Isolation and characterization of human actin genes

Dynamic actin gene family evolution in primates

Actin is one of the most highly conserved proteins and plays crucial roles in many vital cellular functions. In most eukaryotes, it is encoded by a multigene family. Although the actin gene family has been studied a lot, few investigators focus on the comparison of actin gene family in relative species. Here, the purpose of our study is to systematically investigate characteristics and evolutionary pattern of actin gene family in primates. We identified 233 actin genes in human, chimpanzee, gorilla, orangutan, gibbon, rhesus monkey, and marmoset genomes. Phylogenetic analysis showed that actin genes in the seven species could be divided into two major types of clades: orthologous group versus complex group. Codon usages and gene expression patterns of actin gene copies were highly consistent among the groups because of basic functions needed by the organisms, but much diverged within species due to functional diversification. Besides, many great potential pseudogenes were found with incomplete open reading frames due to frameshifts or early stop codons. These results implied that actin gene family in primates went through “birth and death” model of evolution process. Under this model, actin genes experienced strong negative selection and increased the functional complexity by reproducing themselves.

Sertoli-Sertoli and Sertoli-germ cell interactions and their significance in germ cell movement in the seminiferous epithelium during spermatogenesis

Spermatogenesis is the process by which a single spermatogonium develops into 256 spermatozoa, one of which will fertilize the ovum. Since the 1950s when the stages of the epithelial cycle were first described, reproductive biologists have been in pursuit of one question: How can a spermatogonium traverse the epithelium, while at the same time differentiating into elongate spermatids that remain attached to the Sertoli cell throughout their development? Although it was generally agreed upon that junction restructuring was involved, at that time the types of junctions present in the testis were not even discerned. Today, it is known that tight, anchoring, and gap junctions are found in the testis. The testis also has two unique anchoring junction types, the ectoplasmic specialization and tubulobulbar complex. However, attention has recently shifted on identifying the regulatory molecules that "open" and "close" junctions, because this information will be useful in elucidating the mechanism of germ cell movement. For instance, cytokines have been shown to induce Sertoli cell tight junction disassembly by shutting down the production of tight junction proteins. Other factors such as proteases, protease inhibitors, GTPases, kinases, and phosphatases also come into play. In this review, we focus on this cellular phenomenon, recapping recent developments in the field.

Human cytoplasmic actin proteins are encoded by a multigene family

We characterized nine human actin genes that we isolated (Engel et al., Proc. Natl. Acad. Sci. U.S.A. 78:4674-4678, 1981) from a library of cloned human DNA. Measurements of the thermal stability of hybrids formed between each cloned actin gene and alpha-, beta-, and gamma-actin mRNA demonstrated that only one of the clones is most homologous to sarcomeric actin mRNA, whereas the remaining eight clones are most homologous to cytoplasmic actin mRNA. By the following criteria we show that these nine clones represent nine different actin gene loci rather than different alleles or different parts of a single gene: (i) the restriction enzyme maps of the coding regions are dissimilar; (ii) each clone contains sufficient coding region to encode all or most of an entire actin gene; and (iii) each clone contains sequences homologous to both the 5' and 3' ends of the coding region of a cloned chicken beta-actin cDNA. We conclude, therefore, that the human cytoplasmic actin proteins are encoded by a multigene family.

Human cytoplasmic actin proteins are encoded by a multigene family

We characterized nine human actin genes that we isolated (Engel et al., Proc. Natl. Acad. Sci. U.S.A. 78:4674-4678, 1981) from a library of cloned human DNA. Measurements of the thermal stability of hybrids formed between each cloned actin gene and alpha-, beta-, and gamma-actin mRNA demonstrated that only one of the clones is most homologous to sarcomeric actin mRNA, whereas the remaining eight clones are most homologous to cytoplasmic actin mRNA. By the following criteria we show that these nine clones represent nine different actin gene loci rather than different alleles or different parts of a single gene: (i) the restriction enzyme maps of the coding regions are dissimilar; (ii) each clone contains sufficient coding region to encode all or most of an entire actin gene; and (iii) each clone contains sequences homologous to both the 5' and 3' ends of the coding region of a cloned chicken beta-actin cDNA. We conclude, therefore, that the human cytoplasmic actin proteins are encoded by a multigene family.

Production of human skeletal alpha-actin proteins by the baculovirus expression system

Mutations within the human skeletal muscle alpha-actin gene cause three different skeletal muscle diseases. Functional studies of the mutant proteins are necessary to better understand the pathogenesis of these diseases, however, no satisfactory system for the expression of mutant muscle actin proteins has been available. We investigated the baculovirus expression vector system (BEVS) for the abundant production of both normal and mutant skeletal muscle alpha-actin. We show that non-mutated actin produced in the BEVS behaves similarly to native actin, as shown by DNase I affinity purification, Western blotting, and consecutive cycles of polymerisation and depolymerisation. Additionally, we demonstrate the production of mutant actin proteins in the BEVS, without detriment to the insect cells in which they are expressed. The BEVS therefore is the method of choice for studying mutant actin proteins causing human diseases.

Actin of Histriculus cavicola: characteristics of the highly divergent hypotrich ciliate actins

A macronuclear gene-sized molecule carrying an actin gene from the hypotrich ciliate, Histriculus cavicola, was characterized. Southern blot analysis using a coding region probe suggested that actin in H. cavicola is encoded by a single gene. A comparison of the promoter regions indicated that the H. cavicola actin gene has a TATA box in the 5' flanking region in a position identical to those in other oxytrich ciliates. The coding sequence of this gene is not interrupted by any introns, and codes for a protein of 375 amino acid residues. This protein shares a high degree of similarity with other oxytrichid actins, and a relatively low similarity with actins from other eukaryotes. Comparative analyses of sequences indicated that most of the amino acid substitutions in hypotrich actins are found in surface loops, while the core structures are well-conserved. The sites that interact with DNase I and several regions involved in actin-actin contact have diverged considerably in hypotrich actins, while nucleotide-binding sites are the best-conserved interaction motif.

Codon bias in actin multigene families and effects on the reconstruction of phylogenetic relationships

Codon usage patterns and phylogenetic relationships in the actin multigene family have been analyzed for three dipteran species--Drosophila melanogaster, Bactrocera dorsalis, and Ceratitis capitata. In certain phylogenetic tree reconstructions, using synonymous distances, some gene relationships are altered due to a homogenization phenomenon. We present evidence to show that this homogenization phenomenon is due to codon usage bias. A survey of the pattern of synonymous codon preferences for 11 actin genes from these three species reveals that five out of the six Drosophila actin genes show high degrees of codon bias as indicated by scaled chi 2 values. In contrast to this, four out of the five actin genes from the other species have low codon bias values. A Monte Carlo contingency test indicates that for those Drosophila actin genes which exhibit codon bias, the patterns of codon usage are different compared to actin genes from the other species. In addition, the genes exhibiting codon bias also appear to have reduced rates of synonymous substitution. The homogenization phenomenon seen in terms of synonymous substitutions is not observed for nonsynonymous changes. Because of this homogenization phenomenon, "trees" constructed based on synonymous substitutions will be affected. These effects can be overt in the case of multigene families, but similar distortions may underlie reconstructions based on single-copy genes which exhibit codon usage bias.

