Genomic discovery of an evolutionarily programmed modality for small-molecule targeting of an intractable protein surface

The vast majority of intracellular protein targets are refractory toward small-molecule therapeutic engagement, and additional therapeutic modalities are needed to overcome this deficiency. Here, the identification and characterization of a natural product, WDB002, reveals a therapeutic modality that dramatically expands the currently accepted limits of druggability. WDB002, in complex with the FK506-binding protein (FKBP12), potently and selectively binds the human centrosomal protein 250 (CEP250), resulting in disruption of CEP250 function in cells. The recognition mode is unprecedented in that the targeted domain of CEP250 is a coiled coil and is topologically featureless, embodying both a structural motif and surface topology previously considered on the extreme limits of "undruggability" for an intracellular target. Structural studies reveal extensive protein-WDB002 and protein-protein contacts, with the latter being distinct from those seen in FKBP12 ternary complexes formed by FK506 and rapamycin. Outward-facing structural changes in a bound small molecule can thus reprogram FKBP12 to engage diverse, otherwise "undruggable" targets. The flat-targeting modality demonstrated here has the potential to expand the druggable target range of small-molecule therapeutics. As CEP250 was recently found to be an interaction partner with the Nsp13 protein of the SARS-CoV-2 virus that causes COVID-19 disease, it is possible that WDB002 or an analog may exert useful antiviral activity through its ability to form high-affinity ternary complexes containing CEP250 and FKBP12.

Genome-wide siRNA screen for modulators of cell death induced by proteasome inhibitor bortezomib

Multiple pathways have been proposed to explain how proteasome inhibition induces cell death, but mechanisms remain unclear. To approach this issue, we performed a genome-wide siRNA screen to evaluate the genetic determinants that confer sensitivity to bortezomib (Velcade (R); PS-341). This screen identified 100 genes whose knockdown affected lethality to bortezomib and to a structurally diverse set of other proteasome inhibitors. A comparison of three cell lines revealed that 39 of 100 genes were commonly linked to cell death. We causally linked bortezomib-induced cell death to the accumulation of ASF1B, Myc, ODC1, Noxa, BNIP3, Gadd45alpha, p-SMC1A, SREBF1, and p53. Our results suggest that proteasome inhibition promotes cell death primarily by dysregulating Myc and polyamines, interfering with protein translation, and disrupting essential DNA damage repair pathways, leading to programmed cell death.

Abstract A70: A genome-wide siRNA screen for modulators of cell death induced by the proteasome inhibitor bortezomib

Multiple pathways have been proposed as the mechanism by which proteasome inhibition induces cell death. To clarify their relative importance, we performed a genome-wide siRNA screen to evaluate the genetic determinants that confer sensitivity of the HCT-116 colon cancer cell line to bortezomib (VELCADE®, PS-341). The screen identified 100 genes whose knock-down affects the lethality of bortezomib. From this list, the accumulation of the proteins ASF1B, Myc, ODC1, PMAIP1 (Noxa), BNIP3, Gadd45α, p-SMC1A, SREBF1, and p53 by proteasome inhibition was linked to the induction of cell death. Fifty-nine genes in the A375 melanoma cell line and 56 genes in the HeLa cervical cancer cell line showed similar interactions with bortezomib to those seen in HCT-116 and a subset of 39 genes were common to all three cell lines. Finally, knockdown of these 100 genes in HCT-116 cells similarly affected their responsiveness to a structurally diverse set of proteasome inhibitors. Our results suggest that proteasome inhibition promotes cell death primarily by dysregulating Myc and polyamines, interfering with protein translation, and disrupting essential DNA damage repair pathways, leading to programmed cell death.

Citation Information: Mol Cancer Ther 2009;8(12 Suppl):A70.

Choice and manipulation of retroviral vectors

ABSTACT: During the past few years, retroviral vectors have become a very important and widely used means of gene transfer. In the laboratory, their use has expanded the capabilities of investigators to perform important experiments that have solved previously unanswerable biological questions. Retroviral vectors exploit the inherent capacity of retrovkuses to transfer genetic material stably into a cell's genome and subsequently express it in a manner that is generally not detrimental to the host-cell. Initial entry of genetic material into cells via retroviral vectors is so efficient that it permits successful transfer of genes into limiting numbers of cells, such as hematopoietic stem cells in explanted bone marrow, and transfer approaching 100% in tissue-culture experiments. In comparison, chemical and electrical means of gene transfer require large initial numbers of cells and, under optimal circumstances, function at efficiencies several orders of magnitude lower than retroviral vectors, and only in a limited number of specific cell types.

