Responses of mouse cell lines transformed by various means to histidinol/cytosine arabinoside treatment

The responses of the phenotypically-normal BALB/c 3T3 mouse cell line A31 and three of its transformed derivatives, BP-A31 (a benzo(a)pyrene transformant), M-A31 and SV-A31 (Moloney murine sarcomavirus and SV-40 virus transformants, respectively) to L-histidinol and to L-histidinol/cytosine arabinoside(AraC) combinations were investigated. As observed in comparable studies with other normal cell lines, prior treatment of the A31 parental line with L-histidinol led to significant protection from AraC-mediated toxicity. However, in contrast to previous analyses of a number of other transformed cell lines which showed enhanced AraC-mediated toxicity upon prior exposure to L-histidinol, the three types of A31 transformants were afforded varying degrees of protection from AraC when treated with histidinol. These findings support the concept that transformed, tumorigenic cells may retain certain growth properties of their nontumorigenic parental cells.     

Inhibition of voltage-dependent Na+ current in cell-fusion hybrids containing activated c-Ha-ras

Summary The electrophysiological properties of EJ (human bladder carcinoma), GM2291 (human fetal lung fibroblast), and of three hybrid cell lines obtained from their cell fusion were investigated using the patch-clamp technique. GM2291 cells, which are nontumorigenic, express voltage-dependent Na⁺ channels. The pharmacology and gating properties of the Na⁺ channels in GM2291 cells are distinct from neuronal and cardiac Na⁺ channels. EJ cells, which are tumorigenic and contain activated c-Ha-ras, express inward rectifier K⁺ channels. The three cell-fusion hybrid lines, named 145 (nontumorigenic), 145L (non-tumorigenic but morphologically altered), and 147TR2 (fully tumorigenic segregant), have been previously shown to express levels of activated c-Ha-ras similar to those of the EJ parental line. Voltage-dependent Na⁺ channels were observed in none of the hybrid cell lines, while inward rectifier K⁺ channels were observed in each of the hybrid cell lines. The possibility that c-Ha-ras inhibits expression of a voltage-dependent Na⁺ channel is discussed.     

Comparative analysis of involvement of UGT1 and UGT2 splice variants of UDP-galactose transporter in glycosylation of macromolecules in MDCK and CHO cell lines

Nucleotide sugar transporters deliver nucleotide sugars into the Golgi apparatus and endoplasmic reticulum. This study aimed to further characterize mammalian UDP-galactose transporter (UGT) in MDCK and CHO cell lines. MDCK-RCA^r and CHO-Lec8 mutant cell lines are defective in UGT transporter, although they exhibit some level of galactosylation. Previously, only single forms of UGT were identified in both cell lines, UGT1 in MDCK cells and UGT2 in CHO cells. We have identified the second UGT splice variants in CHO (UGT1) and MDCK (UGT2) cells. Compared to UGT1, UGT2 is more abundant in nearly all examined mammalian tissues and cell lines, but MDCK cells exhibit different relative distribution of both splice variants. Complementation analysis demonstrated that both UGT splice variants are necessary for N- and O-glycosylation of proteins. Both mutant cell lines produce chondroitin-4-sulfate at only a slightly lower level compared to wild-type cells. This defect is corrected by overexpression of both UGT splice variants. MDCK-RCA^r mutant cells do not produce keratan sulfate and this effect is not corrected by either UGT splice variant, overexpressed either singly or in combination. Here we demonstrate that both UGT splice variants are important for glycosylation of proteins. In contrast to MDCK cells, MDCK-RCA^r mutant cells may possess an additional defect within the keratan sulfate biosynthesis pathway.     