Molecular analysis of bovine actin gene and pseudogene sequences: expression of nonmuscle and striated muscle isoforms in adult tissues

Most studies on the tissue distribution of actin isoform transcripts have been done in small mammals such as rat and mouse. We have begun a characterization of the actin gene family in a large mammal, the bovine. The alpha skeletal gene was isolated, and an isoform-specific probe to the 3' untranslated region of the transcript identified. This probe, in combination with isoform specific probes for alpha cardiac, beta nonmuscle, and gamma nonmuscle actins, was used to examine expression of nonmuscle and striated muscle actin gene transcription in different tissues. In contrast to other species so far examined, striated muscle isoforms were more strictly tissue specific, with virtually no alpha cardiac isoform transcripts detected in skeletal muscle and almost no alpha skeletal transcripts in cardiac tissue. The distribution of the beta and gamma nonmuscle actins was also unique in bovine compared to other species. A partial beta-actin pseudogene, and the chromosomal DNA flanking one end of it, were also cloned and sequenced. This chromosomal site was found to be homologous to a viral integration site previously identified in simian virus 40 (SV40)-transformed rat cells, suggesting that this region of the chromosome may be a preferred target for insertion events.

SPARC, a secreted protein associated with morphogenesis and tissue remodeling, induces expression of metalloproteinases in fibroblasts through a novel extracellular matrix-dependent pathway

SPARC (osteonectin/BM40) is a secreted protein that modifies the interaction of cells with extracellular matrix (ECM). When we added SPARC to cultured rabbit synovial fibroblasts and analyzed the secreted proteins, we observed an increase in the expression of three metalloproteinases--collagenase, stromelysin, and the 92-kD gelatinase--that together can degrade both interstitial and basement membrane matrices. We further characterized the regulation of one of these metalloproteinases, collagenase, and showed that both collagenase mRNA and protein are upregulated in fibroblasts treated with SPARC. Experiments with synthetic SPARC peptides indicated that a region in the neutral alpha-helical domain III of the SPARC molecule, which previously had no described function, was involved in the regulation of collagenase expression by SPARC. A sequence in the carboxyl-terminal Ca(2+)-binding domain IV exhibited similar activity, but to a lesser extent. SPARC induced collagenase expression in cells plated on collagen types I, II, III, and V, and vitronectin, but not on collagen type IV. SPARC also increased collagenase expression in fibroblasts plated on ECM produced by smooth muscle cells, but not in fibroblasts plated on a basement membrane-like ECM from Engelbreth-Holm-Swarm sarcoma. Collagenase was induced within 4 h in cells treated with phorbol diesters or plated on fibronectin fragments, but was induced after 8 h in cells treated with SPARC. A number of proteins were transiently secreted by SPARC-treated cells within 6 h of treatment. Conditioned medium that was harvested from cultures 7 h after the addition of SPARC, and depleted of residual SPARC, induced collagenase expression in untreated fibroblasts; thus, part of the regulation of collagenase expression by SPARC appears to be indirect and proceeds through a secreted intermediate. Because the interactions of cells with ECM play an important role in regulation of cell behavior and tissue morphogenesis, these results suggest that molecules like SPARC are important in modulating tissue remodeling and cell-ECM interactions.

The use of reverse transcriptase-polymerase chain reaction (RT-PCR) to investigate specific gene expression in multidrug-resistant cells

Expression of specific genes at the level of mRNA can be studied using techniques such as Northern blot, slot/dot blot, RNase protection assay, in situ hybridisation and RT-PCR. In this article these methods of analysis are compared; RT-PCR offers higher levels of specificity and sensitivity than traditional methods of RNA analysis and as such has become the method of choice for the study of gene expression. The RT-PCR technique is described in detail with sections dealing with RNA extraction, choice of primers (including the use of cDNA sequence data bases), PCR and RT-PCR protocols in addition to the limitations of the method. The study of one particular mRNA transcript (MDR1) using RT-PCR is discussed in detail. Recently described methods for quantitation of PCR products are discussed. Quantitative PCR would appear to offer a method of studying gene expression in a more extensive way than has been possible to date.

The mouse smooth muscle gamma actin gene is on chromosome 6

Smooth muscle gamma actin (Actg) is expressed in smooth muscle and in haploid male germ cells. In order to further characterize the Actg gene, a 60-nucleotide-long isotype-specific probe was synthesized. Single bands of DNA were detected when this oligonucleotide was used to probe blots of mouse genomic DNA digested with PstI, EcoRI, KpnI, or XbaI. These results suggest Actg is a single-copy gene with no detectable pseudogenes. The Actg gene was mapped to mouse chromosome 6 by Southern blot analysis of DNA isolated from 15 mouse-hamster hybrid cell lines.

Signal transduction through the fibronectin receptor induces collagenase and stromelysin gene expression

We have investigated the effects of ligation of the fibronectin receptor (FnR) on gene expression in rabbit synovial fibroblasts. Monoclonal antibodies to the FnR that block initial adhesion of fibroblasts to fibronectin induced the expression of genes encoding the secreted extracellular matrix-degrading metalloproteinases collagenase and stromelysin. That induction was a direct consequence of interaction with the FnR was shown by the accumulation of mRNA for stromelysin and collagenase. Monoclonal antibodies to several other membrane glycoprotein receptors had no effect on metalloproteinase gene expression. Less than 2 h of treatment of the fibroblasts with anti-FnR in solution was sufficient to trigger the change in gene expression, and induction was blocked by dexamethasone. Unlike other inducers of metalloproteinase expression, including phorbol diesters and growth factors, addition of the anti-FnR in solution to cells adherent to serum-derived adhesion proteins or collagen produced no detectable change in cell shape or actin microfilament organization. Inductive effects were potentiated by cross-linking of the ligand. Fab fragments of anti-FnR were ineffective unless cross-linked or immobilized on the substrate. Adhesion of fibroblasts to native fibronectin did not induce metallo-proteinases. However, adhesion to covalently immobilized peptides containing the arg-gly-asp sequence that were derived from fibronectin, varying in size from hexapeptides up to 120 kD, induced collagenase and stromelysin gene expression. This suggests that degradation products of fibronectin are the natural inductive ligands for the FnR. These data demonstrate that signals leading to changes in gene expression are transduced by the FnR, a member of the integrin family of extracellular matrix receptors. The signaling of changes in gene expression by the FnR is distinct from signaling involving cell shape and actin cytoarchitecture. At least two distinct signals are generated: the binding of fibronectin-derived fragments and adhesion-blocking antibodies to the FnR triggers events different from those triggered by binding of the native fibronectin ligand. Because the genes regulated by this integrin are for enzymes that degrade the extracellular matrix, these results suggest that information transduced by the binding of various ligands to integrins may orchestrate the expression of genes regulating cell behavior in the extracellular environment.