Identification of leptin-induced transcripts in the mouse hypothalamus

Long-form leptin receptor (OB-R(L)) is a signal-transducing member of the cytokine receptor superfamily that is essential for mediating the effects of leptin on mammalian body weight homeostasis. At present, the range of transcriptional targets responsive to OB-R(L) activation, and consequently, the likely mediators of leptin action, remain undefined. In this report, we have used cDNA subtractive hybridization to identify transcripts induced by leptin in immortalized hypothalamic neurons expressing OB-R(L). Differential expression of the identified transcripts in these cells was confirmed by both array technology and Northern blotting. In situ hybridization studies indicate that these transcripts are expressed in the mouse central nervous system, including nuclei of the hypothalamus that coexpress OB-R(L). Comparative in situ analysis of slices of hypothalami generated from control and leptin-injected ob/ob mice demonstrates that a subset of the identified transcripts is induced in vivo after leptin injection. The potential role of the proteins encoded by these transcripts in mediating the effects of leptin on body weight and energy homeostasis are discussed.

Tyrosine phosphorylation of tub and its association with Src homology 2 domain-containing proteins implicate tub in intracellular signaling by insulin

A mutation in the tub gene leads to maturity-onset obesity, insulin resistance, and progressive retinal and cochlear degeneration in mice. tub is a member of a growing family of genes that encode proteins of unknown function that are remarkably conserved across species. The absence of obvious transmembrane domain(s) or signal sequence peptide motif(s) suggests that Tub is an intracellular protein. Additional sequence analysis revealed the presence of putative tyrosine phosphorylation motifs and Src homology 2 (SH2)-binding sites. Here we demonstrate that in CHO-IR cells, transfected Tub is phosphorylated on tyrosine in response to insulin and insulin-like growth factor-1 and that in PC12 cells, insulin but not EGF induced tyrosine phosphorylation of endogenous Tub. In vitro, Tub is phosphorylated by purified insulin receptor kinase as well as by Abl and JAK 2 but not by epidermal growth factor receptor and Src kinases. Furthermore, upon tyrosine phosphorylation, Tub associated selectively with the SH2 domains of Abl, Lck, and the C-terminal SH2 domain of phospholipase Cgamma and insulin enhanced the association of Tub with endogenous phospholipase Cgamma in CHO-IR cells. These data suggest that Tub may function as an adaptor protein linking the insulin receptor, and possibly other protein-tyrosine kinases, to SH2-containing proteins.

The major dog allergens, Can f 1 and Can f 2, are salivary lipocalin proteins: cloning and immunological characterization of the recombinant forms

Canis familiaris allergen 1 (Can f 1) and Canis familiaris allergen 2 (Can f 2) are the two major allergens present in dog dander extracts. We now report the isolation of cDNAs encoding both proteins and present their nucleotide and deduced amino acid sequences. Can f 1, produced by tongue epithelial tissue, has homology with the von Ebner's gland (VEG) protein, a salivary protein not previously thought to have allergenic properties. Can f 2, produced by tongue and parotid gland, has homology with mouse urinary protein (MUP), a known allergen. Both VEG protein and MUP are members of the lipocalin family of small ligand-binding proteins. Recombinant forms of Can f 1 and Can f 2 were produced and tested for immunoglobulin E (IgE) reactivity. Among dog-allergic subjects, 45% had IgE directed exclusively to rCan f 1, and 25% had IgE to both rCan f 1 and rCan f 2. In addition, both recombinant proteins were able to cross-link IgE and elicit histamine release from peripheral blood leucocytes in vitro. These findings confirm that Can f 1 and Can f 2 are major and minor dog allergens, respectively, and demonstrate that recombinant forms of dog allergens retain at least some IgE-binding epitopes.

