Search

New comparative genome hybridization technology on NotI-microarrays is presented (Karolinska Institute International Patent WO02/086163). The method is based on comparative genome hybridization of NotI-probes from tumor and normal genomic DNA with the principle of new DNA NotI-microarrays. Using this method 181 NotI linking loci from human chromosome 3 were analyzed in 200 malignant tumor samples from different organs: kidney, lung, breast, ovary, cervical, prostate. Most frequently (more than in 30%) aberrations--deletions, methylation,--were identified in NotI-sites located in MINT24, BHLHB2, RPL15, RARbeta1, ITGA9, RBSP3, VHL, ZIC4 genes, that suggests they probably are involved in cancer development. Methylation of these genomic loci was confirmed by methylation-specific PCR and bisulfite sequencing. The results demonstrate perspective of using this method to solve some oncogenomic problems.

Termination of translation in eukaryotes is governed by two polypeptide chain release factors. The middle (M) domain of the class 1 translation termination factor eRF1 contains the strictly conserved GGQ motif and involved in hydrolysis of the peptidyl-tRNA ester bond within the peptidyl transferase center of the large ribosome subunit. Heteronuclear NMR spectroscopy was used to map the interaction interface of the M-domain of human termination factor eRF1 with eukaryotic ribosomes. The protein was found to interact specifically with the large 60S ribosomal subunit: no interaction was detected between the M-domain of eRF1 and the 40S ribosomal subunit. The protein residues at the interaction interface are mainly situated on the long alpha-helix, alpha1 of the M-domain. Some residues adjacent to alpha1, in strand beta5, and in two short helices alpha2 and alpha3 are also involved in the protein-ribosome contact. The interaction of the functionally inactive mutant G183A with the 60S ribosomal subunit is substantially weaker than that found for the wild-type protein. Moreover, the interaction interfaces are not identical in these two cases. The results highlight the functional importance of the long alphal helix and also indicate that conformational flexibility of the GGQ loop is essential for forming tight protein-ribosome contacts.

A properties of atomic models of structure of eukaryotic triple complex eRF1 . mRNA . tRNAPhe containing human class-1 polypeptide release factor eRF1 at the A-site of human 80S ribosome, mRNA and P-site tRNAPhe, obtained before, are considered. The stricture of the complex is described using high resolution NMR structure of eRF1 M-domain. The structural properties of distribution of chemical cross-links are investigated, which allows us to choose correct model of positioning of the eRF1 molecule in ribosome A-site relative to stop codon of mRNA. A distributions of crosslinks between photoactivatable perfluoroaryl azide group of modified nucleotides of mRNA analogues and eRF1 molecule are modeled via molecular dynamics method. Twelve different mRNA analogues with modified nucleotides of stop signal in positions +4 to +9 with respect to the first nucleotide of the P-site codon are modeled. It was shown that only one of the two models of complex eRFI . mRNA . tRNA gives cross-link distribution in a good agreement with experimental data. A new features of the final structure of triple complex eRF1 . mRNA . tRNA is spatial proximity of stop-codon nucleotides to the C-domain of the eRF1, which explains previously obtained cross-link experimental data.

In universal-code eukaryotes, a single class-1 translation termination factor eRF1 decodes all three stop codons, UAA, UAG, and UGA. In some ciliates with variant genetic codes one or two stop codons are used to encode amino acid(s) and are not recognized by eRF1. In Stylonychia, UAG and UAA codons are reassigned as glutamine codons, and in Euplotes, UGA is reassigned as cysteine codon. In omnipotent eRF1s, stop codon recognition is associated with the N-terminal domain of eRF1. Because variant-code ciliates most likely evolved from universal code ancestor(s), structural features should exist in ciliate eRF1s that restrict their stop codon recognition. To find out amino acid residues which confer UAR-only specificity to Euplotes aediculatus eRF1, eRFI chimeras were constructed by swapping eRF1 E. aediculatus N-terminal domain sequences with the matching ones from the human protein. In these chimeras the MC-domain was from human eRF1. Functional analysis of these chimeric eRFI highlighted the crucial role of the two regions (positions 38-50 and 123-145) in the N-terminal domain of E. aediculatus eRF1 that restrict E. aediculatus eRF1 specificity toward UAR codons. Possibly, restriction of eRF1 specificity to UAR codons might have been an early event occurring in independent instances in ciliate evolutionary history, possibly facilitating the reassignment of UGA to sense codons.

