRIBUTION In some cases, genes can be nested. This N Long protein has all 3 exons Introns + exon 2 spliced out occurs when a complete gene is found within the intron of a larger host gene. Nested genes 5 3 often lie on the strand opposite to that of the Jones & host Bartlett gene. Learning, LLC 5 3 mrna NOT FOR SALE In some OR genes DISTRIBUTION there are switches in the pathway for splicing the exons that result in alternative patterns of gene expression. A single gene may generate a variety of mrna products N C that differ in their exon content. Certain exons Short protein has only exon 1 + exon 3 may be optional; in other Jones words, & they Bartlett may be Learning, FIGURE 4.15 LLCAlternative splicing uses the same Jones pre-mrna & to Bartlett generate mrnas Learning, LL included or spliced out. NOT There FOR also may SALE be a OR pair DISTRIBUTION that have different combinations of exons. of exons treated as mutually exclusive one or the other is included in the mature transcript, but not both. The alternative proteins have one primary transcript can be spliced in either of part in common and one unique part. two ways. In the first (more standard) pathway, two introns are spliced Jones out & and Bartlett the three Learning, LLC In some Jones cases, & Bartlett the alternative Learning, means LLC of expression NOT FOR do not SALE affect the OR sequence DISTRIBUTION of the exons are joined together. NOT FOR In the SALE second OR path-distributioway, polypeptide. For example, changes that affect the 59 UTR or the 39 UTR may have regulatory consequences, but the same polypeptide is made. In other cases, one exon is substituted for Jones & another, Bartlett as in Learning, FIGURE LLC In this example, the the second exon is excluded as if a single large intron is spliced out. This intron consists of intron 1 1 exon 2 1 intron 2. In effect, exon 2 has been treated in this pathway as if it were part of Jones a single & intron. Bartlett The Learning, pathways produce LLC NOT FOR polypeptides SALE OR produced DISTRIBUTION by the two mrnas contain sequences that overlap extensively, but are ends, but one has an additional sequence in the two NOT polypeptides FOR SALE that are OR the DISTRIBUTION same at their different within the alternatively spliced region. The 39 half of the troponin T gene of rat muscle contains five exons, but only four are used to middle. (Other types of combinations that are produced by alternative splicing are discussed in the RNA Splicing and Processing chapter). construct an individual mrna. Jones Three & Bartlett exons (W, Learning, Sometimes LLC two alternative splicing Jones pathways & Bartlett Learning, LL operate simultaneously, with NOT a certain FOR SALE OR DISTRIBUT X, and Z) are included NOT in all mrnas. FOR SALE However, OR DISTRIBUTION in one alternative splicing pattern the a exon proportion of the primary RNA transcripts is included between X and Z, whereas in the other pattern it is replaced by the b exon. The being spliced in each way. However, sometimes the pathways are alternatives that are expressed a and b forms of troponin T therefore differ in under different conditions, for example, one the sequence Jones of the & Bartlett amino acids Learning, between W LLC and in one cell type and one Jones in another & Bartlett cell type. Learning, LLC Z, depending NOT FOR on which SALE of the OR alternative DISTRIBUTION exons (a or b) is used. Either one of the a and b exons can be used in an individual mrna, but both cannot be used in the same mrna. So, alternative NOT (or FOR differential) SALE OR splicing can generate different polypeptides with DISTRIBUTION related sequences from a single stretch of DNA. It is curious that the multicellular eukaryotic FIGURE 4.15 shows that alternative splicing genome is often extremely large with long Jones & can Bartlett lead to the Learning, inclusion LLC of an exon in some genes Jones that are & often Bartlett widely Learning, dispersed along LLCa NOT FOR mrnas SALE while OR leaving DISTRIBUTION it out of others. A single chromosome, NOT FOR but SALE at the OR same DISTRIBUTION time there may 4.8 Some DNA Sequences Encode More Than One Polypeptide 91..

12 be multiple products from a single locus. Due we can equate the functional domains of current proteins with the individual exons of the Jones & Bartlett Learning, to alternative LLC splicing, there are,15% more polypeptides than genes in flies and worms, NOT FOR but SALE corresponding OR DISTRIBUTION genes, then this would suggest it is estimated that the majority of human genes selective interdomain interruptions rather than are alternatively spliced (see the chapter titled random ones. Genome Sequences and Gene Numbers). In some cases there is a clear relationship between the structures of a gene and its protein product, but these may be special cases. The 4.9 NOT Some FOR SALE Exons OR Can DISTRIBUTION Be example par excellence NOT is provided FOR SALE by the OR immunoglobulin (antibody) proteins an extracel- DISTRIBUT Equated with Protein lular system for self-/nonself- discrimination Functional Domains that aids in the elimination of foreign pathogens. Immunoglobulins are encoded by genes Jones & Key Bartlett conceptslearning, LLC in which every exon corresponds exactly to NOT FOR SALE Proteins OR can DISTRIBUTION consist of independent functional a NOT known FOR functional SALE protein OR DISTRIBUTION domain. Banks modules, the boundaries of which, in some cases, can be equated with those of exons. of alternate sequence domains are tapped so The exons of some genes appear homologous to that each cell acquires the ability to secrete a the exons of others, suggesting a common exon cell- specific immunoglobulin with distinctive ancestry. binding capacity for a foreign antigen that the organism may one day encounter again (see The issue of the evolution of interrupted NOT genes FOR SALE the chapter OR DISTRIBUTION titled Somatic Recombination and is more fully considered in the Genome Evolution Hypermutation in the Immune System). FIGURE 4.16 chapter. If proteins evolve by recombining compares the structure of an immunoglobulin parts of ancestral proteins that were originally with its gene. separate, the accumulation of protein domains An immunoglobulin is a tetramer of two is likely Jones to have & occurred Bartlett sequentially, Learning, LLC with light chains and two heavy chains that covalently bond to generate NOT a FOR protein SALE with several OR DISTRIBUT one NOT exon added FOR at SALE a time. OR Each DISTRIBUTION addition would have to improve upon the advantages of prior distinct domains. Light chains and heavy chains additions in a sequence of positive selection differ in structure, and there are several types events. Are the different function-encoding of heavy chains. Each type of chain is produced from a gene that has a series of exons segments from which these genes may have originally been pieced together reflected in corresponding to the structural domains of NOT FOR their SALE present OR DISTRIBUTION structures? If a protein sequence the NOT protein. FOR SALE OR DISTRIBUTION were randomly interrupted, sometimes the In many instances, some of the exons of interruption would intersect a domain and a gene can be identified with particular functions. In secretory proteins, such as sometimes it would lie between domains. If insulin, L exon V-J exon C exon Light chain Leader Variable Constant Hinge Constant 2 Constant 3 Heavy chain Protein domains L exon V-D-J exon C1 exon Hinge exon C2 exon C3 exon FIGURE 4.16 Immunoglobulin light chains and heavy chains are encoded by genes whose structures (in their expressed forms) correspond to the Jones distinct & domains Bartlett in the Learning, protein. Each protein LLC domain corresponds to an exon; introns are numbered I1 to I5. 92 chapter 4 The Interrupted Gene..