Constitutive and impaired signaling of leptin receptors containing the Gln --> Pro extracellular domain fatty mutation

Leptin (OB), an adipocyte-secreted circulating hormone, and its receptor (OB-R) are key components of an endocrine loop that regulates mammalian body weight. In this report we have analyzed signal transduction activities of OB-R containing the fatty mutation [OB-R(fa)], a single amino acid substitution at position 269 (Gln --> Pro) in the OB-R extracellular domain that results in the obese phenotype of the fatty rat. We find that this mutant receptor exhibits both ligand-independent transcriptional activation via interleukin 6 and hematopoietin receptor response elements and ligand-independent activation of signal transducer and activator of transcription (STAT) proteins 1 and 3. However, OB-R(fa) is unable to constitutively activate STAT5B and is highly impaired for ligand induced activation of STAT5B compared with OB-R(wt). Introduction of the fatty mutation into a OB-R/G-CSF-R chimera generates a receptor with constitutive character that is similar but distinct from that of OB-R(fa). Constitutive mutant OB-R(fa) receptor signaling is repressed by coexpression of OB-R(wt). The implications of an extracellular domain amino acid substitution generating a cytokine receptor with a partially constitutive phenotype are discussed both in terms of the mechanism of OB-R triggering and the biology of the fatty rat.

Evidence that the diabetes gene encodes the leptin receptor: identification of a mutation in the leptin receptor gene in db/db mice

OB-R is a high affinity receptor for leptin, an important circulating signal for the regulation of body weight. We identified an alternatively spliced transcript that encodes a form of mouse OB-R with a long intracellular domain. db/db mice also produce this alternatively spliced transcript, but with a 106 nt insertion that prematurely terminates the intracellular domain. We further identified G --> T point mutation in the genomic OB-R sequence in db/db mice. This mutation generates a donor splice site that converts the 106 nt region to a novel exon retained in the OB-R transcript. We predict that the long intracellular domain form of OB-R is crucial for initiating intracellular signal transduction, and as a corollary, the inability to produce this form of OB-R leads to the severe obese phenotype found in db/db mice.

Molecular basis for interaction of the protein tyrosine kinase ZAP-70 with the T-cell receptor

The crystal structure of the tandem SH2 domains of human ZAP-70 in complex with a peptide derived from the zeta-subunit of the T-cell receptor reveals an unanticipated interaction between the two domains. A coiled coil of alpha-helices connects the two SH2 domains, producing an interface that constitutes one of the two critical phosphotyrosine binding sites. These and other unique features provide the molecular basis for highly selective association of ZAP-70 with the T-cell receptor.

Interaction of p72syk with the gamma and beta subunits of the high-affinity receptor for immunoglobulin E, Fc epsilon RI

Activation of protein tyrosine kinases is one of the initial events following aggregation of the high-affinity receptor for immunoglobulin E (Fc epsilon RI) on RBL-2H3 cells, a model mast cell line. The protein tyrosine kinase p72syk (Syk), which contains two Src homology 2 (SH2) domains, is activated and associates with phosphorylated Fc epsilon RI subunits after receptor aggregation. In this report, we used Syk SH2 domains, expressed in tandem or individually, as fusion proteins to identify Syk-binding proteins in RBL-2H3 lysates. We show that the tandem Syk SH2 domains selectively associate with tyrosine-phosphorylated forms of the gamma and beta subunits of Fc epsilon RI. The isolated carboxy-proximal SH2 domain exhibited a significantly higher affinity for the Fc epsilon RI subunits than did the amino-proximal domain. When in tandem, the Syk SH2 domains showed enhanced binding to phosphorylated gamma and beta subunits. The conserved tyrosine-based activation motifs contained in the cytoplasmic domains of the gamma and beta subunits, characterized by two YXXL/I sequences in tandem, represent potential high-affinity binding sites for the dual SH2 domains of Syk. Peptide competition studies indicated that Syk exhibits a higher affinity for the phosphorylated tyrosine activation motif of the gamma subunit than for that of the beta subunit. In addition, we show that Syk is the major protein in RBL-2H3 cells that is affinity isolated with phosphorylated peptides corresponding to the phosphorylated gamma subunit motif. These data suggest that Syk associates with the gamma subunit of the high-affinity receptor for immunoglobulin E through an interaction between the tandem SH2 domains of SH2 domains of Syk and the phosphorylated tyrosine activation motif of the gamma subunit and that Syk may be the major signaling protein that binds to Fc epsilon RI tyrosine activation motif of the gamma subunit and that Syk may be the major signaling protein that binds to Dc epsilon tyrosine activation motifs in RBL-2H3 cells.