Expression of RNA transcripts for the postsynaptic 43 kDa protein in innervated and denervated rat skeletal muscle

A cDNA clone encoding the mouse muscle postsynaptic 43 kDa protein was isolated and sequenced. The amino acid sequence of this protein, which is closely associated with nicotinic acetylcholine receptors at Torpedo electrocyte and vertebrate skeletal muscle synapses, is very similar in different species. A cysteine-rich region homologous to part of the regulatory domain of protein kinase C may be important in interactions of this protein with the lipid bilayer. RNA transcripts for the 43 kDa protein increase only 2-3 fold after denervation of rat skeletal muscle, in sharp contrast to the alpha-subunit of the muscle nicotinic receptor which increases more than 30-fold. Thus, the expression of these two proteins is regulated by different mechanisms.

The Drosophila melanogaster tropomyosin II gene produces multiple proteins by use of alternative tissue-specific promoters and alternative splicing

The structure of the Drosophila melanogaster tropomyosin II (TmII) gene has been determined by DNA sequencing of cDNA clones and the genomic DNA coding for the gene. Two overlapping transcriptional units produce at least four different tropomyosin isoforms. A combination of developmentally regulated promoters and alternative splicing produces both muscle and cytoskeletal tropomyosin isoforms. One promoter is a muscle-specific promoter and produces three different tropomyosin isoforms by alternative splicing of the last three 3' exons. The second promoter has the characteristics of a housekeeping promoter and produces a cytoskeletal tropomyosin isoform. Several internal exons along with a final 3' exon are alternatively spliced in the cytoskeletal transcript. The intron-exon boundaries of the TmII gene are identical to the intron-exon boundaries of all vertebrate tropomyosin genes reported, but are very different from the intron-exon boundaries of the D. melanogaster tropomyosin I gene. The TmII gene is the only reported tropomyosin gene that has two promoters and a quadruple alternative splice choice for the final exon. Models for the mechanism of D. melanogaster tropomyosin gene evolution are discussed.

Alternative 5C actin transcripts are localized in different patterns during Drosophila embryogenesis

The Drosophila actin gene located at cytogenetic position 5C forms at least 9 and perhaps as many as 15 different transcripts with the use of alternative transcriptional start points, differential splicing, and different regions of cleavage/polyadenylation. Each transcript contains one of two alternative 5' exons. We have subcloned unique recombinant DNA probes specific for each separate 5' exon and for three polyadenylation regions into vectors containing T3 and T7 promoters. Single stranded, tritium-labeled RNA probes were generated in vitro from these constructs. These probes have been hybridized in situ to RNA transcripts present in tissue sections from Drosophila embryos. The results of these experiments indicate that transcripts homologous to the two separate 5' exon-specific probes accumulate in strikingly different patterns during Drosophila development. Thus the incorporation of a particular 5' exon into a transcript is correlated with tissue-specific localization of that transcript. In contrast, probes for each of the three polyadenylation regions do not detect any tissue-specific localization of transcripts.

A highly conserved mouse gene with a propensity to form pseudogenes in mammals

A mouse cDNA clone corresponding to an abundantly transcribed poly(A)+ mRNA was found to be represented by 200 copies in mammalian genomes. To understand the origin and nature of this sequence family, we studied two genomic members and two cDNA clones from mouse liver. The DNA sequence of the coding strand of a full-length cDNA clone was shown to have an open reading frame capable of encoding a 25-kilodalton polypeptide that has not been previously described. In vitro transcription-translation experiments verified the presence of an open reading frame encoding a protein of the predicted size. Restriction analysis of genomic DNA and DNA sequence analysis of genomic clones indicated that many of the 200 members of this family represent processed pseudogenes, with one or a small number of active structural genes. The vast majority of the genomic copies are heterogeneous in length, truncated at their 5' ends with respect to the mRNA, and do not appear to have intervening sequences. Two distinct genomic members of this family were sequenced and found to represent incomplete copies of the mRNA. Both are 5' truncated at slightly different points with respect to the mRNA. Both pseudogenes have multiple base changes, insertions, and deletions relative to the mRNA, and one of them encodes the poly(A) tail of the mRNA. The expression of this gene family is highest in rapidly dividing cells such as early mouse embryos and testis, but was seen in all tissues tested. This gene shows extremely high sequence conservation, extending to chicken, amphibian, and nematode genomes. Surprisingly, the gene appears to exist in only one copy in these organisms.

The Drosophila melanogaster tropomyosin II gene produces multiple proteins by use of alternative tissue-specific promoters and alternative splicing

The structure of the Drosophila melanogaster tropomyosin II (TmII) gene has been determined by DNA sequencing of cDNA clones and the genomic DNA coding for the gene. Two overlapping transcriptional units produce at least four different tropomyosin isoforms. A combination of developmentally regulated promoters and alternative splicing produces both muscle and cytoskeletal tropomyosin isoforms. One promoter is a muscle-specific promoter and produces three different tropomyosin isoforms by alternative splicing of the last three 3' exons. The second promoter has the characteristics of a housekeeping promoter and produces a cytoskeletal tropomyosin isoform. Several internal exons along with a final 3' exon are alternatively spliced in the cytoskeletal transcript. The intron-exon boundaries of the TmII gene are identical to the intron-exon boundaries of all vertebrate tropomyosin genes reported, but are very different from the intron-exon boundaries of the D. melanogaster tropomyosin I gene. The TmII gene is the only reported tropomyosin gene that has two promoters and a quadruple alternative splice choice for the final exon. Models for the mechanism of D. melanogaster tropomyosin gene evolution are discussed.

Structure, chromosome location, and expression of the human gamma-actin gene: differential evolution, location, and expression of the cytoskeletal beta- and gamma-actin genes

The accumulation of the cytoskeletal beta- and gamma-actin mRNAs was determined in a variety of mouse tissues and organs. The beta-isoform is always expressed in excess of the gamma-isoform. However, the molar ratio of beta- to gamma-actin mRNA varies from 1.7 in kidney and testis to 12 in sarcomeric muscle to 114 in liver. We conclude that, whereas the cytoskeletal beta- and gamma-actins are truly coexpressed, their mRNA levels are subject to differential regulation between different cell types. The human gamma-actin gene has been cloned and sequenced, and its chromosome location has been determined. The gene is located on human chromosome 17, unlike beta-actin which is on chromosome 7. Thus, if these genes are also unlinked in the mouse, the coexpression of the beta- and gamma-actin genes in rodent tissues cannot be determined by gene linkage. Comparison of the human beta- and gamma-actin genes reveals that noncoding sequences in the 5'-flanking region and in intron III have been conserved since the duplication that gave rise to these two genes. In contrast, there are sequences in intron III and the 3'-untranslated region which are not present in the beta-actin gene but are conserved between the human gamma-actin and the Xenopus borealis type 1 actin genes. Such conserved noncoding sequences may contribute to the coexpression of beta- and gamma-actin or to the unique regulation and function of the gamma-actin gene. Finally, we demonstrate that the human gamma-actin gene is expressed after introduction into mouse L cells and C2 myoblasts and that, upon fusion of C2 cells to form myotubes, the human gamma-actin gene is appropriately regulated.

