Determinants of translational fidelity and efficiency in vertebrate mRNAs

Translation from the 5' untranslated region (UTR) of mRNA 1 is repressed, but that from the 5' UTR of mRNA 7 is stimulated in coronavirus-infected cells

Viral gene products are generally required in widely differing amounts for successful virus growth and assembly. For coronaviruses, regulation of transcription is a major contributor to these differences, but regulation of translation may also be important. Here, we examine the possibility that the 5' untranslated regions (UTRs), unique for each of the nine species of mRNA in the bovine coronavirus and ranging in length from 70 nucleotides (nt) to 210 nt (inclusive of the common 5'-terminal 65-nt leader), can differentially affect the rate of protein accumulation. When the natural 77-nt 5' UTR on synthetic transcripts of mRNA 7 (mRNA for N and I proteins) was replaced with the 210-nt 5' UTR from mRNA 1 (genomic RNA, mRNA for viral polymerase), approximately twofold-less N, or (N) CAT fusion reporter protein, was made in vitro. Twofold less was also made in vivo in uninfected cells when a T7 RNA polymerase-driven transient-transfection system was used. In coronavirus-infected cells, this difference surprisingly became 12-fold as the result of both a stimulated translation from the 77-nt 5' UTR and a repression of translation from the 210-nt 5' UTR. These results reveal that a differential 5' UTR-directed regulation of translation can occur in coronavirus-infected cells and lead us to postulate that the direction and degree of regulation is carried out by viral or virally induced cellular factors acting in trans on cis-acting elements within the 5' UTR.

Coding region intron/exon organization, alternative splicing, and X-chromosome inactivation of the KRAB/FPB-domain-containing human zinc finger gene ZNF41

ZNF41 belongs to a cluster of human zinc finger genes residing within a gene-rich region at Xp11.23. ZNF41 encodes a KRAB/FPB (Krüppel-associated/finger preceding box) domain, a potent transcription repression motif present in hundreds of vertebrate zinc finger protein genes, composed of two protein modules, A and B. Three introns, placed at identical positions in paralogous genes, interrupt four exons encoding the ZNF41 N-terminal amino acids, the KRAB/FPB-A and KRAB/FPB-B modules, and the remaining coding region adjoined to the C-terminal zinc finger domain. Since the KRAB/FPB-A and KRAB/FPB-B modules are encoded by dedicated exons in ZNF41 and paralogous genes, exon skipping may lead to differential usage of these modules in alternative gene products. RT-PCR analysis of ZNF41 mRNAs showed that, while skipping of the KRAB/FPB-A and/or KRAB/FPB-B exons was not detected, the use of alternative donor/acceptor sites upstream of the KRAB/FPB-A exon generates multiple ZNF41 transcripts potentially encoding polypeptides differing in the N-terminal region and expressed in different tissues. The expression pattern in cell hybrids containing either active or inactive X chromosomes indicates that ZNF41, which resides within a region of the X chromosome that includes genes that are both subject to and escape X-inactivation, is susceptible to X-chromosome inactivation.

Upregulation of protein synthesis initiation factor eIF-4E is an early event during colon carcinogenesis

A general increase in protein synthesis and a specific increase in the synthesis of growth-promoting proteins are necessary for mitogenesis. Regulation of protein synthesis, as well as preferential translation of some mRNAs coding for growth promoting proteins (e.g. cyclin D1), involves the essential protein synthesis initiation factor eIF-4E. This factor is induced by various oncoproteins, and, when overexpressed, it can transform cultured cells. In this report we explore the roles of eIF-4E in human neoplastic disorders of the colon and in the regulation of general and specific protein synthesis. We find that eIF-4E is increased in colon adenomas and carcinomas, and this increase is accompanied in most but not all cases by elevation of cyclin D1 levels. While general protein synthesis is increased by eIF-4E overexpression in cultured cells, only a small proportion of proteins is preferentially upregulated by eIF-4E, as revealed by two-dimensional gel electrophoresis. These results are consistent with the view that eIF-4E plays a role in carcinogenesis by increasing general protein synthesis and by preferentially upregulating a subset of putative growth promoting proteins. Our results, taken together with the recent findings that c-myc transcription is negatively regulated by APC and our earlier data on transcriptional activation of eIF-4E expression by c-Myc suggest that eIF-4E is a downstream target of the APC/beta-catenin/Tcf-4 pathway, and is strongly involved in colon tumorigenesis.

Sequence and location of SIX3, a homeobox gene expressed in the human eye

Mouse Six3 is a homeobox gene expressed almost exclusively in the developing retina, lens, hypothalamus, and pituitary. It belongs to the same family as sine oculis, a Drosophila regulatory gene that encodes a transcription factor essential for eye development. The optix gene is its closest known Drosophila homologue, with a homeodomain that is 95% identical in sequence to the Six3 protein. We have isolated the homologous human gene, SIX3, which is expressed in the adult retina and encodes a 332 amino acid protein that is 98% identical to its mouse counterpart. The SIX3 protein coding region is interrupted by a single intron located just downstream of the homeobox. A surprising feature of the SIX3 gene is a 533 nucleotide 5' untranslated region that contains long polypyrimidine tracts with 96% identity to mouse Six3. We have used in-situ hybridization to map SIX3 to 2p21-p22, a site that is syntenic with the Six3 region of mouse chromosome 17. Large heterozygous deletions associated with human holoprosencephaly type 2 have been previously mapped to 2p21, opening the possibility that SIX3 could be involved in the development of midline structures of the head. Alternatively, the expression pattern of mouse Six3 suggests that human SIX3 could be involved in disorders of eye and pituitary development.

Parathyroid hormone-regulated production of stem cell factor in human osteoblasts and osteoblast-like cells

We investigated stem cell factor (SCF) expression in osteoblasts because mast cells, which occur ectopically in hyperparathyroid bone, are induced by SCF. Nontransformed osteoblasts and Saos2 or MG63 cells expressed SCF in response to PTH. Western analysis showed only large, cell-associated isoforms, Mrs approximately 40-48 kD. Transfection of MG63 cells with plasmids expressing antisense SCF mRNA eliminated immunoreactive SCF. Sequencing osteoblast SCF cDNAs showed that exon 6 was omitted. mRNAs without exon 6 produce membrane-associated SCF isoforms in rodents, suggesting that human SCFs are processed similarly. The major osteoblastic SCF mRNA, approximately 5 kB, was augmented by PTH. Neither protein or mRNA was increased by vitamin D, however, 6-7 kB transcripts were predominant in other tissues but not detectable in osteoblasts. We conclude that osteoblasts express SCF in response to PTH, with mRNA and protein processing differences relative to other cells. SCF stimulates osteoclasts, suggesting that PTH-induced osteoblastic SCF functions to accelerate bone turnover. Mast cells may occur due to SCF overexpression at extreme PTH levels.

