Gene expression profiling of cultured human bronchial epithelial and lung carcinoma cells.

Lung cancer is a complex collection of diseases that is thought to begin with single mutated progenitor cells and culminates in any of several clinically described pathologies. Our knowledge of the molecular events that lead to different lung cancer types--small cell carcinoma, squamous cell carcinoma, adenocarcinoma, and large cell carcinoma--is incomplete. Nonetheless, it is evident that genetic changes that impact multiple molecular networks are involved in the generation of each specific phenotype. Due to the obvious complexity of these processes, the simultaneous quantitative monitoring of changes in the expression of genes that define these networks can provide mechanistic information to increase our understanding of the molecular basis for human pulmonary carcinogenesis. To this end, we have employed a commercially available human cDNA array (Atlas Human Array, Clontech Laboratories) to systematically screen for alterations in the expression of 600 genes in normal human bronchial epithelial (NHBE) cells as well as in several lung carcinoma lines. Studies on the reproducibility and variability of array results indicate that a 2-fold or greater difference in the expression of a particular gene could be considered a real difference in transcript abundance. Accuracy of gene expression as measured in the array was verified by comparing mRNA levels of the proto-oncogene c-myc in the array with results obtained by traditional Northern blot analysis and by quantitative RT-PCR. Gene expression profiles were compared within and among cell types. The differential expression of 17 genes, including downregulation of MRP8 and MRP14 and upregulation of CYP1B1, was observed in all four carcinoma lines compared to NHBE cells. The direction of all 17 gene expression differences, either upregulation or downregulation relative to NHBE cells, was the same for all four carcinoma lines, underscoring their common molecular features. Each lung tumor line also exhibited a number of unique differences compared to both normal cells and the other tumor cell lines. These differences may be due to differences in the cellular origin and/or pathology of the cell lines studied.

Variations in the VPg protein allow a potyvirus to overcome va gene resistance in tobacco.

The va gene is used in commercial Burley tobacco cultivars including cv TN 86 to confer resistance to tobacco vein mottling virus (TVMV) and, to some extent, other potyviruses. A naturally occurring strain of TVMV (TVMV-S), which overcomes this resistance, was isolated from TN 86 plants. To investigate the mechanism by which TVMV-S overcomes va gene resistance, a cDNA clone encompassing the complete genome of TVMV-S was produced. Using chimeric transcripts combining regions of TVMV-S and regions of the wild-type strain (TVMV-WT) to which TN 86 is resistant, it was demonstrated that a domain within the VPg protein is responsible for overcoming va resistance in TN 86. DNA sequence analysis revealed six amino acid differences between the two strains of TVMV within this domain. Inclusion of sequences for four of the TVMV-S VPg amino acids was sufficient to confer the resistance-overcoming phenotype to all corresponding transcripts. Coinoculation experiments suggested that the resistance of TN 86 to TVMV-WT was not due to the induction of a general host defense response. The results are compatible with the hypothesis that VPg must assume an appropriate configuration in order to interact with appropriate host components and facilitate systemic virus movement.

Molecular cloning, in vitro expression and characterization of a plant squalene synthetase cDNA.

Squalene synthetase (farnesyl-diphosphate:farnesyl-diphosphate farnesyltransferase, EC 2.5.1.21) catalyzes the first committed step for sterol biosynthesis and is thought to play an important role in the regulation of isoprenoid biosynthesis in eukaryotes. Using degenerate oligonucleotides based on a conserved region found in yeast and human squalene synthetase genes, a cDNA was cloned from the plant Nicotiana benthamiana. The cloned cDNA contained an open reading frame of 1234 bp encoding a polypeptide of 411 amino acids (M(r) 47002). Northern blot analysis of poly(A)+ mRNA from N. benthamiana and N. tabacum cv. MD609 revealed a single band of ca. 1.6 kb in both Nicotiana species. The identity and functionality of the cloned plant squalene synthetase cDNA was further confirmed by expression of the cDNA in Escherichia coli and in a squalene synthetase-deficient erg9 mutant of Saccharomyces cerevisiae. Antibodies raised against a truncated form of the protein recognized an endogenous plant protein of appropriate size as well as the full-length bacterially expressed protein as detected by western analysis. Comparison of the deduced primary amino acid sequences of plant, yeast, rat and human squalene synthetase revealed regions of conservation that may indicate similar functions within each polypeptide.