A highly conserved mouse gene with a propensity to form pseudogenes in mammals

A mouse cDNA clone corresponding to an abundantly transcribed poly(A)+ mRNA was found to be represented by 200 copies in mammalian genomes. To understand the origin and nature of this sequence family, we studied two genomic members and two cDNA clones from mouse liver. The DNA sequence of the coding strand of a full-length cDNA clone was shown to have an open reading frame capable of encoding a 25-kilodalton polypeptide that has not been previously described. In vitro transcription-translation experiments verified the presence of an open reading frame encoding a protein of the predicted size. Restriction analysis of genomic DNA and DNA sequence analysis of genomic clones indicated that many of the 200 members of this family represent processed pseudogenes, with one or a small number of active structural genes. The vast majority of the genomic copies are heterogeneous in length, truncated at their 5' ends with respect to the mRNA, and do not appear to have intervening sequences. Two distinct genomic members of this family were sequenced and found to represent incomplete copies of the mRNA. Both are 5' truncated at slightly different points with respect to the mRNA. Both pseudogenes have multiple base changes, insertions, and deletions relative to the mRNA, and one of them encodes the poly(A) tail of the mRNA. The expression of this gene family is highest in rapidly dividing cells such as early mouse embryos and testis, but was seen in all tissues tested. This gene shows extremely high sequence conservation, extending to chicken, amphibian, and nematode genomes. Surprisingly, the gene appears to exist in only one copy in these organisms.

Structure, chromosome location, and expression of the human gamma-actin gene: differential evolution, location, and expression of the cytoskeletal beta- and gamma-actin genes

The accumulation of the cytoskeletal beta- and gamma-actin mRNAs was determined in a variety of mouse tissues and organs. The beta-isoform is always expressed in excess of the gamma-isoform. However, the molar ratio of beta- to gamma-actin mRNA varies from 1.7 in kidney and testis to 12 in sarcomeric muscle to 114 in liver. We conclude that, whereas the cytoskeletal beta- and gamma-actins are truly coexpressed, their mRNA levels are subject to differential regulation between different cell types. The human gamma-actin gene has been cloned and sequenced, and its chromosome location has been determined. The gene is located on human chromosome 17, unlike beta-actin which is on chromosome 7. Thus, if these genes are also unlinked in the mouse, the coexpression of the beta- and gamma-actin genes in rodent tissues cannot be determined by gene linkage. Comparison of the human beta- and gamma-actin genes reveals that noncoding sequences in the 5'-flanking region and in intron III have been conserved since the duplication that gave rise to these two genes. In contrast, there are sequences in intron III and the 3'-untranslated region which are not present in the beta-actin gene but are conserved between the human gamma-actin and the Xenopus borealis type 1 actin genes. Such conserved noncoding sequences may contribute to the coexpression of beta- and gamma-actin or to the unique regulation and function of the gamma-actin gene. Finally, we demonstrate that the human gamma-actin gene is expressed after introduction into mouse L cells and C2 myoblasts and that, upon fusion of C2 cells to form myotubes, the human gamma-actin gene is appropriately regulated.

Regulation of lymphokine production and human T lymphocyte activation by 1,25-dihydroxyvitamin D3. Specific inhibition at the level of messenger RNA

The steroid hormone, 1 alpha,25-dihydroxyvitamin D3 (calcitriol), has been shown to inhibit T cell proliferation, primarily through inhibition of interleukin 2 (IL-2) production. In these experiments, we show that calcitriol also markedly inhibited production of the lymphokine, gamma interferon (IFN-gamma), by activated human T lymphocytes. Regulation of both IL-2 and IFN-gamma production as well as transferrin receptor (TfR) expression by calcitriol was apparent at the messenger RNA (mRNA) level as determined by Northern blotting. The decrease in IL-2 and IFN-gamma mRNA that occurred with calcitriol treatment was coordinate and not apparent up to 12 h after phytohemagglutinin stimulation, whereas decreased accumulation of TfR mRNA was not present before 24-36 h. Furthermore, the effects of calcitriol on IL-2, IFN-gamma, and TfR mRNA accumulation were specific; actin mRNA accumulation was comparable between control and treated cells. These data indicate that calcitriol regulated proteins associated with T cell activation at the transcriptional level and that these effects were mediated in a specific, coordinate fashion.

The molecular evolution of actin

We have investigated the molecular evolution of plant and nonplant actin genes comparing nucleotide and amino acid sequences of 20 actin genes. Nucleotide changes resulting in amino acid substitutions (replacement substitutions) ranged from 3-7% for all pairwise comparisons of animal actin genes with the following exceptions. Comparisons between higher animal muscle actin gene sequences and comparisons between higher animal cytoplasmic actin gene sequences indicated less than 3% divergence. Comparisons between plant and nonplant actin genes revealed, with two exceptions, 11-15% replacement substitution. In the analysis of plant actins, replacement substitution between soybean actin genes SAc1, SAc3, SAc4 and maize actin gene MAc1 ranged from 8-10%, whereas these members within the soybean actin gene family ranged from 6-9% replacement substitution. The rate of sequence divergence of plant actin sequences appears to be similar to that observed for animal actins. Furthermore, these and other data suggest that the plant actin gene family is ancient and that the families of soybean and maize actin genes have diverged from a single common ancestral plant actin gene that originated long before the divergence of monocots and dicots. The soybean actin multigene family encodes at least three classes of actin. These classes each contain a pair of actin genes that have been designated kappa (SAc1, SAc6), lambda (SAc2, SAc4) and mu (SAc3, SAc7). The three classes of soybean actin are more divergent in nucleotide sequence from one another than higher animal cytoplasmic actin is divergent from muscle actin. The location and distribution of amino acid changes were compared between actin proteins from all sources. A comparison of the hydropathy of all actin sequences, except from Oxytricha, indicated a strong similarity in hydropathic character between all plant and nonplant actins despite the greater number of replacement substitutions in plant actins. These protein sequence comparisons are discussed with respect to the demonstrated and implicated roles of actin in plants and animals, as well as the tissue-specific expression of actin.

Nucleotide sequence of the human gamma cytoskeletal actin mRNA: anomalous evolution of vertebrate non-muscle actin genes

Two distinct, but iso-coding, gamma non-muscle actin cDNAs were isolated from an SV40-transformed human fibroblast library. The complete nucleotide sequence of the human gamma non-muscle actin cDNAs indicates that they may have arisen from polymorphic alleles. By using genomic DNA and cellular RNA transfer blots, we demonstrate that the 3' untranslated region (UTR) of the gamma actin mRNA consists of an evolutionarily conserved 5' and more divergent 3' segments. In fact, the conserved segment of the 3' UTR detects a single-copy sequence in the chicken genome and a 20S RNA transcript in chicken non-muscle tissues. The coding regions of these cDNAs were compared with those of other vertebrate non-muscle actin genes. Surprisingly, the percentage of silent base substitutions between the human beta and gamma actin coding regions is anomalously low and indicates greater sequence conservation than would be expected for a gene pair which arose during pre-avian evolution. We discuss gene conversion and recent selective pressure as possible explanations of the apparently anomalous evolution of the gamma non-muscle actin gene.