Position Effects of Variable Region Carbohydrate on the Affinity and In Vivo Behavior of an Anti-(1→6) Dextran Antibody

IgG is a glycoprotein with an N-linked carbohydrate structure attached to the CH2 domain of each of its heavy chains. In addition, the variable regions of IgG often contain potential N-linked carbohydrate addition sequences that frequently result in the attachment of V region carbohydrate. Nonetheless, the precise role of this V region glycan remains unclear. Studies from our laboratory have shown that a naturally occurring somatic mutant of an anti-dextran Ab that results in a carbohydrate addition site at Asn58 of the VH has carbohydrate in the complementarity-determining region 2 (CDR2) of the VH, and the presence of carbohydrate leads to an increase in affinity. However, carbohydrate attached to nearby positions within CDR2 had variable affects on affinity. In the present work we have extended these studies by adding carbohydrate addition sites close to or within all the CDRs of the same anti-dextran Ab. We find that carbohydrate is attached to all the novel addition sites, but the extent of glycosylation varies with the position of the site. In addition, we find that the position of the variable region carbohydrate influences some functional properties of the Ab, including those usually associated with the V region such as affinity for Ag as well as other characteristics typically attributed to the Fc such as half-life and organ targeting. These studies suggest that modification of variable region glycosylation provides an alternate strategy for manipulating the functional attributes of the Ab molecule and may shed light on how changes in carbohydrate structure affect protein conformation.

Comparative analyses of the secondary structures of synthetic and intracellular yeast MFA2 mRNAs

The overall folded (global) structure of mRNA may be critical to translation and turnover control mechanisms, but it has received little experimental attention. Presented here is a comparative analysis of the basic features of the global secondary structure of a synthetic mRNA and the same intracellular eukaryotic mRNA by dimethyl sulfate (DMS) structure probing. Synthetic MFA2 mRNA of Saccharomyces cerevisiae first was examined by using both enzymes and chemical reagents to determine single-stranded and hybridized regions; RNAs with and without a poly(A) tail were compared. A folding pattern was obtained with the aid of the MFOLD program package that identified the model that best satisfied the probing data. A long-range structural interaction involving the 5' and 3' untranslated regions and causing a juxtaposition of the ends of the RNA, was examined further by a useful technique involving oligo(dT)-cellulose chromatography and antisense oligonucleotides. DMS chemical probing of A and C nucleotides of intracellular MFA2 mRNA was then done. The modification data support a very similar intracellular structure. When low reactivity of A and C residues is found in the synthetic RNA, approximately 70% of the same sites are relatively more resistant to DMS modification in vivo. A slightly higher sensitivity to DMS is found in vivo for some of the A and C nucleotides predicted to be hybridized from the synthetic structural model. With this small mRNA, the translation process and mRNA-binding proteins do not block DMS modifications, and all A and C nucleotides are modified the same or more strongly than with the synthetic RNA.

Eukaryotic mRNAs encoding abundant and scarce proteins are statistically dissimilar in many structural features

It is well known that non-coding mRNA sequences are dissimilar in many structural features. For individual mRNAs correlations were found for some of these features and their translational efficiency. However, no systematic statistical analysis was undertaken to relate protein abundance and structural characteristics of mRNA encoding the given protein. We have demonstrated that structural and contextual features of eukaryotic mRNAs encoding high- and low-abundant proteins differ in the 5' untranslated regions (UTR). Statistically, 5' UTRs of low-expression mRNAs are longer, their guanine plus cytosine content is higher, they have a less optimal context of the translation initiation codons of the main open reading frames and contain more frequently upstream AUG than 5' UTRs of high-expression mRNAs. Apart from the differences in 5' UTRs, high-expression mRNAs contain stronger termination signals. Structural features of low- and high-expression mRNAs are likely to contribute to the yield of their protein products.

Effect of short 5' UTRs on protein synthesis in two biological kingdoms

Efficient ribosomal protein synthesis is dependent on cis-acting elements in the 5' untranslated region (UTR) of mRNAs. Between prokaryotes and eukaryotes, the sequence and location of these elements differ to the extent of not being functionally interchangeable. We explored the possibility of constructing bifunctional UTRs that could direct translation in both prokaryotes and eukaryotes. A variant of a UTR from ner of phage Mu (ner-ACC) enhanced protein synthesis in a rabbit reticulocyte lysate, and it was compared to a lacZ-CTA, containing the lambda cro RBS and the Escherichia coli lacZ spacer. Several mutants in the -3 to -1 regions of both lacZ-CTA and ner-ACC were tested in rabbit reticulocyte lysate and E. coli to select UTRs that were optimized simultaneously for both biological kingdoms. The lacZ-ATC proved 217-fold more effective than ner-ACC in this cross-species ability to enhance translation. The lacZ-ACC and ner-ATC were 83- and 78-fold, respectively, better than ner-ACC. We conclude that short UTRs (12-15 nt in length) can be fine-tuned in the -9 to -1 regions to enhance protein synthesis concurrently in prokaryotes and eukaryotes. In related studies, we show that nt at the -3 to -1 region of mRNAs exert an enormous impact on synthesis of proteins in E. coli.

HSP101 functions as a specific translational regulatory protein whose activity is regulated by nutrient status

The 5' leader (Omega) of tobacco mosaic viral RNA functions as a translational enhancer. Sequence analysis of a 102-kD protein, identified previously as a specific Omega RNA-binding protein, revealed homology to the HSP101/HSP104/ClpB family of heat shock proteins and its expression in yeast complemented a thermotolerance defect caused by a deletion of the HSP104 gene. Up to a 50-fold increase in the translation of Omega-luc, but not luc mRNA was observed in yeast expressing the tobacco HSP101 whereas Omega failed to enhance translation in the absence of HSP101. Therefore, HSP101 and Omega comprise a two-component translational regulatory mechanism that can be recapitulated in yeast. Analysis of HSP101 function in yeast translation mutants suggested that the initiation factor (eIF) 3 and specifically one (TIF4632) of the two eIF4G proteins were required for the HSP101-mediated enhancement. The RNA-binding and translational regulatory activities of HSP101 were inactive in respiring cells or in cells subject to nutrient limitation, but its thermotolerance function remained unaffected. This is the first identification of a protein required for specific translational enhancement of capped mRNAs, the first report of a translational regulatory function for any heat-shock protein, and the first functional distinction between the two eIF4G proteins present in eukaryotes.