Potential therapeutic recombinant proteins comprised of peptides containing recombined T cell epitopes

The complete primary structure of Fel d I2 has been determined and shown to be comprised of two separate polypeptide chains (designated chain 1 and 2). Overlapping peptides covering the entire sequence of both chains of Fel d I have been used to map the major areas of human T cell reactivity. The present study describes three non-contiguous T cell reactive regions of < 30 aa in length that were assembled in all six possible configurations using PCR and recombinant DNA methods. These six recombinant proteins comprised of defined non-contiguous T cell epitope regions artificially combined into single polypeptide chains have been expressed in E. coli, highly purified, and examined for their ability to bind to human cat-allergic IgE and for human T cell reactivity. Several of these recombined T cell epitope-containing polypeptides exhibit markedly reduced IgE binding as compared to the native Fel d I. Importantly, the human T cell reactivity to individual T cell epitope-containing regions is maintained even though each was placed in an unnatural position as compared to the native molecule. In addition, T cell responses to potential junctional epitopes were not detected. It was also demonstrated in mice that s.c. injection of T cell epitope-containing polypeptides inhibits the T cell response to the individual peptides upon subsequent challenge in vitro. Thus, these recombined T cell epitope-containing polypeptides, which harbor multiple T cell reactive regions but have significantly reduced reactivity with allergic human IgE, constitute a novel potential approach for desensitization to important allergens.

Recombinant Fel d.I: Expression, purification, IgE binding and reaction with cat-allergic human T cells

This study describes the properties of the two recombinantly expressed polypeptide chains of Fel d I, the major allergen produced by the domestic cat (Felis domesticus). An inframe linker encoding polyhistidine has been added to the 5' ends of the Fel d I chains 1 and 2 cDNAs to facilitate purification using Ni2+ ion affinity chromatography. This method provides high yields in a single step of rchain 1 and rchain 2 of Fel d I with a > 90% level of purity. Polymerase chain reaction (PCR) methods were used to introduce a thrombin cleavage site (LVPR decreases GS) at the N-terminus of both chains. Thrombin cleavage of rchain 1 and rchain 2 followed by HPLC purification of the cleavage products allowed the isolation of each recombinant chain with only two additional residuals (GS) at the N-terminus of the native sequence. Amino acid sequencing analysis of the N-terminus and mass spectrometry of these polypeptides demonstrated that they are highly pure and full-length. Direct ELISA assays showed that IgE from cat-allergic patients binds to both rchain 1 and rchain 2 of Fel d I, demonstrating that both these chains contribute to the allergenicity of this heterodimeric protein. An examination of the reactivity of T cells derived from cat-allergic patients revealed that both polypeptide chains contribute to the T cell response to this allergen. Consequently, it is concluded that the immunological response to Fel d I is composed of a reaction at both the B and T cell level to each of the two chains that constitute the native allergen.

Immunoaffinity chromatography of recombinant Amb a I in the presence of a denaturing agent

Recombinant proteins expressed in E. coli are often sequestered into inclusion bodies and require the use of denaturing agents in order to solubilize them. The recombinant form of Amb a I, the major allergen from short ragweed pollen, is one such protein. In some cases solubility can be maintained after the removal of the denaturing agent, particularly if the protein can be folded into its native conformation. However, not all proteins refold readily and after the removal of the denaturing agent the proteins will reaggregate and/or precipitate. In the case of Amb a I, the recombinant protein stays in solution at low concentrations but aggregates with itself and other proteins. The recombinant Amb a I is not expressed at high levels and may be toxic to E. coli. Therefore, isolation from a complex mixture of E. coli proteins was necessary. Monoclonal antibodies which recognize the denatured form of Amb a I were available, allowing for immunoaffinity purification. However, because the protein was not monomeric, this chromatographic technique did not provide an improvement in the purity level when run in normal buffer solutions. Analysis of one monoclonal antibody's stability to urea indicated it could tolerate the presence of 2 M urea and recover full activity. Use of this antibody as an immunoaffinity reagent in a column run in 2 M urea, which minimized aggregation of the E. coli produced proteins, gave a high degree of purification of recombinant Amb a I in one step. This illustrates the potential for the use of denaturing and other solubilizing agents in immunoaffinity chromatography of recombinant proteins.

