Intronless human dihydrofolate reductase genes are derived from processed RNA molecules.

Three groups of recombinant bacteriophage containing coding sequences for dihydrofolate reductase (DHFR; tetrahydrofolate dehydrogenase; 5,6,7,8-tetrahydrofolate:NADP+ oxidoreductase, EC 1.5.1.3) were isolated from two human DNA clone libraries. One recombinant (lambda hDHFR-1) contains three exons that encode the COOH-terminal portion of human DHFR. The other two human DHFR genes (hDHFR-psi 1 and hDHRF-psi 2) lack introns. hDHFR-psi 2 contains several in-phase termination codons and is only 93% homologous to the normal human DHFR coding sequences, whereas hDHFR-psi 1 has an open reading frame and is virtually identical to the coding sequence of the normal DHFR gene. The region of DNA sequence homology between each intronless gene and the normal DHFR gene extends 2.9 kilobases beyond the end of the coding sequences. At the 3' end of this homologous sequence, each intronless gene has an A-rich tract. The lack of introns and the presence of the 3' A-rich tract suggest that hDHFR-psi 1 and hDHFR-psi 2 were derived from processed RNA molecules. A short DNA sequence, 60 nucleotides 5' to the ATG start codon in lambda hDHFR-psi 2, is directly repeated immediately after the 3' A-rich tract; such terminal direct repeats also flank integrated proretroviruses and transposable DNA elements and are thought to be the hallmark of inserted DNA sequences.     

A strategy to detect and isolate an intron-containing gene in the presence of multiple processed pseudogenes.

We have devised a strategy that utilizes the polymerase chain reaction (PCR) for the detection and isolation of intron-containing genes in the presence of an abundance of processed pseudogenes. The method depends on the genomic DNA sequence between the PCR primers spanning at least one intron in the gene of interest, resulting in the generation of a larger intron-containing PCR product in addition to the smaller PCR product amplified from the intronless pseudogenes. A unique intron probe isolated from the larger PCR product is used for the detection of intron-containing clones from recombinant DNA libraries that also contain pseudogene clones. This method has been used successfully for the selective isolation of an intron-containing rat L19 ribosomal protein gene in the presence of multiple pseudogenes. Analysis of a number of mammalian ribosomal protein multigene families by PCR indicates that they all contain only a single gene with introns.     

Human Cu/Zn superoxide dismutase gene family: molecular structure and characterization of four Cu/Zn superoxide dismutase-related pseudogenes.

Four distinct human Cu/Zn superoxide dismutase (SOD1; EC 1.15.1.1)-related sequences were isolated from genomic DNA libraries. Genomic blots, heteroduplex analyses, and DNA sequencing showed that they are processed pseudogenes not residing on chromosome 21. Three of them originated from the 0.7-kilobase SOD1 mRNA, while the fourth was derived from the 0.9-kilobase mRNA species. Comparison between the coding sequences of the functional gene and two of the processed genes suggested that they integrated into the genome about 25 million years ago.     

Cloning and structural analysis of two distinct families of ovine interferon-alpha genes encoding functional class II and trophoblast (oTP) alpha-interferons

Ovine trophoblast protein (oTP) is a polypeptide secreted by ovine trophectoderm from day 11 to 21, which plays a key role in maternal recognition of pregnancy. Structural analyses established that oTP shares extensive homology with class II alpha-interferon (IFN-alpha II) subfamily. Previous screening of an ovine genomic DNA library probed with an oTP cDNA incidently resulted in the isolation of a functional IFN-alpha II gene and two relevant pseudogenes, as shown by sequence analysis and study of expression in eukaryotic COS cells. The expected oTP gene together with a cognate pseudogene was successfully isolated from the series of clones selected from another genomic library probed with the oTP cDNA, using two specific oligonucleotides, each one complementary to a region of oTP cDNA with little homology with the IFN-alpha II gene and related pseudogenes. Southern blotting of ovine genomic DNA indicated the existence of at least five trophoblast IFN-alpha genes or pseudogenes. Nucleotide sequence comparisons showed that the oTP gene exhibits a higher homology (90%) with bovine trophoblast IFN gene (Stewart et al. (1990) J. Mol. Endocrinol. 4, 275-282) than with oIFN-alpha II gene (70%), thus providing evidence that embryonic IFNs constitute a distinct subfamily of IFN-alpha s.     