Models of atomic structure of eukaryotic translation termination complex containing mRNA, P-site tRNAPhe, human class-1 polypeptide release factor eRF1 and 80S ribosome were constructed. The method of computational modeling was applied. The modeling was based on the functional and structural similarity between tRNA and eFR1 bound in the ribosomal A site. Structural template for the modeling was a known structure of the 70S ribosome complexed with mRNA, P- and A-site tRNAsPhe. The eRF1 molecule bound to the ribosome undergone substantial conformational changes resulting in the mutual configuration of the N- and M-domains similar to tRNA shape. Two models of binding of eRF1 to mRNA at the A-site in the presence of P-site tRNA were generated and characterized by a shape complementarity between the mRNA stop codon and grooves of the different sides of the molecular surface of the fragment of alpha2-helix, NIKS loop and alpha-helix of the N-domain. In the model 1 the stop-codon nucleotides were at the equal distances from the N- and C-domains. In the model 2 the stop-codon was proximal to the NIKS and YxCxxxF motifs of the N-domain. Both models fit the genetic and biochemical data available so far.

Lung cancer is one of the most frequent neoplasia in the Russia, the United States and Europe. This cancer is associated with functional activity changes of many genes. In the present study TIMP3, DAPK1 and AKR1B10 genes transcription analysis of squamous cell lung cancer specimens was carried out using reverse transcription-PCR. Substantial increasing of AKR1B10 transcription level is revealed in 80% tumor samples. TIMP3 and DAPK1 transcription level is considerably decreased in 76 and 72% tumor specimens, accordingly. These results may point out that all three genes are important for squamous cell lung cancer tumorogenesis while AKR1B10 is potential oncogene whereas TIMP3 and DAPK1 are potential tumor suppressor genes. We suggest that revealed substantial transcription level-changes of investigated genes may be used for oncodiagnostics.

RHOA protein, a member of small GTPases family, is implicated in cell morphogenesis, adhesion, and in cell cycle regulation. RHOA gene (3p21.31) exhibits cell transformation activity, and therefore gene is considered as a potential oncogene. The aim of this study was to investigate RHOA transcription and copy number changes in three epithelial tumors (breast, renal cell and epithelial ovarian carcinomas, 45 tumor/normal pairs altogether). EII, HhaI, AciI n Bsh1236I). Hypomethylation of the RHOA promoter region in tumor DNA was observed two times more frequently than increased methylation. Moreover, all (15) cancer cases with hypomethylation of the RHOA gene showed a 2-10 fold increased expression of RHOA. It was concluded that gene copy multiplication and demethylation of the RHOA promoter region can contribute to transcription activation of this gene in epithelial tumors.

Translation termination in eukaryotes is governed by two proteins, belonging to the class-1 (eRF1) and class-2 (eRF3) polypeptide release factors. eRF3 catalyzes hydrolysis of GTP to GDP and inorganic phosphate in the ribosome in the absence of mRNA, tRNA, aminoacyl-tRNA and peptidyl-tRNA but needs the presence of eRF1. It's known that eRF1 and eRF3 interact with each other in vitro and in vivo via their C-terminal regions. eRF1 consists of three domains - N, M, and C. In this study we examined the influence of individual domains of the human eRF1 on induction of the human eRF3 GTPase activity in the ribosome in vitro. It was shown that none of the N-, M-, C- and NM-domains induces eRF3 GTPase activity in presence of the ribosomes. MC-domain does induce GTPase activity of eRF3 but four times less efficient than full-length eRF1, therefore, MC-domain (and very likely M-domain) binds to the ribosome in the presence of eRF3. Based on these data and taking into account the data available in literature, a conclusion was drawn that the N domain of eRF1 is not essential for eRF1-dependent induction of the eRF3 GTPase activity. A working hypothesis is formulated, postulating that GTPase activity eRF3 during the translation termination is associated with the intermolecular interactions of GTP/GDP, GTPase center of the large ribosomal subunit (60S), MC-domain of eRF1, C-terminal region and GTP-binding domains of eRF3, but without participation of the N-terminal region of eRF3.