Construction and analysis of infectious transcripts from a resistance-breaking strain of tobacco vein mottling potyvirus.

The Burley tobacco (Nicotiana tabacum) cultivar TN 86 is "resistant" to most strains of tobacco vein mottling potyvirus (TVMV), the virus being restricted to epidermal cells of inoculated leaves. One strain, designated TVMV-S, overcomes this resistance and infects cv TN 86 systemically. To begin our investigation of the molecular basis for the resistance-breaking phenomenon, we have completed the cloning and sequencing of the TVMV-S RNA genome. The complete cDNA clone, under the control of a T7 RNA polymerase promoter, was used to produce infectious transcripts which were tested for their ability to reproduce the characteristics of TVMV-S RNA on three types of tobacco (N. tabacum cv TN 86, N. tabacum cv KY 14, and N. benthamiana). Timing of symptom appearance, symptom type, and titer of virus were identical to those of plants inoculated with TVMV-S RNA. As a step toward mapping the responsible genetic region(s) that contribute(s) to resistance-breaking by TVMV-S, the nucleotide and deduced amino acid sequences were compared to those of wild-type TVMV, a strain that does not overcome cv TN 86 resistance. Variant TVMV-S transcripts containing changes within the VPg cistron exhibited an altered pattern of infectivity on cv TN 86.

Nucleotide sequence of the capsid protein cistrons from six potato virus Y (PVY) isolates infecting tobacco.

Complementary DNA libraries representing the capsid protein cistron of the potato virus Y (PVY) isolate 'Chilean', 'Hungarian', MsNr, NsNr, O, and 'Potato US' were synthesized and used as template for polymerase chain reaction (PCR) amplification. An AUG codon for initiating a discrete capsid protein (CP) open reading frame was embedded upstream of the first codon of the CP cistrons. PCR-amplified products of the expected size of 0.8 kilo bases were cloned into the transcription vector pBS(+). The fidelity of each PCR-amplified PVY CP cistron was tested by transcribing recombinant plasmids in vitro and translating the transcripts in two cell free translation systems. Translation analysis of in vitro transcribed PVY CP cistrons consistently yielded a polypeptide co-migrating with authentic CP that was immunoprecipitated by anti PVY 'Chilean' antibodies. The nucleotide sequence of each capsid protein gene was determined by dideoxy sequence analysis. Each capsid protein gene was determined to be 801 nucleotides in length, encoding a deduced protein of 267 amino acids with calculated M(r) ranging from 29,799 to 29,980. The nucleic acid sequence similarity between the six isolates ranged between 89 to 97% and the amino acid similarity between 91 to 99%. The high level of amino acid sequence similarity confirms the classification of these viruses as isolates of PVY.

Expression in transgenic plants of a viral gene product that mediates insect transmission of potyviruses.

Helper component (HC) is a virus-encoded nonstructural protein that is required for transmission of potyviruses by their aphid vectors. As a prelude to studies on the molecular basis of HC activity, a cDNA clone (pPB-3) was constructed that contained the first three cistrons (34 kDa-HC-42 kDa) of the RNA genome of the potyvirus tobacco vein mottling virus, the first six nucleotides of the adjacent cylindrical inclusion body protein cistron, and a synthetic translation termination codon. This construction was introduced into tobacco cells via a Ti plasmid-based vector. Northern blot analysis of transgenic plants demonstrated the presence of an RNA of the size expected from the construction of pPB-3, and Western blot analysis revealed the presence of a protein that comigrated with authentic HC, indicating that the proteolytic activity necessary to produce mature-sized HC was encoded by pPB-3. The HC produced in the transgenic plants was demonstrated to be active in a virus transmission bioassay with aphids.

Expression of a potyvirus non-structural protein in transgenic tobacco.