Genetic alteration of the c-myc protooncogene (MYC) in human primary breast carcinomas

We have studied the genomic organization of the c-myc locus (MYC) from 121 human primary breast carcinomas. Two types of alterations were observed: (i) the c-myc protooncogene appeared to be amplified 2- to 15-fold in 38 (32%) of the carcinoma DNAs and (ii) a non-germ-line c-myc-related fragment of variable size was detected in 5 primary breast carcinoma DNAs. With three exceptions, all the tumors containing a genetic alteration of the c-myc locus were invasive ductal carcinomas. A significant correlation (P less than 0.02) was observed between patients more than 50 years old and the presence of a genetically altered c-myc. Enhanced levels of c-myc RNA were observed in 10 of 14 breast carcinomas examined. The c-myc gene was genetically altered in 6 of these 10 tumors. The frequency with which the c-myc gene is altered and its correlation with age suggest that it may play a role in the development of breast carcinomas.

RNA metabolism in nuclei: selective transport of kappa exons from myeloma nuclei and adenoviral transcripts from infected HeLa nuclei

Two independent systems and several analytical procedures have been used to establish that isolated mammalian nuclei selectively transport mature RNA polymerase I and II products. Murine myeloma nuclei retain physiologic restriction in our transport assay as assessed by the transport of the immunoglobulin kappa light chain mRNA and 18S and 28S rRNAs. Nearly 50% of the total kappa exons are transported as structurally intact mature mRNA molecules while less than 8% of either pulse-labeled or steady state kappa intron sequences are detected in the transported fraction. Ribosomal external transcribed spacer sequences also are absent in transported RNA. Release of cytoplasmic RNA from the outer nuclear membrane during the transport assay accounts for less than 10% of transported RNA. Nuclei isolated from adenovirus-infected HeLa cells at 20 hours post infection retain cellular actin mRNA and transport viral poly A+RNA. Ribosomal RNA is transported from infected nuclei although at a reduced rate compared to transport from mock-infected nuclei. Inhibition of transport of host mRNA is paralleled by the absence of pulse-labeled actin mRNA in the cytoplasm of infected cells. The implications of our transport data in relationship to intranuclear RNA trafficking are discussed.

Comparison of three actin-coding sequences in the mouse; evolutionary relationships between the actin genes of warm-blooded vertebrates

We have determined the sequences of three recombinant cDNAs complementary to different mouse actin mRNAs that contain more than 90% of the coding sequences and complete or partial 3' untranslated regions (3'UTRs): pAM 91, complementary to the actin mRNA expressed in adult skeletal muscle (alpha sk actin); pAF 81, complementary to an actin mRNA that is accumulated in fetal skeletal muscle and is the major transcript in adult cardiac muscle (alpha c actin); and pAL 41, identified as complementary to a beta nonmuscle actin mRNA on the basis of its 3'UTR sequence. As in other species, the protein sequences of these isoforms are highly (greater than 93%) conserved, but the three mRNAs show significant divergence (13.8-16.5%) at silent nucleotide positions in their coding regions. A nucleotide region located toward the 5' end shows significantly less divergence (5.6-8.7%) among the three mouse actin mRNAs; a second region, near the 3' end, also shows less divergence (6.9%), in this case between the mouse beta and alpha sk actin mRNAs. We propose that recombinational events between actin sequences may have homogenized these regions. Such events distort the calculated evolutionary distances between sequences within a species. Codon usage in the three actin mRNAs is clearly different, and indicates that there is no strict relation between the tissue type, and hence the tRNA precursor pool, and codon usage in these and other muscle mRNAs examined. Analysis of codon usage in these coding sequences in different vertebrate species indicates two tendencies: increases in bias toward the use of G and C in the third codon position in paralogous comparisons (in the order alpha c less than beta less than alpha sk), and in orthologous comparisons (in the order chicken less than rodent less than man). Comparison of actin-coding sequences between species was carried out using the Perler method of analysis. As one moves backward in time, changes at silent sites first accumulate rapidly, then begin to saturate after -(30-40) million years (MY), and actually decrease between -400 and -500 MY. Replacements or silent substitutions therefore cannot be used as evolutionary clocks for these sequences over long periods. Other phenomena, such as gene conversion or isochore compartmentalization, probably distort the estimated divergence time.

Molecular structure of the human cytoplasmic beta-actin gene: interspecies homology of sequences in the introns

A recombinant phage that carries the cytoplasmic beta-actin gene was isolated from a human DNA library. The nucleotide sequence of this gene was determined. The amino acid sequence deduced from the nucleotide sequence matches perfectly that of beta-actin from human fibroblasts. The gene contains five introns. A large intron was found in the 5' untranslated region six nucleotides upstream from the ATG initiation codon. Four introns were found within the coding region at codons specifying amino acids 41/42, 121/122, 267, and 327/328. In contrast to the human cardiac muscle actin gene, the aorta-type smooth muscle actin gene, and the stomach-type smooth muscle actin gene, the beta-actin gene lacks the codon for cysteine between the ATG initiation codon and the codon for the NH2-terminal amino acid of the mature protein. Hybridization of genomic DNA with DNA fragments derived from intron I in the 5' untranslated region and from intron III strongly suggests the presence of a single beta-actin gene in the human genome. The DNA sequences of the coding region, of the 3' untranslated region, and of the sequence block between the "CCAAT" box and "TATA" box in the 5' flanking DNA of the human beta-actin gene are highly homologous to the corresponding sequences of the rat and chicken beta-actin genes. Unexpectedly, the sequence of intron III of the human beta-actin gene shows considerable homology to that of the rat beta-actin gene.

Novel chicken actin gene: third cytoplasmic isoform

We identified a novel chicken actin gene. The actin protein deduced from its nucleotide sequence very closely resembles the vertebrate cytoplasmic actins; accordingly, we classified this gene as a nonmuscle type. We adopted the convention for indicating the nonmuscle actins of the class Amphibia (Vandekerckhove et al., J. Mol. Biol. 152:413-426) and denoted this gene as type 5. RNA blot analysis demonstrated that the type 5 actin mRNA transcripts accumulate in adult tissues in a pattern indicative of a nonmuscle actin gene. Genomic DNA blots indicated that the type 5 actin is a single copy gene and a distinct member of the chicken actin multigene family. Inspection of the nucleotide sequence revealed many features that distinguished the type 5 gene from all other vertebrate actin genes examined to date. These unique characteristics include: (i) an initiation Met codon preceding an Ala codon, a feature previously known only in plant actins, (ii) a single intron within the 5' untranslated region, with no interruptions in the coding portion of the gene, and (iii) an atypical Goldberg-Hogness box (ATAGAA) preceding the mRNA initiation terminus. These unusual features have interesting implications for actin gene diversification during evolution.