Thermal denaturation of eukaryotic class-1 translation termination factor eRF1 and its mutants was examined using differential scanning microcalorimetry (DSK). Changes of free energy caused by mutants in the N domain of human eRF1 were calculated. Melting of eRF1 molecule composed of three individual domains is cooperative. Some amino acid substitutions did not affect protein thermostability and in some other cases even slightly stabilize the protein globule. These imply that these amino acid residues are not involved in maintenance of the 3D structure of human eRF1. Thus, in Glu55Asp, Tyr125Phe, Asn61Ser, Glu55Arg, Glu55A1a, Asn61Ser + Ser64Asp, Cys127Ala and Ser64Asp mutants selective inactivation of release activity is not caused by a destabilization of protein 3D structure and, most likely, is associated with local stereochemical changes introduced by substitutions of amino acid side chains in the functionally essential sites of N-domain molecule. Some residues (Asn129, Phe131) as shown by calorimetric measurements are essential for preservation of stable protein structure, but at the same time they affect selective stop codon recognition probably via their neighboring amino acids. Recognition of UAG and UAA stop codons in vitro is more sensitive to preservation of protein stability than the UGA recognition.

Ninety four NotI-STS markers to seventy two individual NotI clones were developed basing on DNA nucleotide sequences from NotI-"jumping" and "linking" NotI-libraries of human chromosome 3. The localization of NotI-STS markers and their ordering on chromosome was established by combined data of RH-mapping (our data), contig-mapping, cytogenetic mapping and in silico mapping. Performed comparison of NotI-STS DNAs with human genome sequences revealed two gaps in the regions, 3p21.33 (marker NLI-256) and 3p21.31 (NL3-005), and segmental duplication. Identical DNA fragments are localized in the regions 12q and 3p22-21.33 (marker NL3-007). In the region 3q28-q29 (marker NLM-084) a fragment was detected with its identical copies present also on chromosomes 1, 2, 15 and 19. For 69 NotI-STSs, significant homologies with nucleotide sequences of 70 genes and two cDNAs were detected taking in consideration homologies to NotI-STS 5'- and 3'-terminal sequences. Association of NotI-STSs with genes is confirmed by high correlation of gene density distribution with the density of NotI-STS markers on the map of human chromosome 3. Obtained data evidence possibility of NotI-STS marker application as gene markers and allow considering constructed NotI-map as gene map of human chromosome 3.

Methylation of the promoter CpG-islands of the candidate tumor suppressor gene RASSF1A (3p21.31) was studied in primary tumors of kidney, breast and ovary (172 cases). Methylation-specific PCR (MSP) and methyl-sensitive restriction endonuclease digestion followed by PCR (MSRA) were applied. Statistically significant correlation (P << 10(-6)) was shown for the results of the MSP and MSRA, and the data of bisulfite sequencing reported earlier. The frequency of RASSF1A methylation according to MSP and MSRA was 86% (25/29) and 94% (50/53) in renal cell carcinoma (RCC) and 64% (18/28) and 78% (32/41)--in breast carcinoma (BC) samples, and 59% (17/29) and 73% (33/45) in ovarian epithelial tumors (OET), respectively. The use of several methyl-sensitive restriction enzymes (HpaII, HhaI, Bsh12361, AciI) enhanced the sensitivity of MSRA and allowed to analyze methylation status of 18 CpG-pairs in the RASSF1A CpG-island. Density of methylation of the RASSF1A CpG-island was 72% (644/900) in RCC, 63% (361/576) in BC, and 58% (346/594) in OET samples (18 CpG-pairs multiplied to the number of samples shown methylation were assumed as 100%). The RASSF1A gene methylation was also observed in samples of morphologically normal tissues adjacent to corresponding tumors (11-35%), but it was not detected in blood DNAs of healthy donors (0/15). The RASSF1A methylation frequency did not show significant correlation to tumor stage, grade and metastases (P = 0.3-1.0). The RASSF1A gene methylation was observed more frequently than other investigated aberrations--hemi- and homozygous deletions inside or around this gene. These observations are consistent with the hypothesis that the RASSF1A gene methylation is an early event in the carcinogenesis and one of the dominant ways of its inactivation.