A cDNA fragment encoding the cytoplasmic inclusion protein of tobacco vein mottling virus was inserted into the plant expression cassette of a Ti plasmid-based binary vector. The vector was transferred to Agrobacterium tumifaciens, and following a modified leaf disc procedure, transformed tobacco plants were obtained. Analysis of poly(A)+ RNA from transgenic plants revealed a novel RNA of approximately 2100 nucleotides possessing tobacco vein mottling virus sequences. Also, immunoprecipitation of protein extracts of [35S]methionine-labeled transformed callus using anti-cytoplasmic inclusion protein antiserum revealed a polypeptide of approximately 70 kDa. This size is consistent with that predicted from the inserted tobacco vein mottling virus coding sequences. Together these data demonstrate the expression of the cytoplasmic inclusion protein in the absence of viral infections.

In vitro analysis of tobacco vein mottling virus NIa cistron: evidence for a virus-encoded protease.

Potential protease functions associated with the NIa nuclear inclusion protein of tobacco vein mottling virus (TVMV) were investigated. In the absence of treatments, in vitro translation of viral RNA produced various polypeptides representing each of the proposed TVMV cistrons--28K-HC-42K-CI-5.5K-NIa-NIb-CP. When viral RNA was first hybridized to DNA probes complementary to the NIa cistron, and then treated with RNase H prior to translation, a 48-kDa processing product, immunologically identified as the NIa protein, was abolished. In its place was observed a series of larger polypeptides, immunologically identified as fusion products of the cylindrical inclusion (CI) and NIa cistrons. The use of probes which permitted translation through as few as 15 nucleotide residues beyond the sequences encoding the proposed carboxyl terminus of NIa resulted in normal processing. None of the DNA probes affected an apparent cleavage between the helper component (HC) and 42K proteins. Cloned cDNA regions representing the NIa cistron and flanking sequences were inserted in transcription vectors. Translation of the in vitro transcript resulted in synthesis, not of a large fusion polyprotein, but, of a mature-sized NIa polypeptide. In vitro transcripts, lacking the 3'-most sequences that were expected to encode the protease active site of the NIa protein, were translated. These generated an apparent fusion polypeptide that reacted with antisera to both CI and NIa. The results indicate that the NIa gene product functions as a protease responsible for some but not all of the cleavage events which lead to the production of the mature forms of TVMV proteins.

Juvenile hormone regulation of mRNA levels for a highly abundant hemolymph protein in larval Trichoplusia ni.

Several hemolymph proteins ranging in size from 73 to 76 kDa increase to very high levels just prior to metamorphosis in Trichoplusia ni (Lepidoptera). One of these proteins (pI = 5.8, Mr = 76,000) was selected for a study of hormonal regulation. The appearance of this protein could be suppressed in vivo by topical treatment with the juvenile hormone analog fenoxycarb. An antiserum for this protein was prepared and shown to react selectively with the 76-kDa protein in whole hemolymph. Translation of poly(A)-containing RNA from untreated larvae yielded the 76-kDa protein, whose identity was verified with the antibody, whereas mRNA from juvenile hormone analog-treated larvae did not. These data indicate that juvenile hormone acts to regulate the level of the mRNA of this hemolymph protein.

Amino acid sequence of the mRNA cap-binding protein from human tissues.

The 25-kDa mRNA cap-binding protein (CBP) involved in translation was purified by affinity chromatography from human erythrocytes and rabbit reticulocytes. The sequences of eight human and seven rabbit tryptic and V8 proteolytic peptides were determined. Based on the peptide sequence data, oligodeoxynucleotide probes were synthesized and used to screen human fibroblast and lymphocyte lambda cDNA libraries. The DNA sequence obtained from recombinant lambda phage inserts was found to code for all but one peptide. A 23-base oligonucleotide was synthesized based on the DNA sequence and used to prime synthesis of cDNA from human placental mRNA to construct a third library in lambda gt10. Screening with a 22-base oligonucleotide, whose sequence was upstream from the 23-base primer, yielded numerous recombinant phages with approximately equal to 250-base inserts. The 1900-base-pair cDNA sequence compiled from all phage inserts appeared to represent the entire primary sequence of CBP (Mr 25,117). Blot analysis of human placental and HeLa mRNA revealed multiple CBP mRNA species ranging from 1925 to 2250 bases. The amino acid sequence of CBP showed homology to the cap-binding PB2 protein of influenza virus.