A Breast and Melanoma-Shared Tumor Antigen: T Cell Responses to Antigenic Peptides Translated from Different Open Reading Frames

Infusion of TIL586 along with IL-2 into the autologous patient with metastatic melanoma resulted in the objective regression of tumor. Here, we report that screening a cDNA library from the 586mel cell line using CTL clones derived from TIL586 resulted in the isolation of a gene, CAG-3 (cancer Ag gene 3). Sequence analysis revealed that CAG-3 encodes an open reading frame identical to NY-ESO-1, which was recently reported to be recognized by autologous serum from a patient with esophageal cancer. Thus, NY-ESO-1 appears to be an immune target for both Ab- and T cell-mediated responses. Significantly, NY-ESO-1-specific CTL clones were capable of recognizing two HLA-A31-positive fresh and cultured breast tumors. To our knowledge, this represents the first direct demonstration that tumor-specific CTL clones can recognize both breast and melanoma tumor cells. A 10-mer antigenic peptide ESO10–53 (ASGPGGGAPR) was identified from the normal open reading frame of NY-ESO-1 based on its ability to sensitize HLA-A31-positive target cells for cytokine release and specific lysis. Interestingly, two additional CTL clones that were sensitized with NY-ESO-1 recognized two overlapping antigenic peptides derived from an alternative open reading frame of the same gene. These findings indicate that CTLs simultaneously responded to two different gene products translated from the normal and alternative reading frames of the same gene. Understanding of this mechanism by which the alternative reading frame is translated may have important implications in tumor immunology.

An avian cDNA encoding a tyrosine-phosphorylated protein with PDZ, coiled-coil, and SAM domains

Tyrosine phosphoproteins of size 115-120 kDa were purified from membranes of chicken embryo fibroblasts (CEF) infected with Rous sarcoma virus (RSV). A mouse was immunized with these proteins, and the immune serum was used to screen a CEF cDNA expression library. A highly immunoreactive clone (KS5) was identified and characterized. The cDNA of this clone is 2.3 kb in length with a short 5' UTR and a single major open reading frame (ORF) encoding a polypeptide of 719 amino acids, with a calculated molecular weight of 81.1 kDa. The encoded protein contains an amino terminal PDZ domain, followed by a predicted coiled-coil region, a PEST domain, and a carboxy-terminal SAM domain. Consensus sequence motifs for tyrosine phosphorylation are also present, as are consensus sequences for the binding of SH2 and PDZ domains. Antisera from mice immunized with bacterially expressed fragments of the KS5 protein recognized proteins of size 230, 116, and 65 kDa in CEF. In other chicken embryo tissues, a 116-kDa species was the predominant protein recognized. The 116-kDa species is tyrosine-phosphorylated in RSV-CEF. The presence of PDZ and SAM domains in the KS5 protein suggests that it may act as a molecular adaptor, promoting and relaying information in a signal transduction pathway. It is a member of a family of related proteins, all of which have a highly conserved PDZ domain adjacent to a coiled-coil region. Two other members of this family are the neuronal proteins spinophilin (Allen, P.B., Ouimet, C.C., Greengard, P., 1997. Spinophilin, a novel protein phosphatase 1 binding protein localized to dendritic spines. Proc. Natl. Acad. Sci. USA 94, 9956-9961) and neurabin (Nakanishi, H., Obaishi, H., Satoh, A., Wada, M., Mandai, K., Satoh, K., Nishioka, H., Matsuura, Y., Mizoguchi, A. , Takai, Y., 1997. Neurabin: A novel neural tissue-specific actin filament-binding protein involved in neurite formation. J. Cell Biol. 139, 951-961).

The optx2 homeobox gene is expressed in early precursors of the eye and activates retina-specific genes

Vertebrate eye development begins at the gastrula stage, when a region known as the eye field acquires the capacity to generate retina and lens. Optx2, a homeobox gene of the sine oculis-Six family, is selectively expressed in this early eye field and later in the lens placode and optic vesicle. The distal and ventral portion of the optic vesicle are fated to become the retina and optic nerve, whereas the dorsal portion eventually loses its neural characteristics and activates the synthesis of melanin, forming the retinal pigment epithelium. Optx2 expression is turned off in the future pigment epithelium but remains expressed in the proliferating neuroblasts and differentiating cells of the neural retina. When an Optx2-expressing plasmid is transfected into embryonic or mature chicken pigment epithelial cells, these cells adopt a neuronal morphology and express markers characteristic of developing neural retina and photoreceptors. One explanation of these results is that Optx2 functions as a determinant of retinal precursors and that it has induced the transdifferentiation of pigment epithelium into retinal neurons and photoreceptors. We also have isolated optix, a Drosophila gene that is the closest insect homologue of Optx2 and Six3. Optix is expressed during early development of the fly head and eye primordia.

Two birds with one stone: genes that encode products targeted to two or more compartments

Eukaryotic cells are divided into multiple membrane-bound compartments, all of which contain proteins. A large subset of these proteins perform functions that are required in more than one compartment. Although in most cases proteins carrying out the same function in different compartments are encoded by different genes, this is not always true. Numerous examples have now been found where a single gene encodes proteins (or RNAs) found in two (or more) cell organelles or membrane systems. Some particularly clear examples come from protein synthesis itself: plant cells contain three protein-synthesizing compartments, the cytosol, the mitochondrial matrix and the plastid stroma. All three compartments thus require tRNAs and aminoacyl-tRNA synthetases. Some mitochondrial tRNAs and their aminoacyl-tRNA synthetases are identical to their cytosolic counterparts and they are encoded by the same genes. Similarly, some mitochondrial and plastid aminoacyl-tRNA synthetases are encoded by the same nuclear genes. The various ways in which differentially targeted products can be generated from single genes is discussed.