Expression and genomic structure of the genes encoding FdI, the major allergen from the domestic cat

The genes encoding chain 1 (Ch1) and chain 2 (Ch2) of the major allergen of the domestic cat, Felis domesticus I, have been analyzed by genomic cloning and by polymerase chain reaction (PCR). Ch1 has two potential leader sequences, designated A and B. Analysis of a genomic clone encoding Ch1 demonstrated that one structural gene contains sequences corresponding to both leaders, which utilize different Met start codons. PCR analysis showed that genes encoding Ch1 and Ch2 are co-expressed in both the salivary glands and the skin, and that leader sequence A of Ch1 is utilized preferentially in both tissues. Ch2 was shown to have two dominant forms that are differentially expressed in the aforementioned tissues. The long form (Ch2L), composed of 92 amino acids (aa), is preferentially expressed in the salivary glands, while the short form (Ch2S), composed of 90 aa, is preferentially expressed in the skin. There is minor sequence polymorphism in both forms of Ch2. A genomic clone for Ch2 only contained sequences for Ch2S, suggesting that Ch2L is encoded by an exon not contained within this genomic clone.

Amino acid sequence of Fel dI, the major allergen of the domestic cat: protein sequence analysis and cDNA cloning

The complete primary structure of Fel dI (International Union of Immunological Societies nomenclature), the major allergen produced by the domestic cat, Felis domesticus, was determined by protein sequence analysis and cDNA cloning. Protein sequencing of Fel dI from an immunoaffinity-purified extract of house dust revealed that the allergen is composed of two polypeptide chains. Degenerate oligonucleotides derived from the protein sequence were used in polymerase chain reaction amplification of cat salivary gland cDNA to demonstrate that the two chains are encoded by different genes. Chain 1 of Fel dI shares amino acid homology with rabbit uteroglobin, while chain 2 is a glycoprotein with N-linked oligosaccharides.

Complete sequence of the allergen Amb alpha II. Recombinant expression and reactivity with T cells from ragweed allergic patients

This study defines the complete primary structure of Amb alpha II, an important allergen produced by short ragweed (Ambrosia artemisiifolia). The deduced amino acid sequence derived from the cDNA indicates that Amb alpha II shares approximately 65% sequence identity with the Amb alpha I multigene family of allergens. Full-length cDNA encoding Amb alpha I.1 and Amb alpha II have been expressed in E. coli and purified. An in-frame linker encoding polyhistidine has been added to the 5' end of the cDNA to facilitate purification using Ni2+ ion affinity chromatography, yielding greater than 90% pure recombinant protein in a single step. T cells from patients allergic to ragweed proliferate in response to pollen extract as well as purified recombinant Amb alpha I.1 and Amb alpha II. T cell lines established using either Amb alpha I.1 or II as the stimulating Ag exhibit a high level of cross-reactivity to both proteins. This result is entirely consistent with the extensive primary sequence identity shared by these two proteins. These data suggest that allergic humans recognize shared T cell epitopes on these two related molecules.

Complete sequence of the allergen Amb alpha II. Recombinant expression and reactivity with T cells from ragweed allergic patients

This study defines the complete primary structure of Amb alpha II, an important allergen produced by short ragweed (Ambrosia artemisiifolia). The deduced amino acid sequence derived from the cDNA indicates that Amb alpha II shares approximately 65% sequence identity with the Amb alpha I multigene family of allergens. Full-length cDNA encoding Amb alpha I.1 and Amb alpha II have been expressed in E. coli and purified. An in-frame linker encoding polyhistidine has been added to the 5' end of the cDNA to facilitate purification using Ni2+ ion affinity chromatography, yielding greater than 90% pure recombinant protein in a single step. T cells from patients allergic to ragweed proliferate in response to pollen extract as well as purified recombinant Amb alpha I.1 and Amb alpha II. T cell lines established using either Amb alpha I.1 or II as the stimulating Ag exhibit a high level of cross-reactivity to both proteins. This result is entirely consistent with the extensive primary sequence identity shared by these two proteins. These data suggest that allergic humans recognize shared T cell epitopes on these two related molecules.

Domains of human c-myc protein required for autosuppression and cooperation with ras oncogenes are overlapping

Amino acids 106 to 143 and 354 to 433 of the human c-myc protein (439 amino acids) were shown to be required for the protein to suppress c-myc gene transcription and were found to exactly overlap with those necessary for c-myc to cooperate with ras oncogenes in the transformation of rat embryo fibroblasts. The essential carboxyl-terminal region harbors structural motifs (a basic region, a helix-loop-helix motif, and a "leucine zipper"), which, in other proteins, can mediate dimerization and sequence-specific DNA binding.