Glucocerebrosidase recombinant allele: molecular evolution of the glucocerebrosidase gene and pseudogene in primates

Glucocerebrosidase is a lysosomal enzyme that hydrolyses the beta-glycosidic linkage of glucocerebroside, a ubiquitous sphingolipid present in the plasma membrane of mammalian cells. Deleterious mutations in the glucocerebrosidase gene result in Gaucher disease, the most prevalent lysosomal storage disease. Humans have one glucocerebrosidase functional gene and pseudogene that were located 16 kb apart on chromosome 1q21 and share 96% overall sequence similarity. Recombination between the two genes creates a 'complex allele' that renders glucocerebrosidase non-functional and accounts for >20% of the total Gaucher disease mutations in some population. The glucocerebrosidase pseudogene is absent in all other mammalian species surveyed so far. In order to learn more about the molecular evolution of the glucocerebrosidase functional gene and pseudogene, we have sequenced approximately 1.1 kb of the C-terminal region of these genes that encodes the enzyme catalytic site, from PCR-amplified genomic DNA of gorilla, chimpanzee, orangutan (the great apes), and squirrel monkey (a new-world monkey). In gorilla, chimpanzee, and orangutan, there are two copies of the glucocerebrosidase gene while the squirrel monkey possesses only a single copy. Similar to human, the second copy of glucocerebrosidase gene in gorilla and chimpanzee is non-functional because of a 55-bp deletion in exon 9, while that in orangutan appears to be unaffected and may still be functional. These data suggest that the glucocerebrosidase gene duplication event occurred after squirrel monkey divergence from the great apes, and that the exon 9 deletion that rendered the second copy of the glucocerebrosidase gene non-functional occurred prior to the divergence of gorilla and chimpanzee but after the divergence of orangutan from their common ancestor to human. The two genes in each species are least similar in gorilla and chimpanzee (97.8%) and most similar in orangutan (99.5%). None of the nucleotide variations in the GBA gene among the primates correspond to known mutations in Gaucher disease. Phylogenetic tree analysis using DNAstar and PAUP4.0 software indicates that gene conversion caused the evolution of glucocerebrosidase functional gene and pseudogene to be concerted.     

Rapid evolution of the human gene for cytochrome c oxidase subunit IV.

We have compared the DNA sequences of nine mammalian genes for cytochrome c oxidase subunit IV (COX4 genes)--four expressed genes (human, bovine, rat, and mouse) and five pseudogenes (human, chimpanzee, orangutan, squirrel monkey, and bovine)--and constructed the sequence of the ancestral mammalian COX4 gene. By analyzing these sequences to determine the pattern and rate of nucleotide substitution in each branch of the evolutionary tree, we deduced that the human gene has evolved rapidly since the origin of the primate pseudogene approximately 41 million years ago, and we discuss the suggestion that this results from coevolution of nuclear and mitochondrial genes for cytochrome c oxidase.     

The nucleotide sequence of chromosome I from Saccharomyces cerevisiae.

Chromosome I from the yeast Saccharomyces cerevisiae contains a DNA molecule of approximately 231 kbp and is the smallest naturally occurring functional eukaryotic nuclear chromosome so far characterized. The nucleotide sequence of this chromosome has been determined as part of an international collaboration to sequence the entire yeast genome. The chromosome contains 89 open reading frames and 4 tRNA genes. The central 165 kbp of the chromosome resembles other large sequenced regions of the yeast genome in both its high density and distribution of genes. In contrast, the remaining sequences flanking this DNA that comprise the two ends of the chromosome and make up more than 25% of the DNA molecule have a much lower gene density, are largely not transcribed, contain no genes essential for vegetative growth, and contain several apparent pseudogenes and a 15-kbp redundant sequence. These terminally repetitive regions consist of a telomeric repeat called W', flanked by DNA closely related to the yeast FLO1 gene. The low gene density, presence of pseudogenes, and lack of expression are consistent with the idea that these terminal regions represent the yeast equivalent of heterochromatin. The occurrence of such a high proportion of DNA with so little information suggests that its presence gives this chromosome the critical length required for proper function.     

Cloning of human immunoglobulin epsilon chain genes: evidence for multiple C epsilon genes.

An active human epsilon chain gene was cloned from a phage library containing partial EcoRI digests of IgE-producing myeloma DNA, using the human JH (joining) gene fragment as a probe. The epsilon chain gene clone was identified by partial nucleotide sequence determination. The germ-line constant region gene of the epsilon chain (C epsilon gene) was cloned from a human fetal liver DNA library, using the cloned epsilon chain gene as a probe. Comparative studies on the human and mouse germ-line epsilon chain genes revealed that the switch (S) sequence is more conserved than the coding sequence. Restriction endonuclease BamHI digestion of human DNA produced three C epsilon fragments of 3.0, 6.5, and 9.2 kilobases, which were named C epsilon 1, C epsilon 2, and C epsilon 3 genes, respectively. We found the three C epsilon gene fragments in all of the human DNA preparations from eleven individuals. The C epsilon gene expressed in the myeloma was identified as the C epsilon 1 gene. Because the C epsilon 2 gene is deleted from the myeloma DNA, the order of the C epsilon genes is likely to be 5'-C epsilon 2-C epsilon 1-C epsilon 3-3', assuming that all the C epsilon genes are on chromosome 14. The germ-line C epsilon 3 gene was also cloned from the myeloma DNA. Characterization of the C epsilon 3 gene revealed that it does not have the S region, suggesting that it might be a pseudogene.     