Tetracycline-regulated suppression of amber codons in mammalian cells

As an approach to inducible suppression of nonsense mutations in mammalian cells, we described recently an amber suppression system in mammalian cells dependent on coexpression of Escherichia coli glutaminyl-tRNA synthetase (GlnRS) along with the E. coli glutamine-inserting amber suppressor tRNA. Here, we report on tetracycline-regulated expression of the E. coli GlnRS gene and, thereby, tetracycline-regulated suppression of amber codons in mammalian HeLa and COS-1 cells. The E. coli GlnRS coding sequence attached to a minimal mammalian cell promoter was placed downstream of seven tandem tetracycline operator sequences. Cotransfection of HeLa cell lines expressing a tetracycline transactivator protein, carrying a tetracycline repressor domain linked to part of a herpesvirus VP16 activation domain, with the E. coli GlnRS gene and the E. coli glutamine-inserting amber suppressor tRNA gene resulted in suppression of the amber codon in a reporter chloramphenicol acetyltransferase gene. The tetracycline transactivator-mediated expression of E. coli GlnRS was essentially completely blocked in HeLa or COS-1 cells grown in the presence of tetracycline. Concomitantly, both aminoacylation of the suppressor tRNA and suppression of the amber codon were reduced significantly in the presence of tetracycline.

Molecular basis of non-responsiveness to peroxisome proliferators: the guinea-pig PPARalpha is functional and mediates peroxisome proliferator-induced hypolipidaemia

The guinea pig does not undergo peroxisome proliferation in response to peroxisome proliferators, in contrast with other rodents. To understand the molecular basis of this phenotype, the peroxisome proliferator activated receptor alpha (PPARalpha) from guinea-pig liver was cloned; it encodes a protein of 467 amino acid residues that is similar to rodent and human PPARalpha. The guinea-pig PPARalpha showed a high substitution rate: maximum likelihood analysis was consistent with rodent monophyly, but could not exclude rodent polyphyly (P approximately 0.06). The guinea-pig PPARalpha cDNA was expressed in 293 cells and mediated the induction of the luciferase reporter gene by the peroxisome proliferator, Wy-14,643, dependent on the presence of a peroxisome proliferator response element. Moreover the PPARalpha RNA and protein were expressed in guinea-pig liver, although at lower levels than in a species which is responsive to peroxisome proliferators, the mouse. To determine whether the guinea-pig PPARalpha mediated any physiological effects, guinea pigs were exposed to two selective PPARalpha agonists, Wy-14, 643 and methylclofenapate; both compounds induced hypolipidaemia. Thus the guinea pig is a useful model for human responses to peroxisome proliferators.

The Arabidopsis thaliana cDNAs coding for eIF4E and eIF(iso)4E are not functionally equivalent for yeast complementation and are differentially expressed during plant development

Two cDNAs (At.EIF4E1 and At.EIF4E2) encoding, respectively, the eukaryotic initiation factors eIF4E and eIF(iso)4E of Arabidopsis thaliana were isolated by complementation of a Saccharomyces cerevisiae conditional mutant. The deduced amino acid sequences of the proteins are homologous to those from monocotyledonous plants, yeast and mammals. The corresponding genes were identified in YAC clones mapping to chromosome IV (At.EIF4E1) and to chromosome V (At.EIF4E2). The yeast strain complemented by At.EIF4E2 grew poorly compared with an isogenic strain expressing At.EIF4E1. Northern and in situ hybridization analysis show that both Arabidopsis At.EIF4E1 and At.EIF4E2 mRNAs are differentially accumulated in plant tissues. The At.EIF4E1 mRNA is expressed in all tissues except in the cells of the specialization zone of the roots; the At.EIF4E2 mRNA is particularly abundant in floral organs and in young developing tissues. This work further demonstrates an association between a high level of EIF4E mRNAs and cell proliferation and suggests that the plant eIF4E isoforms may have distinct functions in cell development and metabolism.

Nucleosomal DNA regulates the core-histone-binding subunit of the human Hat1 acetyltransferase

BACKGROUND

In eukaryotic cells, newly synthesized histone H4 is acetylated at lysines 5 and 12, a transient modification erased by deacetylases shortly after deposition of histones into chromosomes. Genetic studies in Saccharomyces cerevisiae revealed that acetylation of newly synthesized histones H3 and H4 is likely to be important for maintaining cell viability; the precise biochemical function of this acetylation is not known, however. The identification of enzymes mediating site-specific acetylation of H4 at Lys5 and Lys12 may help explain the function of the acetylation of newly synthesized histones.

RESULTS

A cDNA encoding the catalytic subunit of the human Hat1 acetyltransferase was cloned and, using specific antibodies, the Hat1 holoenzyme was purified from human 293 cells. The human enzyme acetylates soluble but not nucleosomal H4 at Lys5 and Lys12 and acetylates histone H2A at Lys5. Unexpectedly, we found Hat1 in the nucleus of S-phase cells. Like its yeast counterpart, the human holoenzyme consists of two subunits: a catalytic subunit, Hat1, and a subunit that binds core histones, p46, which greatly stimulates the acetyltransferase activity of Hat1. Both p46 and the highly related p48 polypeptide (the small subunit of human chromatin assembly factor 1; CAF-1) bind directly to helix 1 of histone H4, a region that is not accessible when H4 is in chromatin.

CONCLUSIONS

We suggest that p46 and p48 are core-histone-binding subunits that target chromatin assembly factors, chromatin remodeling factors, histone acetyltransferases and histone deacetylases to their histone substrates in a manner that is regulated by nucleosomal DNA.

In vitro suppression as a tool for the investigation of translation initiation

An in vitro protein synthesizing system that employs rabbit reticulocyte lysates has been employed for protein production from mRNAs containing nonsense (UAG) codons in the presence of misacylated suppressor tRNAs.The system includes a misacylated Escherichia coli tRNAAlaCUA that functions at least as efficiently as any suppressor tRNA transcript reported to date and which has been shown not to be a substrate for (re)activation by alanyl-tRNA synthetase. Application of the optimized system for preparation of dihydrofolate analogs has also permitted analysis of competing mechanisms that control the sites(s) of translation initiation.

hSK4, a member of a novel subfamily of calcium-activated potassium channels

The gene for hSK4, a novel human small conductance calcium-activated potassium channel, or SK channel, has been identified and expressed in Chinese hamster ovary cells. In physiological saline hSK4 generates a conductance of approximately 12 pS, a value in close agreement with that of other cloned SK channels. Like other members of this family, the polypeptide encoded by hSK4 contains a previously unnoted leucine zipper-like domain in its C terminus of unknown function. hSK4 appears unique, however, in its very high affinity for Ca2+ (EC50 of 95 nM) and its predominant expression in nonexcitable tissues of adult animals. Together with the relatively low homology of hSK4 to other SK channel polypeptides (approximately 40% identical), these data suggest that hSK4 belongs to a novel subfamily of SK channels.