Differential methylation of mitochondrial and nuclear DNA in cultured mouse, hamster and virus-transformed hamster cells. In vivo and in vitro methylation

Information has been lacking as to whether mitochondrial DNA of animal cells is methylated. The methylation patterns of mitochondrial and nuclear DNAs of several mammalian cell lines have therefore been compared by four methods: (1) in vivo transfer of the methyl group from [methyl-³H]methionine; (2) in vivo incorporation of [³²P]orthophosphate and a combination of (1) and (2); (3) in vivo incorporation of [³H]deoxycytidine; (4) in vitro methylation of DNAs with ³H-labeled S-adenosylmethionine as methyl donor and DNA methylase preparations from L cell nuclei. The cell lines were mouse L cells, $\text{BHK21}\text{C13}$, $\text{C13}\text{B4}$ (baby hamster kidney cells transformed by the Bryan strain of Rouse sarcoma virus), and PyY (BHK cells transformed by polyoma virus). DNA bases were separated chromatographically, using 5-methylcytosine, 6-methylaminopurine and, in some cases, 7-methylguanine as markers.Mitochondrial DNA was found to be significantly less methylated than nuclear DNA with respect to 5-methylcytosine in all cell types studied and by all methods used. The relative advantages and disadvantages of each method have been discussed. The level of 5-methylcytosine in mitochondrial DNA as compared with that in nuclear DNA was estimated as one-fourth to one-fourteenth in various cell lines. The estimated 5-methylcytosine content per circular mitochondrial DNA molecule (mol. wt 10 × 10⁶) was about 12 methylcytosine residues for L cells and 24, 30 and 36 methylcytosine residues for BHK, B4 and PyY cells, respectively. Relative to cytosine residues, the estimate was one 5-methylcytosine per 500 cytosine residues of mitochondrial DNA and one 5-methylcytosine per 36 cytosine residues of nuclear DNA from L-cells. The values for methylcytosine of mitochondrial DNA are presumed to be maximal. PyY cells as compared with other cells had the highest methylcytosine content of both mitochondrial and nuclear DNA as estimated by method (3). No methylation of nuclear DNA was observed in confluent L cells.Evidence for the presence of DNA methylase activity associated with mitochondrial fractions was obtained. This activity could be distinguished from other cellular DNA methylase activity by differential response to mercaptoethanol. Radioactivity from ³H-labeled S-adenosylmethionine was found only in 5-methyl-cytosine of DNA.     

Alkaline phosphatase activity and the regulation of growth in transformed mammalian cells

The relationship between alkaline phosphatase activity and cell growth has been studied in hamster cells transformed by different carcinogens. About 90% of normal hamster embryo cells were constitutively positive for alkaline phosphatase activity (AP⁺). However, there were no AP⁺ cells in cell lines transformed after treatment with the chemical carcinogens dimethylnitrosamine or 4-nitro-quinoline-N-oxide and 0.02% and 4% AP⁺ cells in cell lines transformed by polyoma virus or Simian virus 40. The glucocorticoid hormone, prednisolone, induced alkaline phosphatase activity in 12% and 44% of the enzyme-negative (AP^?) cells in cell lines transformed by polyoma or Simian virus 40, but this hormone did not induce alkaline phosphatase activity in AP^? cells from cell lines transformed after treatment with the chemical carcinogens. Treatment of polyoma transformed AP^? cells with the mutagen N-methyl-N′-nitro-N-nitro-soguanidine produced AP⁺ cells, whereas no AP⁺ cells were found after mutagen treatment of AP^? cells from the chemically transformed cell lines. Studies on spontaneous segregation in the polyoma transformed cell line has shown that AP⁺ cells segregated AP^? cells both in vitro and in vivo, although no spontaneous segregation was observed from AP^? to AP⁺ cells. AP⁺ cells, compared to AP^? cells, showed a decrease in DNA synthesis, cell multiplication, the ability to form colonies in soft agar and tumorogenicity in animals. AP^? cells induced for alkaline phosphatase activity by prednisolone, showed the same growth properties in vitro as uninduced AP^? cells. The decreased cell growth found in AP⁺ cells which were constitutive for alkaline phosphatase activity was therefore not found in the hormone induced AP^? cells. The results indicate that constitutive alkaline phosphatase activity appears to be related to the regulation of cell growth and that AP^? cells have a selective advantage over AP⁺ cells.     