Identification of a novel NCF-1 (p47-phox) pseudogene not containing the signature GT deletion: significance for A47 degrees chronic granulomatous disease carrier detection

The p47-phox gene, NCF-1, has 2 nearly identical pseudogenes (psiNCF-1) in proximity at chromosomal locus 7q11.23. A dinucleotide deletion (DeltaGT) at the beginning of exon 2 that leads to a frameshift and premature stop codon is considered the signature sequence of the pseudogenes. It is also the most prevalent mutation in p47-phox-deficient (A47 degrees ) chronic granulomatous disease (CGD) as a result of the insertion of a DeltaGT-containing fragment of pseudogene into NCF-1. Extending our study of the relationship between NCF-1 and psiNCF-1 to 53 unaffected control individuals, we found that although in most (n = 44), the ratio of pseudogene (DeltaGT) to functional gene (GTGT) sequence in amplicons spanning exon 2 was 2:1, as previously observed, surprisingly, in 7 persons the ratio was 1:1, and in 2 persons the ratio was 1:2. The lowered ratios are explained by the presence, in a heterozygous or homozygous state, respectively, of a pseudogene that contains GTGT rather than DeltaGT. It is possible that this pseudogene has not undergone deletion of GT, but more likely, based on analysis of additional NCF-1/psiNCF-1 markers, it represents the previously unidentified product of the reciprocal crossover of DNA fragments between the functional gene and one of its pseudogenes. The mutated NCF-1 resulting from this event is the predominant A47 degrees CGD allele. The existence of 2 extended haplotypes encompassing NCF-1/psiNCF-1 further complicates the detection of A47 degrees CGD carriers. Although most have a DeltaGT/GTGT ratio of 5:1, some have a ratio of 2:1 and are indistinguishable by this means from unaffected individuals.     

Extraordinarily high evolutionary rate of pseudogenes: evidence for the presence of selective pressure against changes between synonymous codons.

Comparisons of nucleotide sequences of several pseudogenes described to date, including alpha- and beta-globin and immunoglobulin kappa-type variable domain pseudogenes, with those of functional counterparts revealed that pseudogenes accumulate mutations at an extremely high rate uniformly over their entirety. It is remarkable that the evolutionary rate exceeds the rate of changes between synonymous codons, the highest known rate, in functional genes. Because no pseudogenes appear to function, this result strongly supports the neutral theory. In addition this result apparently indicates the presence of selective pressure against changes between synonymous codons in functional genes. Close examinations of codon utilization patterns in pseudogenes and functional genes revealed a significant correlation between the rate of changes at synonymous codon sites and the strength of bias in code word usage. This implies that even synonymous codon changes are not completely free from selective pressure but are constrained in part, although presumably weakly, depending on the degree of bias in code word usage. We also reexamined alignment between mouse beta h3 (pseudogene) and beta maj sequences and found a unique structure of the beta h3 that is homologous in sequence to the beta maj gene overall but contains a long deletion (about 150 base pairs) in the middle of the gene.     

The human somatic cytochrome c gene: two classes of processed pseudogenes demarcate a period of rapid molecular evolution

We have isolated and determined the DNA sequences of the human somatic cytochrome c gene (HCS) and 11 processed pseudogenes. HCS is the functional homologue to the previously characterized rat somatic gene because it correctly encodes the human heart protein, is present in single copy in the human genome, is nearly identical in both size and intron/exon structure to rodent somatic genes, and shares a high degree of sequence homology with its rat counterpart including a well-conserved promoter region (77% over 250 nucleotides). In contrast to the rodent system, however, where the known pseudogenes all originated from a locus encoding the present day cytochrome c, the human pseudogenes are of two types. A predominant class of older pseudogenes came from a progenitor of HCS that encoded an ancestral form of the protein, while a second group of only a few young pseudogenes originated from a recent parent of HCS that encoded the current cytochrome c polypeptide. These two distinct classes of human pseudogenes provide a molecular record of the history of cytochrome c evolution in primates and demarcate a short period of rapid evolution of the functional gene.     