Advanced mammalian gene transfer: high titre retroviral vectors with multiple drug selection markers and a complementary helper-free packaging cell line

We report the development of an advanced system for transfer and expression of exogenous genes in mammalian cells based on Moloney murine leukemia virus (Mo MuLV). Extensive deletion/mutagenesis analysis to identify cis-acting signals involved in virus transmission has led to the design of a family of novel, highly efficient retroviral vectors and a partner helper-free packaging cell line. The pBabe retroviral vector constructs transmit inserted genes at high titres and express them from the Mo MuLV Long Terminal Repeat (LTR). Each of these vectors has been constructed with one of four different dominantly acting selectable markers, allowing the growth of infected mammalian cells in the presence of G418, hygromycin B, bleomycin/phleomycin or puromycin, respectively. The high titre ecotropic helper free packaging cell line, omega E, was designed in conjunction with the pBabe vectors to reduce the risk of generation of wild type Mo MuLV via homologous recombination events. The omega E cell line was generated with separate gagpol and ecotropic env expression constructs with minimal sequence overlap and decreased sequence homology achieved by 'codon wobbling'. Homologous env coding sequences were deleted from the pBabe vectors without diminishing recombinant vector titre. Together, the pBabe vectors and omega E cell line should prove useful in experiments where highest frequencies of gene transfer, or concomitant expression of several different genes within a single cell are required with minimal risk of helper virus contamination.

Continued expression of HPV-16 E7 protein is required for maintenance of the transformed phenotype of cells co-transformed by HPV-16 plus EJ-ras

The close association between HPV-16 and cervical cancer implies some role for the virus in development of this cancer. Recent studies have shown that the HPV-16 E7 gene encodes the major transforming activity of the virus in baby rat kidney (BRK) cell transformation assays. To investigate the requirement for continued E7 expression in BRK cells transformed by HPV-16 E7 plus EJ-ras, we have developed a system for inducible expression of the E7 gene. The studies reported here show that continued expression of the HPV-16 E7 gene is required for maintenance of the transformed phenotype in these cells. The implications these observations bear on the role of the E7 gene in cervical carcinoma are discussed.

Analysis of HLA-DR glycoproteins by DNA-mediated gene transfer. Definition of DR2 beta gene products and antigen presentation to T cell clones from leprosy patients

We have used DNA-mediated gene transfer to express HLA class II molecules in mouse L cells for serological, biochemical, and functional analysis. cDNA clones encoding the DR2 beta a and DR2 beta b products of the DR2Dw2 haplotype were subcloned into a mouse Moloney leukemia virus-based expression vector (pJ4) and transfected separately into mouse L cells together with a HLA-DR alpha/pJ4 construct. These transfectants have allowed differential analysis of the two DR2 beta products in a manner normally prohibited by the concomitant expression seen in B cells. Two-dimensional SDS-PAGE analysis of the transfectants defines the more acidic beta chain as the product of the DR2 beta a sequence, and the more basic chain as the product of the DR2 beta b sequence. The LDR2a transfectants present antigen efficiently to M.leprae-specific T cell clones and are capable of presenting synthetic peptide, 65-kD recombinant mycobacterial antigen and M.leprae. Of the DR2Dw2-restricted T cell clones we have tested, all use the DR2 beta a chain as their restriction element. Inhibition studies with mAbs demonstrate the dependence of presentation by the transfectant on class II and CD4, while mAbs against LFA-1, which substantially inhibit presentation by B-lymphoblastoid cell lines, do not inhibit transfectant presentation.

Behavior of myc and ras oncogenes in transformation of rat embryo fibroblasts

The requirements for transformation of rat embryo fibroblasts (REFs) by transfected ras and myc oncogenes were explored. Under conditions of dense monolayer culture, neither oncogene was able to transform REFs on its own. However, the introduction of a ras oncogene together with a selectable neomycin resistance marker into REFs allowed killing of the normal nontransfected cells and the outgrowth of colonies of ras transformants, 10% of which survived crisis and became tumorigenic. These cells expressed greater than 10-fold-higher levels of ras p21 than tumorigenic cells cotransfected with ras and myc oncogenes. The myc oncogene similarly was unable to induce tumorigenic conversion of REFs unless especially refractile colonies of oncogene-bearing cells, produced by use of a cotransfected selectable marker, were picked and subcultured. Tumorigenic conversion of REFs by single transfected oncogenes appears to require special culture conditions and high levels of gene expression.