The nucleotide sequence of tobacco vein mottling virus RNA.

The nucleotide sequence of the RNA of tobacco vein mottling virus, a member of the potyvirus group, was determined. The RNA was found to be 9471 residues in length, excluding a 3'-terminal poly(A) tail. The first three AUG codons from the 5'-terminus were followed by in-frame termination codons. The fourth, at position 206, was the beginning of an open reading frame of 9015 residues which could encode a polyprotein of 340 kDa. No other long open reading frames were present in the sequence or its complement. This AUG was present in the sequence AGGCCAUG, which is similar to the consensus initiation sequence shared by most eukaryotic mRNAs. The chemically-determined amino acid compositions of the helper component and coat proteins were similar to those predicted from the nucleotide sequence. Amino acid sequencing of coat protein from which an amino-terminal peptide had been removed allowed exact location of the coat protein cistron. A consensus sequence of V-(R or K)-F-Q was found on the N-terminal sides of proposed cleavage sites for proteolytic processing of the polyprotein.

Cistron mapping of tobacco vein mottling virus.

The location and order of cistrons in the RNA of the potyvirus tobacco vein mottling virus (TVMV) were investigated. Hybrid-arrested translation, using cloned single-stranded DNA probes complementary to various regions of the viral RNA, was performed and the resulting translation products were analyzed by electrophoresis. The pattern of polypeptides produced with each probe was different from that of control reactions containing RNA alone. Immunoprecipitation of reaction products with antisera to five potyviral proteins revealed that, in some cases, portions of cistrons were masked by the DNA probe resulting in the precipitation of altered polypeptides. In other cases, the entire cistron was affected, resulting in the total loss of immunoreactive material. By correlating the location of each DNA probe with the resulting pattern of translation products, it was possible to construct a map of known cistrons and potential coding regions for additional cistrons in the genomic RNA. DNA probes representing several regions of the viral RNA were expressed in E. coli. Immunoprecipitation of cell proteins revealed the presence of TVMV-related polypeptides; the results were consistent with the cistron order determined by hybrid-arrested translation experiments. Our proposed cistron order and the sizes in kilodaltons of the corresponding polypeptides are: 5'-25 kDa unidentified protein-53 kDa helper component protein-50 kDa unidentified protein-70 kDa cylindrical inclusion protein-52 kDa nuclear inclusion protein-56 kDa nuclear inclusion protein-32 kDa coat protein-3'.

Purification and characterization of potyvirus helper component.

Helper component (HC) was purified from tobacco vein mottling (TVMV)- and potato virus Y (PVY)- infected tobacco plants by sucrose gradient fractionation followed by affinity chromatography on oligo(dT)-cellulose and by gel electrophoresis. The subunit apparent molecular weights (M(r)) of the purified HCs were 53,000 (53K) and 58K for TVMV and PVY, respectively. Antisera to these purified polypeptides specifically blocked the activity of the homologous HC, as determined by aphid transmission assays, and specifically precipitated 75K products of the cell-free translation of the homologous RNA. The molecular weight of undissociated, biologically active TVMV or PVY HC, as determined by high-pressure liquid chromatography (HPLC)-gel permeation chromatography was found to be between 100K and 150K, suggesting that the active molecule is a dimer.

On the origin of the helper component of tobacco vein mottling virus: translational initiation near the 5' terminus of the viral RNA and termination by UAG codons.