The gene for the human erythropoietin receptor: analysis of the coding sequence and assignment to chromosome 19p

The full-length coding sequence of the human erythropoietin receptor has been assembled from cDNA and genomic DNA. The derived 508 amino acid sequence is 82% identical to the murine erythropoietin (Epo) receptor with one single residue gap in alignment. There is no major structural difference between the human and murine receptor molecules. Nucleotide sequence homology is, as expected, very high within the coding domain. Unexpectedly, there are two distinct, short stretches of 3' untranslated sequence homology between human and murine cDNAs. The functional significance of this sequence conservation is unknown. The human Epo receptor gene is localized to human chromosome 19p by in situ hybridization. This chromosome assignment is confirmed by hybridization to a panel of sorted human chromosomes.     

The olfactory receptor gene repertoire in primates and mouse: evidence for reduction of the functional fraction in primates

Olfactory receptors (ORs) located in the cell membrane of olfactory sensory neurons of the nasal epithelium are responsible for odor detection by binding specific odorant ligands. Primates are thought to have a reduced sense of smell (microsmatic) with respect to other mammals such as dogs or rodents. We have previously demonstrated that over 70% of the human OR genes have become nonfunctional pseudogenes, leading us to hypothesize that the reduced sense of smell could correlate with the loss of functional genes. To extend these results, we sampled the OR gene repertoire of 10 primate species, from prosimian lemur to human, in addition to mouse. About 221 previously unidentified primate sequences and 33 mouse sequences were analyzed. These sequences encode ORs distributed in seven families and 56 subfamilies. Analysis showed a high fraction ( approximately 50% on average) of pseudogenes in hominoids. In contrast, only approximately 27% of OR genes are pseudogenes in Old World monkeys, and New World monkeys are almost free of pseudogenes. The prosimian branch seems to have evolved differently from the other primates and has approximately 37% pseudogene content. No pseudogenes were found in mouse. With the exception of New World monkeys, we demonstrate that primates have a high fraction of OR pseudogenes compared with mouse. We hypothesize that under relaxed selective constraints, primates would have progressively accumulated pseudogenes with the highest level seen in hominoids. The fraction of pseudogenes in the OR gene repertoire could parallel the evolution of the olfactory sensory function.     

The emergence of new DNA repeats and the divergence of primates.

We have identified four genetic novelties that are fixed in specific primate lineages and hence can serve as phylogenetic time markers. One Alu DNA repeat is present in the human lineage but is absent from the great apes. Another Alu DNA repeat is present in the gorilla lineage but is absent from the human, chimpanzee, and orangutan. A progenitor Xba1 element is present in the human, chimpanzee, gorilla, and orangutan, but only in the human lineage did it give rise to a transposed progeny, Xba2. The saltatory appearance of Xba2 is an example of a one-time event in the evolutionary history of a species. The enolase pseudogene, known to be present as a single copy in the human, was found to be present in four other primates, including the baboon, an Old World monkey. Using the accepted value of 5 x 10(-9) nucleotide substitutions per site per year as the evolutionary rate for pseudogenes, we calculated that the enolase pseudogene arose approximately 14 million years ago. The calculated age for this pseudogene and its presence in the baboon are incongruent with each other, since Old World monkeys are considered to have diverged from the hominid lineage some 30 million years ago. Thus the rate of evolution in the enolase pseudogene is only about 2.5 x 10(-9) substitutions per site per year, or half the rate in other pseudogenes. It is concluded that rates of substitution vary between species, even for similar DNA elements such as pseudogenes. We submit that new DNA repeats arise in the genomes of species in irreversible and punctuated events and hence can be used as molecular time markers to decipher phylogenies.     

Active late-appearing variable surface antigen genes in Trypanosoma equiperdum are constructed entirely from pseudogenes.

The expression of genes coding for variable surface glycoproteins (VSGs) in Trypanosoma equiperdum is linked to duplicative transpositions of silent, basic copy sequences into telomere-linked expression sites. Examination of three independently derived late-appearing trypanosome clones expressing VSG-78 revealed that the expressed gene in all cases is composed of sequences derived from three or four individual silent genes. The 182 base pairs at the 3' end of the coding sequence are derived from one silent gene, the 3' donor. The remaining 5' segment is a mosaic structure containing variable-length segments derived from two, or perhaps three, related silent genes. All of the silent genes that participate in the construction of the VSG-78 expression-linked copy (ELC) genes contain multiple stop codons and are unable to code for VSGs. Individual silent pseudogenes complement one another in the mosaic structure of the 5' portions of the ELC genes and create functional VSG genes. The joining of the 3' and 5' portions of the composite genes occurs in short regions of homology and suggests a mechanism by which the ordered expression of the VSG genes is generated.