A Xenopus zinc finger protein that specifically binds dsRNA and RNA-DNA hybrids

Proteins containing C2H2 type zinc finger motifs represent one of the largest classes of nucleic acid-binding proteins found in nature. We describe a novel zinc finger protein, dsRBP-ZFa, isolated by screening an expression library with dsRNA. The dsRBP-ZFa cDNA encodes a protein containing seven zinc finger motifs and an acidic C-terminal domain. Mobility shift experiments demonstrate that dsRBP-ZFa binds dsRNA and RNA-DNA hybrids with nanomolar dissociation constants and in a sequence independent manner. We also show that DNA and single stranded RNA fail to compete with dsRNA for binding suggesting dsRBP-ZFa prefers to bind an A-form helix. Using western analyses we have localized dsRBP-ZFa primarily to the nucleus of Xenopus laevis oocytes. The identification of dsRBP-ZFa provides the first example of a zinc finger protein that is specific for dsRNA. In addition, dsRBP-ZFa does not contain the previously described dsRNA binding motif, suggesting certain zinc fingers may provide an alternative way to recognize the A-form helix.

The position dependence of translational regulation via RNA-RNA and RNA-protein interactions in the 5'-untranslated region of eukaryotic mRNA is a function of the thermodynamic competence of 40 S ribosomes in translational initiation

Cap proximity is a requirement to enable secondary structures and RNA-binding proteins to repress translational initiation via the 5'-untranslated region (5'-UTR) of mammalian mRNAs. We show that in Saccharomyces cerevisiae, unlike mammalian cells, the in vitro translational repressive effect of the mammalian iron regulatory protein 1 (IRP1) is independent of the site of its target in the 5'-UTR, the iron-responsive element (IRE). In vitro studies demonstrate that the binding affinity of IRP1 is also unaffected by the position of the IRE. Using IRE loop mutants, we observe an almost complete loss of IRP1-dependent repression in yeast concomitant with a 150-fold reduction in binding affinity for the IRE target. This mirrors the natural quantitative range of iron-induced adjustment of IRE/IRP1 affinity in mammalian cells. By enhancing the stability of the IRE stem-loop, we also show that its intrinsic folding energy acts together with the binding energy of IRP1 to give an additive capacity to restrict translational initiation. An IRE.IRP1 complex in a cap-distal position in yeast blocks scanning 40 S ribosomes on the 5'-UTR. It follows that the position effect of mammalian site-specific translational repression is dictated by the competence of the mammalian preinitiation complex to destabilize inhibitory structures at different steps of the initiation process.

Sequence and RT-PCR expression analysis of two peroxidases from Arabidopsis thaliana belonging to a novel evolutionary branch of plant peroxidases

cDNA clones encoding two new Arabidopsis thaliana peroxidases, ATP 1a and ATP 2a, have been identified by searching the Arabidopsis database of expressed sequence tags (dbEST). They represent a novel branch of hitherto uncharacterized plant peroxidases which is only 35% identical in amino acid sequence to the well characterized group of basic plant peroxidases represented by the horseradish (Armoracia rusticana) isoperoxidases HRP C, HRP E5 and the similar Arabidopsis isoperoxidases ATP Ca, ATP Cb, and ATP Ea. However ATP 1a is 87% identical in amino acid sequence to a peroxidase encoded by an mRNA isolated from cotton (Gossypium hirsutum). As cotton and Arabidopsis belong to rather diverse families (Malvaceae and Crucifereae, respectively), in contrast with Arabidopsis and horseradish (both Crucifereae), the high degree of sequence identity indicates that this novel type of peroxidase, albeit of unknown function, is likely to be widespread in plant species. The atp 1 and atp 2 types of cDNA sequences were the most redundant among the 28 different isoperoxidases identified among about 200 peroxidase encoding ESTs. Interestingly, 8 out of totally 38 EST sequences coding for ATP 1 showed three identical nucleotide substitutions. This variant form is designated ATP 1b. Similarly, six out of totally 16 EST sequences coding for ATP 2 showed a number of deletions and nucleotide changes. This variant form is designated ATP 2b. The selected EST clones are full-length and contain coding regions of 993 nucleotides for atp 1a, and 984 nucleotides for atp 2a. These regions show 61% DNA sequence identity. The predicted mature proteins ATP 1a, and ATP 2a are 57% identical in sequence and contain the structurally and functionally important residues, characteristic of the plant peroxidase superfamily. However, they do show two differences of importance to peroxidase catalysis: (1) the asparagine residue linked with the active site distal histidine via hydrogen bonding is absent; (2) an N-glycosylation site is located right at the entrance to the heme channel. The reverse transcriptase polymerase chain reaction (RT-PCR) was used to identify mRNAs coding for ATP 1a/b and ATP 2a/b in germinating seeds, seedlings, roots, leaves, stems, flowers and cell suspension culture using elongation factor 1alpha (EF-1alpha) for the first time as a positive control. Both mRNAs were transcribed at levels comparable to EF-1alpha in all plant tissues investigated which were more than two days old, and in cell suspension culture. In addition, the mRNA coding for ATP 1a/b was found in two day old germinating seeds. The abundant transcription of ATP 1a/b and ATP 2a/b is in line with their many entries in dbEST, and indicates essential roles for these novel peroxidases.

Cloning and characterization of HUPF1, a human homolog of the Saccharomyces cerevisiae nonsense mRNA-reducing UPF1 protein

Levels of most nonsense mRNAs are normally reduced in prokaryotes and eukaryotes when compared with that of corresponding functional mRNAs. Genes encoding polypeptides that selectively reduce levels of nonsense mRNA have so far only been identified in simple eukaryotes. We have now cloned a human cDNA whose deduced amino acid sequence shows the highest degree of homology to that of UPF1, a bona fide Saccharomyces cerevisiae group I RNA helicase required for accelerated degradation of nonsense mRNA. Based on the total sequence of the shorter yeast UPF1 protein, the overall identity between the human protein and UPF1 is 51%. Besides NTPase and other RNA helicase consensus motifs, UPF1 and its human homolog also share similar putative zinc finger motifs that are absent in other group I RNA helicases. Northern blot analysis with the human cDNA probe revealed two transcripts in several human cell lines. Further, antibodies raised against a synthetic peptide of the human polypeptide detected a single 130 kDa polypeptide on Western blots from human and mouse cells. Finally, immunofluorescence and Western blot analyses revealed that the human and mouse polypeptides, like yeast UPF1, are expressed in the cytoplasm, but not in the nucleus. We have thus identified the first mammalian homolog of yeast UPF1, a protein that regulates levels of nonsense mRNA, and we tentatively name this protein human HUPF1 (for human homolog of UPF1).