The integral structural parameters and the shape of the molecule of human translation termination factor eRF1 were determined from the small-angle X-ray scattering in solution. The molecular shapes were found by bead modeling with nonlinear minimization of the root-mean-square deviation of the calculated from the experimental scattering curve. Comparisons of the small-angle scattering curves computed for atomic-resolution structures of eRF1 with the experimental data on scattering from solution testified that the crystal and the solution conformations are close. In the ribosome, the distance between the eRF1 motifs GGQ and NIKS must be shorter than in crystal or solution (75 versus 107-112 A). Therefore, like its bacterial counterpart RF2, the eukaryotic eRF1 must change its conformation as it binds to the ribosome. The conformational mobility of eukaryotic and prokaryotic class-1 release factors is another feature making them functionally akin to tRNA.

The review considers the results obtained by several groups in the fields of identification of polymorphic loci in the human genome, localization and analysis of genes associated with epithelial tumors of various origins, and generation of molecular markers of socially important oncological diseases. In the first two cases, work was initiated and supported by the Russian program Human Genome. To find new polymorphic loci in the human genome, di-, tri-, and tetranucleotide repeats were searched for in an ordered cosmid library of chromosome 13, NotI and cosmid clones of chromosome 3, and in brain EST. In total, nine polymorphisms and almost 200 STS were identified. Markers of NotI clones of chromosome 3 were associated with particular genes. Polymorphic loci NL1-024, NL2-007, and EST04896 were employed in analysis of deletions from chromosome 3p in tumor DNA. Deletion mapping of 3p in epithelial tumors of five types revealed six critical regions containing potential tumor suppressor genes. Of these, two were in the distal region of chromosome 3p and four, in region 3p21.3. A significant correlation was observed for the frequency of allelic deletions and the stage and the grade of tumors (P < 0.05). On the strength of these findings, genes of region 3p were associated with both tumor development and progression, and proposed as prognostic markers. Regions LUCA and AP20 (3p21.3) showed a high (90%) frequency of aberrations, including homozygous deletions in almost 20% cases. The peak of allelic deletions from region D3S2409-D3S3667 (600 kb) was statistically valid (P = 10(-3)). Regions AP20 and D3S2409-D3S3667 (3p21.3) were for the first time associated with tumorigenesis. Clusters of tumor suppressor genes were identified in regions LUCA, AP20, and D3S2409-D3S3667. Methylation of RASSF1A and RARbeta2 (3p) was associated with early carcinogenesis, and that of SEMA3B, with tumor progression. These findings are useful for early diagnostics and post-surgery prognosis of tumors.

Reviewed and discussed are the recent data demonstrating profound functional similarity between class-1 translation termination factors (RF1 and RF2 in prokaryotes, aRF1 and eRF1 in Archaea and eukaryotes, respectively) and aminoacyl-tRNA as regards their roles in the course of translation on the ribosome. Functional analogy of these two components of the cell protein-synthesizing machinery was suggested long ago; however, numerous experimental proofs have been obtained only recently. This similarity implies that decoding of the genetic information by the ribosomal machine is performed similarly at all stages of translation, though tRNA plays the main role at initiation and elongation, while the protein is most important for termination. Earlier it was found that nucleic acids (ribozymes) can operate like the protein enzymes, and now we have got evidence for the reverse: a protein (translation termination factor) can act like a nucleic acid (tRNA). Thus one can speak of "exchange" of molecular functions between proteins and nucleic acids. Therefore, the profound chemical difference between proteins and nucleic acids is not an insuperable barrier to their mutual functional replacement in certain situations.