The nature of the polypeptide products encoded by the 5'-terminal region of the RNA of the potyvirus, tobacco vein mottling virus (TVMV), was investigated. Single-stranded DNA probes complementary to either nucleotides 1100-2100 or 2100-2820 from the 5' terminus of the RNA were prepared by subcloning recombinant plasmids in bacteriophage M13. These were hybridized to TVMV RNA, the DNA:RNA hybrids translated in a reticulocyte lysate cell-free translation system (hybrid-arrested translation), and the products analyzed by electrophoresis on SDS-containing polyacrylamide gels. The hybrids produced altered patterns of polypeptides which indicated that the major product, P75, was translated from the 5' terminus of the RNA. The N-terminal portion (molecular weight approximately 35,000) of P75 did not react with antisera to any of the five known potyviral proteins, suggesting the existence of a previously unidentified cistron at the 5' terminus. Translation of additional RNA sequences produced a polypeptide with a molecular weight of 68,000 which was immunoprecipitable by antiserum to TVMV helper component, establishing the coding region for helper component near the 5' terminus. Limited DNA sequence analysis of the region encoding the C terminus of P75 revealed an open reading frame of 369 nucleotides followed by a pair of UAG Codons. Pulse-chase experiments demonstrated that in the first 10 min of incubation, the only polypeptide initiated was P75, suggesting that products downstream from P75 arose from in vitro fragmentation of TVMV RNA and translation of the RNA fragments.

Translational recognition of messenger ribonucleic acid caps as a function of pH.

The degree to which cell-free translation of eukaryotic mRNA is stimulated by the presence of a 5'-terminal 7-methylguanosine-containing cap is affected by a variety of factors including ionic strength, temperature, mRNA concentration, and the type of mRNA. In this report, we show that pH also affects cap dependence. Translation of globin mRNA from which the cap had been enzymatically removed was relatively insensitive to pH compared with capped mRNA. Also, at low pH (6.6-7.2), the cap analogue m7GTP caused little inhibition of globin mRNA translation in a cell-free system whereas at higher pH the degree of inhibition increased. Finally, the overall extent to which globin mRNA translation could be inhibited at saturating concentrations of m7GTP increased with increasing pH. It is also shown that the pKa of the N-1 proton of m7GTP is affected by mono- and divalent cations. At the K+ and Mg2+ concentrations optimal for cell-free translation, the pKa is approximately 7.4. Since the pH optimum for translation is near 7.6, both keto and enolate forms of m7G are present in appreciable amounts. One interpretation for the observed change in cap dependence with pH is that only the enolate form of m7G is recognized by the cap-binding protein.

Molecular cloning of DNA complementary to tobacco vein mottling virus RNA.

RNA isolated from tobacco vein mottling virus (TVMV) was used as a template for avian myeloblastosis virus RNA-dependent DNA polymerase, primed with oligo(dT). The largest single-stranded cDNA synthesized was 10 kb, the same as the viral RNA. This material was converted to double-stranded cDNA with Escherichia coli DNA polymerase I and digested with restriction endonuclease HindIII. The cDNA fragments were ligated to HindIII-digested plasmid pBR322 and the product used to transform E. coli strain DG-75. Clones containing recombinant plasmids were selected by ampicillin resistance, and those containing TVMV RNA sequences were selected by colony hybridization with a single-stranded cDNA probe. Four different sizes of recombinant plasmid were reproducibly observed. The inserted DNA portion could be excised in each case with HindIII. The lengths of inserted DNA were 3.0, 1.85, 1.1, and 0.72 kb. A similar procedure was used with PstI-digested cDNA and pBR322. A single type of recombinant plasmid, containing a DNA insertion of 1.85 kb, was reproducibly observed. Hybridization with TVMV RNA confirmed that the five inserted DNA segments were derived from the viral RNA. Hybridization of each recombinant plasmid with the others established that each of the cloned HindIII fragments was unique and that one of them overlapped the cloned PstI fragment. The cloned cDNA fragments were ordered by establishing a restriction map of the cDNA. Together the cloned cDNA fragments account for over 80% of the viral genome.

Cell-free translation of tobacco vein mottling virus RNA. II. Immunoprecipitation of products by antisera to cylindrical inclusion, nuclear inclusion, and helper component proteins.