An orphan nuclear receptor lacking a zinc-finger DNA-binding domain: interaction with several nuclear receptors

The yeast two-hybrid screening was applied to cloning cDNAs of proteins that interact with peroxisome proliferator-activated receptor alpha (PPAR alpha). We obtained from a rat liver cDNA library a clone encoding a protein related to the ligand-binding domain of the members of nuclear hormone receptor superfamily, whereas apparently lacking the zinc-finger DNA-binding domain. This protein interacted with the activated forms of several nuclear receptors, and thus is a novel type of heterodimer-forming nuclear receptor.

Mouse RAD50 has limited epitopic homology to p53 and is expressed in the adult myocardium

Previous studies have identified a 180-kDa mouse cardiomyocyte phosphoprotein with limited epitopic homology to p53. In this study, the protein was purified and partially sequenced. Oligonucleotide probes based on the available amino acid sequence data were used to isolate cDNA clones. Sequence analyses revealed that the clones encoded a protein with regional homology to the yeast RAD50 gene product. Expression of the mouse cDNA rescued the methyl methanesulfonate-sensitive phenotype in rad50 mutant yeast, indicating that the cardiomyocyte phosphoprotein is the mammalian homologue of the yeast RAD50 gene product. Fluorescence in situ hybridization analyses localized the mouse RAD50 gene to the A5-B1 region of chromosome 11. Northern blot analyses demonstrated a complex pattern of RAD50 expression during mouse development which was further complicated by the presence of several alternatively spliced transcripts. High levels of RAD50 expression was evident in the adult myocardium, a somewhat surprising observation given the absence of DNA synthesis in adult cardiomyocytes.

Cloning of a third mammalian Na+-Ca2+ exchanger, NCX3

NCX3 is the third isoform of a mammalian Na+-Ca2+ exchanger to be cloned. NCX3 was identified from rat brain cDNA by polymerase chain reaction (PCR) using degenerate primers derived from the sequences of two conserved regions of NCX1 and NCX2. The NCX3 PCR product was used to isolate two overlapping clones totalling 4.8 kilobases (kb) from a rat brain cDNA library. The overlapping clones were sequenced and joined at a unique Bsp106I restriction enzyme site to form a full-length cDNA clone. The NCX3 cDNA clone has an open reading frame of 2.8 kb encoding a protein of 927 amino acids. At the amino acid level, NCX3 shares 73% identity with NCX1 and 75% identity with NCX2 and is predicted to share the same membrane topology as NCX1 and NCX2. Following addition of a poly(A)+ tail to the NCX3 clone, exchanger activity could be expressed in Xenopus oocytes. NCX3 was also expressed in the mammalian BHK cell line. NCX3 transcripts are 6 kb in size and are highly restricted to brain and skeletal muscle. Linkage analysis in the mouse indicated that the NCX family of genes is dispersed, since the NCX1, NCX2, and NCX3 genes mapped to mouse chromosomes 17, 7, and 12, respectively.

Translation in plants--rules and exceptions

Translation processes in plants are very similar to those in other eukaryotic organisms and can in general be explained with the scanning model. Particularly among plant viruses, unconventional mRNAs are frequent, which use modulated translation processes for their expression: leaky scanning, translational stop codon readthrough or frameshifting, and transactivation by virus-encoded proteins are used to translate polycistronic mRNAs; leader and trailer sequences confer (cap-independent) efficient ribosome binding, usually in an end-dependent mechanism, but true internal ribosome entry may occur as well; in a ribosome shunt, sequences within an RNA can be bypassed by scanning ribosomes. Translation in plant cells is regulated under conditions of stress and during development, but the underlying molecular mechanisms have not yet been determined. Only a small number of plant mRNAs, whose structure suggests that they might require some unusual translation mechanisms, have been described.

Both 5' and 3' sequences of maize adh1 mRNA are required for enhanced translation under low-oxygen conditions

Alcohol dehydrogenase-1 (ADH1) synthesis in O2-deprived roots of maize (Zea mays L.) results from induced transcription and selective translation of ADH1 mRNA. The effect of ADH1 mRNA sequences on message stability and translation was studied in protoplasts of the maize cell line P3377.5' capped and 3' polyadenylated mRNA constructs containing the firefly gene (luc) for luciferase (LUC) or the Escherichia coli gene (uidA) for beta-glucuronidase (GUS) coding region were synthesized in vitro and electroporated into protoplasts that were cultured at 40 or 5% O2. A LUC mRNA with a 17-nucleotide polylinker 5' untranslated region (UTR) was expressed 10-fold higher under aerobic conditions than under hypoxic conditions. Expression of five chimeric ADH1-GUS mRNAs was measured relative to this LUC mRNA. An mRNA containing the 5'-UTR and the first 18 codons of adh1 in a translational fusion with the GUS coding region and followed by the 3'-UTR of adh1 was expressed 57-fold higher at 5% O2. Progressive deletion of adh1 5'-UTR and coding sequences reduced expression of the GUS-mRNA at 5% O2, but had little impact on expression of 40% O2. Enhancement of expression in hypoxic protoplasts conferred by the adh1 5'-UTR and the first 26 codons decreased more than 3-fold when the adh1 3'-UTR was removed. In addition, the adh1 3'-UTR slightly inhibited expression in aerobic protoplasts. The physical half-lives of the GUS and LUC mRNAs were similar under both anaerobic and hypoxic conditions, indicating that expression levels were largely independent of mRNA stability. Thus, both adh1 5' and 3' mRNA sequences are required for enhanced translation in protoplasts under O2 deprivation.