Studies of the recent decade, including sequencing of numerous human genome regions, allowed a great progress in detection of new tumor suppressor genes (TSG) and development of new means of their identification and analysis. Effective methods of genome scanning and TSG identification combine DNA array techniques and subtraction hybridization. Alternative ways take advantage of new extrachromosomal vector systems (pETE, pETR) and the functional gene inactivation test. A breakthrough was made in localizing new TSG on the human chromosome 3 short arm, which harbors tumor-suppressing regions and is often rearranged in various tumors and in early carcinogenesis. On 3p, only three putative TSG were known five years ago, and at least ten were identified by the end of 2002. The role of new TSG in carcinogenesis is commonly inferred from a decrease in their transcription in tumor cell lines or primary tumors and from their ability to suppress the growth of these. Protein products of 3p TGS play an important part, constraining cell malignization. Some are directly involved in regulating the cell cycle and inducing apoptosis (RASSFIA), others suppress angiogenesis (Sema3B) or metastasis (Hyal-1). Numerous attempts to find mutations in exons of silent genes failed, and at least half of the new candidate genes (RASSFIA, CACNA2D2, BLU, HYAL1, SEMA3B, RAR-beta) proved to be inactivated by promoter methylation.

Termination of protein synthesis (hydrolysis of the last peptidyl-tRNA on the ribosome) takes place when the ribosomal A site is occupied simultaneously by one of the three stop codons and by a class-1 translation termination factor. The existing procedures to measure the functional activity of this factor both in vitro and in vivo have serious drawbacks, the main of which are artificial conditions for in vitro assays, far from those in the cell, and indirect evaluation of activity in in vivo systems. A simple reliable and sensitive system to measure the functional activity of class-1 translation termination factors could considerably expedite the study of the terminal steps of protein synthesis, at present remaining poorly known, especially in eukaryotes. We suggest a novel system to test the functional activity in vitro using native functionally active mRNA, rather than tri-, tetra-, or oligonucleotides as before. This mRNA is specially designed to contain one of the three terminating (stop) codons within the coding nucleotide sequence. Plasmids have been generated that carry the genes of suppressor tRNAs each of which is specific toward one of the three stop codons. They were shown to support normal synthesis of a reporter protein, luciferase, by reading through the stop codon within the coding mRNA sequence. We have demonstrated that human class-1 translation termination factor eRF1 is able to compete with suppressor tRNA for a stop codon and to completely prevent its suppressive effect at a sufficient concentration. Forms of eRF1 with point mutations in functionally essential regions have lower competitive ability, demonstrating the sensitivity of the method to the eRF1 structure. The enzymatic reaction catalyzed by the full-size reporter protein is accompanied by emission of light quanta. Therefore, competition between suppressor tRNA and eRF1 can be measured using a luminometer, and this allows precise kinetic measurements in a continuous automatic mode.

Nonsense mutations in the dystrophin gene are the cause of Duchenne muscular dystrophy (DMD) in 10-15% of patients. In such an event, one approach to gene therapy for DMD is the use of suppressor tRNAs to overcome the premature termination of translation of the mutant mRNA. We have carried out cotransfection of the HeLa cell culture with constructs containing a suptRNA gene (pcDNA3suptRNA) and a marker LacZ gene (pNTLacZhis) using their polymer VSST-525 complexes. It was found that the number of cells producing beta-galactosidase depends inversely on the dose of the suptRNA gene. A single in vivo injection of the construct providing for expression of the suptRNAochre gene into mdx mouse muscle resulted in the production of dystrophin in 2.5% of fibers. This suggests that suppressor tRNAs are applicable in gene therapy for hereditary diseases caused by nonsense mutations.