Novel chicken actin gene: third cytoplasmic isoform

We identified a novel chicken actin gene. The actin protein deduced from its nucleotide sequence very closely resembles the vertebrate cytoplasmic actins; accordingly, we classified this gene as a nonmuscle type. We adopted the convention for indicating the nonmuscle actins of the class Amphibia (Vandekerckhove et al., J. Mol. Biol. 152:413-426) and denoted this gene as type 5. RNA blot analysis demonstrated that the type 5 actin mRNA transcripts accumulate in adult tissues in a pattern indicative of a nonmuscle actin gene. Genomic DNA blots indicated that the type 5 actin is a single copy gene and a distinct member of the chicken actin multigene family. Inspection of the nucleotide sequence revealed many features that distinguished the type 5 gene from all other vertebrate actin genes examined to date. These unique characteristics include: (i) an initiation Met codon preceding an Ala codon, a feature previously known only in plant actins, (ii) a single intron within the 5' untranslated region, with no interruptions in the coding portion of the gene, and (iii) an atypical Goldberg-Hogness box (ATAGAA) preceding the mRNA initiation terminus. These unusual features have interesting implications for actin gene diversification during evolution.

Molecular cloning and characterization of mutant and wild-type human beta-actin genes

There are more than 20 beta-actin-specific sequences in the human genome, many of which are pseudogenes. To facilitate the isolation of potentially functional beta-actin genes, we used the new method of B. Seed (Nucleic Acids Res. 11:2427-2446, 1983) for selecting genomic clones by homologous recombination. A derivative of the pi VX miniplasmid, pi AN7 beta 1, was constructed by insertion of the 600-base-pair 3' untranslated region of the beta-actin mRNA expressed in human fibroblasts. Five clones containing beta-actin sequences were selected from an amplified human fetal gene library by homologous recombination between library phage and the miniplasmid. One of these clones contained a complete beta-actin gene with a coding sequence identical to that determined for the mRNA of human fibroblasts. A DNA fragment consisting of mostly intervening sequences from this gene was then used to identify 13 independent recombinant copies of the analogous gene from two specially constructed gene libraries, each containing one of the two types of mutant beta-actin genes found in a line of neoplastic human fibroblasts. The amino acid and nucleotide sequences encoded by the unmutated gene predict that a guanine-to-adenine transition is responsible for the glycine-to-aspartic acid mutation at codon 244 and would also result in the loss of a HaeIII site. Detection of this HaeIII polymorphism among the fibroblast-derived clones verified the identity of the beta-actin gene expressed in human fibroblasts.

Molecular cloning and characterization of mutant and wild-type human beta-actin genes

There are more than 20 beta-actin-specific sequences in the human genome, many of which are pseudogenes. To facilitate the isolation of potentially functional beta-actin genes, we used the new method of B. Seed (Nucleic Acids Res. 11:2427-2446, 1983) for selecting genomic clones by homologous recombination. A derivative of the pi VX miniplasmid, pi AN7 beta 1, was constructed by insertion of the 600-base-pair 3' untranslated region of the beta-actin mRNA expressed in human fibroblasts. Five clones containing beta-actin sequences were selected from an amplified human fetal gene library by homologous recombination between library phage and the miniplasmid. One of these clones contained a complete beta-actin gene with a coding sequence identical to that determined for the mRNA of human fibroblasts. A DNA fragment consisting of mostly intervening sequences from this gene was then used to identify 13 independent recombinant copies of the analogous gene from two specially constructed gene libraries, each containing one of the two types of mutant beta-actin genes found in a line of neoplastic human fibroblasts. The amino acid and nucleotide sequences encoded by the unmutated gene predict that a guanine-to-adenine transition is responsible for the glycine-to-aspartic acid mutation at codon 244 and would also result in the loss of a HaeIII site. Detection of this HaeIII polymorphism among the fibroblast-derived clones verified the identity of the beta-actin gene expressed in human fibroblasts.

The 3'-untranslated sequence of human skeletal muscle alpha-actin mRNA

A human actin cDNA clone pGF3 isolated from a fetal skeletal muscle cDNA library is described. The insert cDNA is homologous to skeletal muscle alpha-actin as judged by restriction mapping and nucleotide sequencing. The recombinant contains a substantial portion of the coding and the complete 3'-untranslated region. Comparison of the 3' ends of human and rat skeletal muscle and human cardiac alpha-actins reveals little homology between different types of actin genes in man but marked conservation of this region in the skeletal muscle actins of man and rat.

Evolution of the human sarcomeric-actin genes: evidence for units of selection within the 3' untranslated regions of the mRNAs

The complete 3' untranslated region (3'UTR) sequence of the human skeletal-actin gene has been compared with the corresponding regions of the rat and chicken skeletal-actin genes. This comparison reveals that the skeletal-actin 3'UTR is composed of conserved and nonconserved segments. By using genomic Southern transfer blots and thermal stability (Tm) measurements, we found that the cardiac-actin gene 3'UTR also consists of conserved and nonconserved segments. Comparison of human and Xenopus laevis cardiac-actin mRNA sequences confirms the presence of a region of high similarity in the 3'UTR. We conclude that subsegments of the 3'UTRs of both skeletal- and cardiac-actin genes of birds and mammals are under considerable selective pressure. This suggests that these conserved sequences may have functional roles in actin-gene expression or regulation, and that these roles might be different for each actin isoform.

Isolation and characterization of human random cDNA clones homologous to DNA from the X chromosome

To search for human X-chromosome-specific probes useful for molecular mapping, we studied recombinant clones isolated from a human cDNA library. DNA preparations from 150 randomly selected clones were labeled and annealed to XY and 4XY human DNA, and to DNA from a human-mouse hybrid cell line that had retained only the human X-chromosome (A9/HRBC2). cDNA clones sharing homology with DNA from the X chromosome annealed to A9/HRBC2-DNA and hybridized more intensely to 4XY DNA than to XY DNA. Eleven such clones were identified. Of these, three hybridized only to X chromosomal DNA while the rest also annealed to DNA from one or more autosomes. Chromosomal assignment of the autosomal DNA fragments showed that, in addition to hybridization to X chromosomal DNA, four of the clones hybridized to DNA sequences from chromosome 2 and two clones to chromosome 7. Subregional mapping of the relevant X chromosomal DNA fragments indicated that one clone is homologous to DNA sequences located at Xp21-Xp22, whereas the others are located in the telomeric region of the long arm. The cDNA clones were used to search for restriction fragment length polymorphisms. Several restriction-site polymorphisms were detected. Some corresponded to variants of X chromosomal DNA sequences while others were from autosomes such as chromosomes 2 and 7.