Gene expression from viral RNA genomes

This review is centered on the major strategies used by plant RNA viruses to produce the proteins required for virus multiplication. The strategies at the level of transcription presented here are synthesis of mRNA or subgenomic RNAs from viral RNA templates, and 'cap-snatching'. At the level of translation, several strategies have been evolved by viruses at the steps of initiation, elongation and termination. At the initiation step, the classical scanning mode is the most frequent strategy employed by viruses; however in a vast number of cases, leaky scanning of the initiation complex allows expression of more than one protein from the same RNA sequence. During elongation, frameshift allows the formation of two proteins differing in their carboxy terminus. At the termination step, suppression of termination produces a protein with an elongated carboxy terminus. The last strategy that will be described is co- and/or post-translational cleavage of a polyprotein precursor by virally encoded proteinases. Most (+)-stranded RNA viruses utilize a combination of various strategies.

Genomic and cDNA structures of the gene encoding the chicken ZF5 DNA binding protein

Genomic and cDNA clones encoding the chicken ZF5 transcription factor were isolated and sequenced. Conceptual translation of the cDNAs indicates that chicken ZF5 has five zinc fingers at its C-terminal end and a POZ/BTB protein-interaction domain at its N-terminal end. These two functional domains are more than 99% identical in amino acid sequence between the chicken and mouse proteins, whereas the region separating the POZ/BTB and DNA binding domains is only 74% identical. The entire 458 amino acid open reading frame is contained within a single exon, except for the initiator methionine codon which alone is supplied from an upstream exon by mRNA splicing.

The testis isoform of the phosphorylase kinase catalytic subunit (PhK-gammaT) plays a critical role in regulation of glycogen mobilization in developing lung

In order to identify the form of phosphorylase kinase catalytic subunit expressed in developing lung, degenerate polymerase chain reaction primers were designed based on conserved domains of the two known catalytic subunits, expressed primarily in muscle and testis. Amplification of cDNA from day 19 fetal rat lung followed by cloning and sequence analyses indicated that only the testis isoform of phosphorylase kinase (PhK-gammaT) was detectable in fetal lung. In situ hybridization analyses indicated that expression of PhK-gammaT RNA in developing lung tissue was widespread and not restricted to Type II epithelial cells; PhK-gammaT protein expression was temporally and spatially correlated with expression of PhK-gammaT RNA. PhK-gammaT RNA and protein expression was also characterized in the PhK-deficient glycogen storage disease (gsd) rat. PhK-gammaT RNA levels were similar in Type II cells isolated from wild type and gsd/gsd fetuses; in contrast, PhK-gammaT protein was virtually undetectable in gsd/ gsd Type II cells and enzyme activity was very low. These results suggest that PhK-gammaT plays a critical role in mobilization of glycogen during fetal lung development and that failure to catabolize glycogen in the gsd/gsd rat is related to an untranslatable PhK-gammaT RNA or unstable protein.

De novo expression of transfected human class 1 aldehyde dehydrogenase (ALDH) causes resistance to oxazaphosphorine anti-cancer alkylating agents in hamster V79 cell lines. Elevated class 1 ALDH activity is closely correlated with reduction in DNA interstrand cross-linking and lethality

Human class 1 aldehyde dehydrogenase (hALDH-1) can oxidize aldophosphamide, a key aldehyde intermediate in the activation pathway of cyclophosphamide and other oxazaphosphorine (OAP) anti-cancer alkylating agents. Overexpression of class 1 ALDH (ALDH-1) has been observed in cells selected for survival in the presence of OAPs. We used transfection to induce de novo expression of human ALDH-1 in V79/SD1 Chinese hamster cells to clearly quantitate the role of hALDH-1 expression in OAP resistance. Messenger RNA levels correlated well with hALDH-1 protein levels and enzyme activities (1.5-13.6 milliunits/mg with propionaldehyde/NAD+ substrate, compared to < 1 milliunit/mg in controls) in individual clonal transfectant lines, and slot blot analysis confirmed the presence of the transfected cDNA. Expressed ALDH activity was closely correlated (r = 0.99) with resistance to mafosfamide, up to 21-fold relative to controls. Transfectants were cross-resistant to other OAPs but not to phosphoramide mustard, ifosfamide mustard, melphalan, or acrolein. Resistance was completely reversed by pretreatment with 25 microM diethylaminobenzaldehyde, a potent ALDH inhibitor. Alkaline elution studies showed that expression of ALDH-1 reduced the number of DNA cross-links commensurate with mafosfamide resistance, and this reduction in cross-links was fully reversed by the inhibitor. Thus, overexpression of human class 1 ALDH alone is sufficient to confer OAP-specific drug resistance.

Tissue specific and androgen-regulated expression of human prostate-specific transglutaminase

Transglutaminases (TGases) are calcium-dependent enzymes catalysing the post-translational cross-linking of proteins. In the prostate at least two TGases are present, the ubiquitously expressed tissue-type TGase (TGC), and a prostate-restricted TGase (TGP). This paper deals with the molecular cloning and characterization of the cDNA encoding the human prostate TGase (hTGP). For this purpose we have screened a human prostate cDNA library with a probe from the active-site region of TGC. The largest isolated cDNA contained an open reading frame encoding a protein of 684 amino acids with a predicted molecular mass of 77 kDa as confirmed by in vitro transcription-translation and subsequent SDS/PAGE. The hTGP gene was tissue-specifically expressed in the prostate, yielding an mRNA of approx. 3.5 kb. Furthermore, a 3-fold androgen-induced upregulation of hTGP mRNA expression has been demonstrated in the recently developed human prostate cancer cell line, PC346C. Other well established human prostate cancer cell lines, LNCaP and PC-3, showed no detectable hTGP mRNA expression on a Northern bolt. The gene coding for prostate TGase was assigned to chromosome 3.

Initiation of protein synthesis in eukaryotic cells

It is becoming increasingly apparent that translational control plays an important role in the regulation of gene expression in eukaryotic cells. Most of the known physiological effects on translation are exerted at the level of polypeptide chain initiation. Research on initiation of translation over the past five years has yielded much new information, which can be divided into three main areas: (a) structure and function of initiation factors (including identification by sequencing studies of consensus domains and motifs) and investigation of protein-protein and protein-RNA interactions during initiation; (b) physiological regulation of initiation factor activities and (c) identification of features in the 5' and 3' untranslated regions of messenger RNA molecules that regulate the selection of these mRNAs for translation. This review aims to assess recent progress in these three areas and to explore their interrelationships.