Analysis of DNA sequences of the human chromosomes 21 and 22 performed using a specially designed MegaGene software allowed us to obtain the following results. Purine and pyrimidine nucleotide residues are unevenly distributed along both chromosomes, displaying maxima and minima (Y waves phi) with a period of about 3 Mbp. Distribution of G + C along both chromosomes has no distinct maxima and minima, however, chromosome 21 contains considerably less G + C than chromosome 22. Both exons and Alu repeats are unevenly distributed along chromosome 21: they are scarce in its left part and abundant in the right part, while MIR elements are quite monotonously spread along this chromosome. The Alu repeats show a wave-like distribution pattern similar for both repeat orientations. The number of the Alu repeats of opposite orientations was equal for both studied chromosomes, and this may be considered a new property of the human genome. The positive correlation between the exon and Alu distribution patterns along the chromosome, the concurrent distribution of Alu repeats in both orientations along the chromosome, and the equal copy numbers for Alu in direct and inverted orientations within an individual chromosome point to their important role in the human genome, and do not fit the notion that Alu repeats belong to parasitic (junk) DNA.

Physical mapping of the human genome appears to be a complicated problem for the molecular genetics of higher organisms. The difficulties are not only due to the great amount of work to be done, but mostly due to the peculiarities of genome organization (repeated nucleotide sequences, non-clonable, methylated and toxic fragments, etc.). Mapping procedures based on molecular hybridization do not allow to obtain the unambiguous results due to the occurrence of the repeats of different types common for the genome as a whole. Giant DNA fragments (e.g. YAC--Yeast Artificial Chromosomes) undergo deletions, transformations and chimerism and cannot be the main approach for genome mapping. The new strategy for genome mapping is described with reference to the human chromosome 3--one of the largest human chromosomes. The main steps of this new strategy are the following: 1) sequencing of STS (Sequence Tagged Sites) flanking DNA rare-cutting restriction sites; 2) alignment of the STS along the chromosomal DNA (generation of the contigs) based on the comparative computer analysis of STS from the same and from the neighboring restriction sites, 3) determination of the distance between STS by hybridization of large DNA fragments with STS. In order to apply this strategy called "shot-gun sequencing strategy for long range genome mapping" a new family of vectors (SK series) specially designed for the genome cloning was constructed and the simplified inexpensive technology for preparing the jumping/linking libraries was developed. Chromosome 3 fragments were sequenced around NotI sites, and the contigs covering about 50 Mbp of genomic DNA were constructed. The perspectives for the whole human genome mapping are discussed.

Antisense RNA transcription of human c-myc gene has been examined in HeLa, Burkitt lymphoma BL-60 t(8;22) cells, and diploid fibroblasts. By means of the primer extension technique two startpoints of antisense transcription have been detected and mapped with the first (untranslated) exon of the c-myc gene. Similarity between the antisense nucleotide sequence of the first c-myc intron and the SV40 DNA fragment containing the binding sites for transcriptions factors GT-I, GT-II, TC-I, and TC-II has been revealed by computer analysis. It has been established that the DNA fragment of the first c-myc intron is able to form complexes with proteins from the HeLa cell extract. Three nucleotide sequences (TTTCTG, TTTTTA, and TGACTTGTC) are involved in the reactions. These data imply that the c-myc antisense transcripts might take part in the regulation of human c-myc gene expression.

By means of small angle X-ray scattering, an aggregation of beef pancreas Trp-tRNA synthetase (EC 6.1.1.2) was observed at physiological temperatures. A Trp-tRNA synthetase preparation which is homogeneous after PAGE in beta-ME-SDS was found to be heterogeneous in particle sizes even at low (4-8 degrees C) temperature. At heating up to 30-45 degrees C, the oligomer sizes increased as well as its proportion depending on the incubation time and temperature; very large aggregates were observed 10 times exceeding the sizes of initial particles. Cooling to 20 degrees C caused no disaggregation due to disulphide bond formation between associated subunits of Trp-tRNA synthetase. A hypothesis is proposed that the aggregation of bovine Trp-tRNA synthetase evaluated in vitro and not observed earlier with any aminoacyl-tRNA synthetases of unicellular organisms might serve as one of the mechanisms of its compartmentation in pancreas.