The genomic RNA of tobacco vein mottling virus (TVMV) was translated in a cell-free system derived from rabbit reticulocytes. Antisera against TVMV coat protein, TVMV cylindrical inclusions, the helper component required for aphid transmission of TVMV, and the 49- and 54-kd nuclear inclusion proteins of tobacco etch virus (TEV) were used to characterize the translational products. Each of the five antisera precipitated a distinctive pattern of polypeptides. Specificity of immunoprecipitation was shown by competing with the various proteins to which antisera were raised and by sequentially precipitating with two different antisera. These experiments showed that the five antisera define five different nonoverlapping regions of the TVMV genome.

A polypeptide which reverses cap analogue inhibition of cell-free protein synthesis. Purification and binding to capped oligonucleotides.

An assay was developed to detect the component which recognizes the methylated 5'-terminus of messenger RNA (cap) during initiation of translation. Globin mRNA translation in a reticulocyte cell-free system was partially inhibited with cap analogues, and protein fractions were added to the system in an attempt to reverse inhibition. Such an activity was detected in the 500 mM KCl extract of rabbit reticulocyte ribosomes. The activity (cap analogue inhibition reversal) was purified 800-fold by precipitation with ammonium sulfate saturation, batchwise chromatography on DEAE-cellulose, centrifugation on sucrose gradients containing 100 and 500 mM KCl, and column chromatography on DEAE-cellulose and phosphocellulose. At some stages multiple peaks of activity were detected. Electrophoretic analysis of the final preparation revealed a single polypeptide of 24,000 daltons, making it likely that it is the same as the cap-binding protein detected by Sonenberg et al. (Sonenberg, N., Morgan, M. A., Merrick, W. C., and Shatkin, A. J. (1978) Proc. Natl. Acad. Sci. U. S. A. 75, 4843-4847), using a cross-linking assay. Direct binding of purified fractions of cap analog inhibition reversal factor to capped oligonucleotides of the form m7Gppp(Np)6-8G[32P]Cp could be demonstrated by both gel filtration on Sephadex G-50 and nitrocellulose membrane filtration. Binding was stimulated by MgCl2 and nucleoside triphosphates. An approximate association constant between oligonucleotide and protein of 3 X 10(8) M-1 was obtained.

Cell-free translation of tobacco vein mottling virus RNA.

Tobacco vein mottling virus (TVMV), a member of the potyvirus group, was translated in a rabbit reticulocyte cell-free system. The RNA was approximately 5% as active as rabbit globin mRNA in directing protein synthesis. Translation was stimulated approximately 15% by the cap analog pppm7G, a phenomenon which has also been observed with uncapped viral RNAs. Treatment with NaIO4 and NaB(3H]4 under conditions which label the caps of globin and ovalbumin mRNA failed to produce labeled cap structures in TVMV RNA. Approximately 20 polypeptides, ranging from 20,000 to 100,000 daltons, were produced in the reticulocyte system programmed with TVMV RNA. The predominant species (P75) had a molecular weight of 75,000. The same major polypeptides were produced in the wheat germ system, but there was relatively greater synthesis of the smaller polypeptides. Incubation of the reticulocyte system in the presence of cycloheximide following an initial period of synthesis failed to cause breakdown of larger polypeptides into smaller polypeptides. Preincubation of TVMV RNA with the reticulocyte system before addition of labeled amino acids caused greater synthesis of the smaller polypeptides, suggesting that they are derived from fragments of the RNA produced during the incubation. Translation of TVMV RNA which had been bound to oligo(dT)-cellulose yielded nearly the same spectrum of polypeptides, but synthesis of polypeptides smaller than 52,000 daltons was reduced. At low ionic strength, only polypeptides of 52,000 daltons and smaller were synthesized. At high ionic strength, P75 was essentially the only product synthesized. Antibody to whole virus precipitated many of the polypeptides, including P75, suggesting that they contain the coat protein sequence. No polypeptide with the electrophoretic mobility of authentic coat protein, however, was precipitated. Comparison of partial proteolytic digests of P75 and authentic coat protein provided further evidence for sequence homology.