DNA restriction fragment length polymorphisms and heterozygosity in the human genome

A list is presented of published reports of DNA polymorphisms found in the human genome by restriction enzyme analysis. While the list indicates the large number of restriction fragment length polymorphisms (RFLPs) detected to date, the information collated is insufficient to permit an estimate of heterozygosity for the genome as a whole. Data from our laboratory are therefore also presented on RFLPs detected using a random sample of cloned DNA segments. Such an analysis has permitted a first unbiassed estimate of heterozygosity for the human genome. Since this figure is an order of magnitude higher than previous estimates derived from protein data, the majority of polymorphic variation present in the human genome must, by implication, occur in noncoding sequences. In addition it was confirmed that enzymes containing the dinucleotide CpG in their recognition sequences detect more polymorphic variation than those that do not contain a CpG. Also presented are the clinical applications of DNA polymorphisms in the diagnosis of human genetic disease.

alpha-skeletal and alpha-cardiac actin genes are coexpressed in adult human skeletal muscle and heart

We determined the actin isotypes encoded by 30 actin cDNA clones previously isolated from an adult human muscle cDNA library. Using 3' untranslated region probes derived from alpha-skeletal, beta- and gamma-actin cDNAs and from an alpha-cardiac actin genomic clone, we showed that 28 of the cDNAs correspond to alpha-skeletal actin transcripts. Unexpectedly, however, the remaining two cDNA clones proved to derive from alpha-cardiac actin mRNA. Sequence analysis confirmed that the two skeletal muscle alpha-cardiac actin cDNAs are derived from transcripts of the cloned alpha-cardiac actin gene. Direct measurements of actin isotype mRNA expression in human skeletal muscle showed that alpha-cardiac actin mRNA is expressed at 5% the level of alpha-skeletal actin. Furthermore, the alpha-cardiac actin gene expressed in skeletal muscle is the same gene which produces alpha-cardiac actin mRNA in the human heart. Of equal surprise, we found that alpha-skeletal actin mRNA accounts for about half of the total actin mRNA in adult heart. Comparison of total actin mRNA levels in adult skeletal muscle and adult heart revealed that the steady-state levels in skeletal muscle are about twofold greater, per microgram of total cellular RNA, than those in heart. Thus, in skeletal muscle and in heart, both of the sarcomeric actin mRNA isotypes are quite abundant transcripts. We conclude that alpha-skeletal and alpha-cardiac actin genes are coexpressed as an actin pair in human adult striated muscles. Since the smooth-muscle actins (aortic and stomach) and the cytoplasmic actins (beta and gamma) are known to be coexpressed in smooth muscle and nonmuscle cells, respectively, we postulate that coexpression of actin pairs may be a common feature of mammalian actin gene expression in all tissues.

Localization of actin-related sequences by in situ hybridization to R-banded human chromosomes

Using a cytoplasmic actin cDNA probe we have localized a number of actin sequences in the human genome using a novel in situ hybridization technique. Metaphase chromosomes treated to produce R-bands were directly annealed with 125I-labeled actin probe. Under these conditions many regions of the genome were apparently denatured enough to be capable of hybridizing with the probe. Most of the actin sites detected in prior experiments using chromosome preparations, which had been completely denatured, were recognized in this experiment. The major advantage of this method over standard in situ hybridization techniques is the marked increase in the resolution of subregional localization.

Human actin genes are single copy for alpha-skeletal and alpha-cardiac actin but multicopy for beta- and gamma-cytoskeletal genes: 3' untranslated regions are isotype specific but are conserved in evolution

We have constructed isotype-specific subclones from the 3' untranslated regions of alpha-skeletal, alpha-cardiac, beta-cytoskeletal, and gamma-cytoskeletal actin cDNAs. These clones have been used as hybridization probes to assay the number and organization of these actin isotypes in the human genome. Hybridization of these probes to human genomic actin clones (Engel et al., Proc. Natl. Acad. Sci. U.S.A. 78:4674-4678, 1981; Engel et al., Mol. Cell. Biol. 2:674-684, 1982) has allowed the unambiguous assignment of the genomic clones to isotypically defined actin subfamilies. In addition, only one isotype-specific probe hybridizes to each actin-containing gene, with a single exception. This result suggests that the multiple actin genes in the human genome are not closely linked. Genomic DNA blots probed with these subclones under stringent conditions demonstrate that the alpha-skeletal and alpha-cardiac muscle actin genes are single copy, whereas the cytoskeletal actins, beta and gamma, are present in multiple copies in the human genome. Most of the actin genes of other mammals are cytoplasmic as well. These observations have important implications for the evolution of multigene families.

alpha-skeletal and alpha-cardiac actin genes are coexpressed in adult human skeletal muscle and heart

We determined the actin isotypes encoded by 30 actin cDNA clones previously isolated from an adult human muscle cDNA library. Using 3' untranslated region probes derived from alpha-skeletal, beta- and gamma-actin cDNAs and from an alpha-cardiac actin genomic clone, we showed that 28 of the cDNAs correspond to alpha-skeletal actin transcripts. Unexpectedly, however, the remaining two cDNA clones proved to derive from alpha-cardiac actin mRNA. Sequence analysis confirmed that the two skeletal muscle alpha-cardiac actin cDNAs are derived from transcripts of the cloned alpha-cardiac actin gene. Direct measurements of actin isotype mRNA expression in human skeletal muscle showed that alpha-cardiac actin mRNA is expressed at 5% the level of alpha-skeletal actin. Furthermore, the alpha-cardiac actin gene expressed in skeletal muscle is the same gene which produces alpha-cardiac actin mRNA in the human heart. Of equal surprise, we found that alpha-skeletal actin mRNA accounts for about half of the total actin mRNA in adult heart. Comparison of total actin mRNA levels in adult skeletal muscle and adult heart revealed that the steady-state levels in skeletal muscle are about twofold greater, per microgram of total cellular RNA, than those in heart. Thus, in skeletal muscle and in heart, both of the sarcomeric actin mRNA isotypes are quite abundant transcripts. We conclude that alpha-skeletal and alpha-cardiac actin genes are coexpressed as an actin pair in human adult striated muscles. Since the smooth-muscle actins (aortic and stomach) and the cytoplasmic actins (beta and gamma) are known to be coexpressed in smooth muscle and nonmuscle cells, respectively, we postulate that coexpression of actin pairs may be a common feature of mammalian actin gene expression in all tissues.