The aim of this review is to summarize the data obtained in the author's laboratory during the last decade. The main objects of these investigations were mammalian aminoacyl-tRNA synthetases, mainly bovine tryptophanyl-tRNA synthetase (EC 6.1.1.2). The data are discussed and compared with those described in literature. In the course of these studies it turned out that some properties of mammalian aminoacyl-tRNA synthetases for instance, nuclear location of some of the synthetases, presence of extra-domain in bovine tryptophanyl-tRNA synthetase capable of catalyzing hydrolysis of ATP and GTP in the absence of Zn2+ ions and normal aminoacylation capacity, ability to bind to one of the glycolytic enzymes, glyceraldehyde-3-phosphate dehydrogenase, formation of aminoacylated and pyrophosphorylated forms of tryptophanyl-tRNA synthetase etc., seem to be unrelated to the main function of the synthetases, catalysis of aminoacyl-tRNA formation, and, therefore, might be classified as noncanonical ones. Comparison of prokaryotic and eukaryotic aminoacyl-tRNA synthetases indicates the multipotential nature of the latter.

Part of the transcriptionally active rat genomic locus, K51, has been sequenced (2,321 b.p.). The coding DNA chains was identified and two long open reading frames (ORF) were found by computer analysis one of which seems to be much less probable. By transcriptional mapping, at least two exons (A - about 110 b.p., B - 549 b.p.) were revealed. Computer search showed that the deciphered nucleotide and amino acid sequence was absent in the data-banks up to the end of 1988. Numerous GC rich direct repeats were noticed in the fragment. At protein level, resemblance with hypervariable N- and C-terminal domains of cytokeratins was visualized. Some regions of the putative protein product possess relatedness with some other gene products, particularly with ATP binding domains of mos oncoprotein.

After prolonged cultivation in the presence of increasing amounts of carboxyl-substituted tryptophan analogs (tryptamine and tryptophanol), cell lines resistant to high concentrations of these compounds were obtained. The initial culture was the Madin-Darby line of spontaneously transformed bovine kidney cells. In the resistant lines the amount of tryptophanyl-tRNA synthetase (E. C. 6.1.1.2) is manyfold increased as shown by two criteria: (i) enzymatic activity (ATP-PPi isotopic exchange) per mg of protein, (ii) binding of in vivo 35S-labeled proteins to polyclonal antibodies against tryptophanyl-tRNA synthetase. It was shown that tryptophanyl-tRNA synthetase is phosphorylated in vivo, and the degree of phosphorylation of the enzyme in initial cells seems to be higher then in the resistant ones. The Km value for tryptophan is not significantly changed for the enzyme from resistant cells. The permeability for tryptophan and its analogs is reduced in the resistant cells. It is proposed that the acquisition of the resistance against tryptophan analogs are due to alterations at the genomic level (for example, gene amplification etc.).

The review summarizes the data mostly published in 1982 and 1983 concerning isolation, identification, cellular localization, deciphering of primary structure, and properties of the proteins coded by the retroviral oncogenes. A tentative classification of oncoproteins is suggested. The main part of the review is devoted to oncoproteins with tyrosine specific protein kinase activity. A hypothesis is proposed concerning the possible mode of action of these proteins on transforming cells. It is stressed in conclusion that the remarkable progress of our knowledge in this field has been mostly achieved due to recombinant DNA technique successfully and widely applied for studying oncogene structure and function.

A method is proposed for analysis of natural and chemically modified polynucleotides which consists in enzymatic conversion of the polymer or oligomer into nucleosides followed by cation-exchange chromotography on the microcolumns. By using the method developed it was shown that after treatment of the yeast tRNAVal and tRNAPhe with monoperphthalic acid N-oxides of adenosine and cytidine were formed. Poly (U, G) was not modified at a measurable extent whereas GMP was decomposed. In tRNAVal (yeast)the adenosines and cytosines of the anticodon loop and 3'-end are most reactive; it is the case for the C17 of the diHU-loop as well. These data are in agreement with the results obtained for tRNA modification with other reagents and for limited enzymatic hydrolysis of the tRNAVal. The limitations of the reaction of the monoperphthalate with nucleic acids are briefly discussed.