A large variety of alternatively spliced and differentially expressed mRNAs are encoded by the human acute myeloid leukemia gene AML1

The human chromosome 21 acute myeloid leukemia gene AML1 is frequently rearranged in the leukemia-associated translocations t(8;21) and t(3;21), generating fused proteins containing the amino-terminal part of AML1. In normal blood cells, five size classes (2-8 kb) of AML1 mRNAs have been previously observed. We isolated seven cDNAs corresponding to various AML1 mRNAs. Sequencing revealed that their size differences were mainly due to alternatively spliced 5' and 3' untranslated regions, some of which were vast, exceeding 1.5 kb (5') and 4.3 kb (3'). These untranslated regions contain sequences known to control mRNA translation and stability and seem to modulate AML1 mRNA stability. Further heterogeneity was found in the coding region due to the presence of alternatively spliced stop codon-containing exons. The latter led to production of polypeptides that were smaller than the full-length AML1 protein; they lacked the trans-activation domains but maintained DNA binding and heterodimerization ability. The size of these truncated products was similar to the AML1 segment in the fused t(8;21) and t(3;21) proteins. In thymus, only one mRNA species of 6 kb was detected. Using in situ hybridization, we showed that its expression was confined to the cortical region of the organ. The 6-kb mRNA was also prominent in cultured peripheral blood T cells, and its expression was markedly reduced upon mitogenic activation by phorbol myristate acetate (TPA) plus concanavalin A (ConA). These results and the presence of multiple coding regions flanked by long complex untranslated regions, suggest that AML1 expression is regulated at different levels by several control mechanisms generating the large variety of mRNAs and protein products.

Kinetics of hydrogen bond breakage in the process of unfolding of ribonuclease A measured by pulsed hydrogen exchange

A sensitive test for kinetic unfolding intermediates in ribonuclease A (EC 3.1.27.5) is performed under conditions where the enzyme unfolds slowly (10 degrees C, pH 8.0, 4.5 M guanidinium chloride). Exchange of peptide NH protons (2H-1H) is used to monitor structural opening of individual hydrogen bonds during unfolding, and kinetic models are developed for hydrogen exchange during the process of protein unfolding. The analysis indicates that the kinetic process of unfolding can be monitored by EX1 exchange (limited by the rate of opening) for ribonuclease A in these conditions. Of the 49 protons whose unfolding/exchange kinetics was measured, 47 have known hydrogen bond acceptor groups. To test whether exchange during unfolding follows the EX2 (base-catalyzed) or the EX1 (uncatalyzed) mechanism, unfolding/exchange was measured both at pH 8.0 and at pH 9.0. A few faster-exchanging protons were found that undergo exchange by both EX1 and EX2 processes, but the 43 slower-exchanging protons at pH 8 undergo exchange only by the EX1 mechanism, and they have closely similar rates. Thus, it is likely that all 49 protons undergo EX1 exchange at the same rate. The results indicate that a single rate-limiting step in unfolding breaks the entire network of peptide hydrogen bonds and causes the overall unfolding of ribonuclease A. The additional exchange observed for some protons that follows the EX2 mechanism probably results from equilibrium unfolding intermediates and will be discussed elsewhere.

Adherence to the first-AUG rule when a second AUG codon follows closely upon the first

The rule that eukaryotic ribosomes initiate translation exclusively at the 5' proximal AUG codon is abrogated under rare conditions. One circumstance that has been suggested to allow dual initiation is close apposition of a second AUG codon. A possible mechanism might be that the scanning 40S ribosomal subunit flutters back and forth instead of stopping cleanly at the first AUG. This hypothesis seems to be ruled out by evidence presented herein that in certain mRNAs, the first of two close AUG codons is recognized uniquely. To achieve this, the 5' proximal AUG has to be provided with the full consensus sequence; even small departures allow a second nearby AUG codon to be reached by leaky scanning. This context-dependent leaky scanning unexpectedly fails when the second AUG codon is moved some distance from the first. A likely explanation, based on analyzing the accessibility of a far-downstream AUG codon under conditions of initiation versus elongation, is that 80S elongating ribosomes advancing from the 5' proximal start site can mask potential downstream start sites.

Participation of initiation factors in the recruitment of mRNA to ribosomes

The step of protein synthesis which is normally rate limiting, formation of the 48S initiation complex, is catalyzed by the group 4 initiation factors. Collectively they recognize the 7-methylguanosine-containing cap of mRNA, unwind mRNA secondary structure, and allow scanning for the initiation codon by the small ribosomal subunit. The activities of the eIF-4 polypeptides are modulated by phosphorylation. Recent studies shed new light on the mechanism of assembly of the 48S initiation complex and the effect of phosphorylation of one of the eIF-4 polypeptides, the cap-binding protein eIF-4E.

Translational control during amino acid starvation

Amino acid starvation of mammalian cells results in a pronounced fall in the overall rate of protein synthesis. This is associated with increased phosphorylation of the alpha-subunit of the initiation factor eIF-2, which in turn impairs the activity of the guanine nucleotide exchange factor, eIF-2B. Similar mechanisms have now been found to operate in the yeast, Saccharomyces cerevisiae, where the major physiological result is to circumvent the lack of external amino acids by promoting the translation of a transcription factor, GCN4, that facilitates the expression of a number of enzymes required for amino acid biosynthesis. This article reviews current knowledge of these mechanisms in both mammalian and yeast cells and identifies questions still requiring elucidation.

Regulation and function of non-AUG-initiated proto-oncogenes

A small, yet growing, number of cellular eukaryotic mRNAs encoding important regulatory proteins, such as c-myc and other proto-oncogenes, initiate translation from a non-AUG codon, usually in addition to initiating at a downstream AUG. The efficiency of non-AUG initiation on these natural cellular mRNAs varies considerably and appears to be governed by several features, including the codon sequence, the context surrounding the codon and the secondary structure of the transcript. In addition to factors which control the overall efficiency of c-myc non-AUG initiation, the relative efficiency of the upstream non-AUG initiation compared with the AUG initiation changes during the growth of cells. As lymphoid and fibroblast cells approach high densities in culture there is a sustained 5-10-fold induction in the synthesis of the non-AUG-initiated c-Myc 1 protein to levels comparable to or greater than the AUG-initiated c-Myc 2 protein. This increased efficiency of c-myc non-AUG initiation, due to methionine depletion of the growth medium, suggests that the scanning preinitiation complex can be regulated to enhance the recognition of a suboptimal non-AUG codon. The significance of non-AUG initiation for the growth-regulatory genes is illustrated by the different localizations of the int-2, bFGF and hck non-AUG-initiated proteins, the disruption of the c-myc and lyl-1 non-AUG initiation in tumor-derived cell lines, and the distinct biological function of the non-AUG-initiated forms of bFGF.