Human actin genes are single copy for alpha-skeletal and alpha-cardiac actin but multicopy for beta- and gamma-cytoskeletal genes: 3' untranslated regions are isotype specific but are conserved in evolution

We have constructed isotype-specific subclones from the 3' untranslated regions of alpha-skeletal, alpha-cardiac, beta-cytoskeletal, and gamma-cytoskeletal actin cDNAs. These clones have been used as hybridization probes to assay the number and organization of these actin isotypes in the human genome. Hybridization of these probes to human genomic actin clones (Engel et al., Proc. Natl. Acad. Sci. U.S.A. 78:4674-4678, 1981; Engel et al., Mol. Cell. Biol. 2:674-684, 1982) has allowed the unambiguous assignment of the genomic clones to isotypically defined actin subfamilies. In addition, only one isotype-specific probe hybridizes to each actin-containing gene, with a single exception. This result suggests that the multiple actin genes in the human genome are not closely linked. Genomic DNA blots probed with these subclones under stringent conditions demonstrate that the alpha-skeletal and alpha-cardiac muscle actin genes are single copy, whereas the cytoskeletal actins, beta and gamma, are present in multiple copies in the human genome. Most of the actin genes of other mammals are cytoplasmic as well. These observations have important implications for the evolution of multigene families.

Isolation and characterization of full-length cDNA clones for human alpha-, beta-, and gamma-actin mRNAs: skeletal but not cytoplasmic actins have an amino-terminal cysteine that is subsequently removed

cDNA clones encoding three classes of human actins have been isolated and characterized. The first two classes (gamma and beta, cytoplasmic actins) were obtained from a cDNA library constructed from simian virus 40-transformed human fibroblast mRNA, and the third class (alpha, muscle actin) was obtained from a cDNA library constructed from adult human muscle mRNA. A new approach was developed to enrich for full-length cDNAs. The human fibroblast cDNA plasmid library was linearized with restriction enzymes that did not cut the inserts of interest; it was then size-fractionated on gels, and the chimeric molecules of optimal length were selected for retransformation of bacteria. When the resulting clones were screened for actin-coding sequences it was found that some full-length cDNAs were enriched as much as 50- to 100-fold relative to the original frequency of full-length clones in the total library. Two types of clones were distinguished. One of these clones encodes gamma actin and contains 100 base pairs of 5' untranslated region, the entire protein coding region, and the 3' untranslated region. The second class encodes beta actin, and the longest such clone contains 45 base pairs of 5' untranslated region plus the remainder of the mRNA extending to the polyadenylic acid tail. A third class, obtained from the human muscle cDNA library, encodes alpha actin and contains 100 base pairs of 5' untranslated region, the entire coding region, and the 3' untranslated region. Analysis of the DNA sequences of the 5' end of the clones demonstrated that although beta- and gamma-actin genes start with a methionine codon (MET-Asp-Asp-Asp and MET-Glu-Glu-Glu, respectively), the alpha-actin gene starts with a methionine codon followed by a cysteine codon (MET-CYS-Asp-Glu-Asp-Glu). Since no known actin proteins start with a cysteine, it is likely that post-translational removal of cysteine in addition to methionine accompanies alpha-actin synthesis but not beta- and gamma-actin synthesis. This observation has interesting implications both for actin function and actin gene regulation and evolution.

Number and organization of actin-related sequences in the mouse genome

Recombinant plasmids containing cDNA sequences complementary to the two mouse striated-muscle actin messenger RNAs (pAF81, pAM91) and to a non-muscle actin mRNA (pAL41) have been used to examine the number and organization of actin-related sequences in the mouse genome. A large number (greater than 20) of actin-related sequences are detected on Southern blots of restricted mouse DNA, the majority of which hybridize to both the 5' and 3' ends of the actin-coding sequence, even under conditions revealing only sequences greater than 80% homologous to the actin cDNA probes. More stringent washing of these blots indicates that the two striated muscle actins are each encoded by single genes, and that a non-muscle (beta or gamma) actin cDNA detects one homologous and two closely related sequences in mouse DNA. The segregation of the two striated-muscle actin genes in recombinant inbred mouse strains shows that these genes are not closely linked (greater than 1 centimorgan), and that the skeletal muscle actin gene is not linked to a non-muscle actin gene. Screening a bank of mouse genomic DNA, cloned in Charon 4A, indicates that the number of actin-related sequences in the mouse genome is much higher than 20. In particular, five phages have been isolated representing part of a sub-family of 20 to 50 similar but non-identical sequences, only weakly homologous to actin cDNA probes (probably a family of actin pseudogenes), which are the result of a recent amplification of a greater than 17 X 10(3) base region of mouse DNA.

Isolation and characterization of cDNA clones for human skeletal muscle alpha actin

Two cDNA libraries corresponding to polyA+ RNA from human adult skeletal muscle have been constructed by cloning in the PstI site of pBR322. Skeletal alpha actin cDNA clones have been isolated and characterized. Three of these plasmids have overlapping inserts which together contain the complete 5' non-coding and protein-coding region and part of the 3' untranslated region. Determination of the sequence of the cloned cDNA confirms the complete conservation in human of the amino-acid sequence of skeletal alpha actin compared to the rabbit or rat proteins. The 5' untranslated region, but not the 3' untranslated region, shows good homology with the corresponding one in the rat gene. Analysis of changes at silent sites within the protein-coding region suggests that the divergence of skeletal and cardiac alpha actin took place much earlier than the mammalian radiation. The plasmids described here have been used as probes to detect the homologous gene among the about thirty actin sequences present in the human genome.

Isolation and characterization of full-length cDNA clones for human alpha-, beta-, and gamma-actin mRNAs: skeletal but not cytoplasmic actins have an amino-terminal cysteine that is subsequently removed

cDNA clones encoding three classes of human actins have been isolated and characterized. The first two classes (gamma and beta, cytoplasmic actins) were obtained from a cDNA library constructed from simian virus 40-transformed human fibroblast mRNA, and the third class (alpha, muscle actin) was obtained from a cDNA library constructed from adult human muscle mRNA. A new approach was developed to enrich for full-length cDNAs. The human fibroblast cDNA plasmid library was linearized with restriction enzymes that did not cut the inserts of interest; it was then size-fractionated on gels, and the chimeric molecules of optimal length were selected for retransformation of bacteria. When the resulting clones were screened for actin-coding sequences it was found that some full-length cDNAs were enriched as much as 50- to 100-fold relative to the original frequency of full-length clones in the total library. Two types of clones were distinguished. One of these clones encodes gamma actin and contains 100 base pairs of 5' untranslated region, the entire protein coding region, and the 3' untranslated region. The second class encodes beta actin, and the longest such clone contains 45 base pairs of 5' untranslated region plus the remainder of the mRNA extending to the polyadenylic acid tail. A third class, obtained from the human muscle cDNA library, encodes alpha actin and contains 100 base pairs of 5' untranslated region, the entire coding region, and the 3' untranslated region. Analysis of the DNA sequences of the 5' end of the clones demonstrated that although beta- and gamma-actin genes start with a methionine codon (MET-Asp-Asp-Asp and MET-Glu-Glu-Glu, respectively), the alpha-actin gene starts with a methionine codon followed by a cysteine codon (MET-CYS-Asp-Glu-Asp-Glu). Since no known actin proteins start with a cysteine, it is likely that post-translational removal of cysteine in addition to methionine accompanies alpha-actin synthesis but not beta- and gamma-actin synthesis. This observation has interesting implications both for actin function and actin gene regulation and evolution.