Lactobacillus reuteri DSM 17938 Changes the Frequency of Foxp3+ Regulatory T Cells in the Intestine and Mesenteric Lymph Node in Experimental Necrotizing Enterocolitis

Necrotizing enterocolitis (NEC) is an inflammatory disease of the intestine in premature infants. Lactobacillus reuteri DSM 17938 improves survival and reduces the incidence and severity of NEC in a rodent model. Foxp3⁺ regulatory T cells (Tregs) maintain intestinal homeostasis by controlling inflammation and inducing tolerance. To determine whether there are insufficient numbers of Tregs to control inflammation in NEC and to determine if LR17938 increases the frequency of Tregs, we studied selected groups of newborn Sprague-Dawley rats according to feeding plan: dam±LR17938, formula±LR17938, and NEC±LR17938. NEC was induced by gavage feeding with special formula and exposure to hypoxic conditions. Lymphocytes isolated from ileum, mesenteric lymph nodes (MLN), spleen and thymus were labeled for T cell surface markers (CD3, CD4, CD8) and intracellular Foxp3; and labeled cells were analyzed by flow cytometry. The percentage of CD3⁺ T cells and Foxp3⁺ Tregs in the ileum significantly decreased in pups with NEC, compared to normal controls. Feeding LR17938 to neonatal rats with NEC increased the % of Foxp3⁺ T cells in the ileum while decreasing the percentage of cells in the MLN. Administration of LR17938 to dam-fed rats significantly increased Foxp3⁺Tregs in the ileum as early as day of life (DOL)1 but did not produce an increase in Tregs in formula-fed rats on DOL1. These results suggest that factors in breast milk may enhance the early immunomodulatory effects of LR17938. An anti-inflammatory effect of LR17938 in NEC was associated with the modulation of immune responses and induction and what appears to be migration of Foxp3⁺ Tregs to the diseased gut. Probiotic-facilitated development of Tregs might play an important role in the prevention of NEC.     

A preliminary study of the role of extracellular -5′- nucleotidase in breast cancer stem cells and epithelial-mesenchymal transition

Tumor stem cell theory may well explain a variety of malignant behaviors of tumors. Cells undergoing epithelial-mesenchymal transition (EMT) share many characteristics with tumor stem cells. Our previous studies showed that extracellular -5'- nucleotidase (CD73), one of the important surface markers of mesenchymal stem cells, may promote growth and metastasis of breast cancer cells both in vivo and in vitro. In this study, we assessed breast cancer stem cell (BCSC) markers [acetaldehyde dehydrogenase (ALDH)⁺ and CD44⁺CD24^?] in various breast cancer cell lines with flow cytometry after overexpression (by lentivirus infection) or suppression (by siRNA interference) of CD73. We measured CD73 expression in breast cancer mammospheres with real-time PCR and western blots. Finally, we examined the expression of CD73 and EMT markers in different breast cancer cell lines, as well as in mammary cells (MCF10A) that underwent EMT induced by transforming growth factor beta (TGF-β). We found that CD73 positively correlated with ALDH⁺ or CD44⁺CD24^? subsets of breast cancer cells. CD73 was expressed more in breast cancer mammospheres than in adherent cells. CD73 and mesenchymal marker expression was higher in breast cancer cells with more malignant features, while CD73 was lower in low malignant breast cancer cells with higher epithelial markers. Furthermore, CD73 expression increased during the process of TGF-β-induced EMT. Our results indicate that CD73 may play an important role in BCSCs.     

General organization of the genes specifically involved in the diaminopimelate-lysine biosynthetic pathway of Corynebacterium glutamicum

Summary To obtain animal cell lines carrying nonsense mutations and the corresponding suppressors, we used a supersuppressor selection strategy on the CHO cell line. The wild-type strain is resistant to the aminopterin present in HAT medium (i.e., it is HAT^r) because it contains the enzymes hypoxanthine-guanine phosphoribosyl transferase (HPRT) and thymidine kinase (TK), whereas both HPRT mutants — selected by their resistance to 6-thioguanine (TG^r) — and TK^- mutants — selected by their resistance to 5-bromodeoxyuridine (BrdUrd^r) — are HAT^s. Therefore, from HPRT^- TK^- double nonsense mutants, whose phenotype would be TG^r BrdUrd^r (HAT^s), simultaneous HPRT⁺ TK⁺ double phenotypic revertants could be obtained by selecting HAT^r (TG^s BrdUrd^s) variants carrying the corresponding nonsense supersuppressors. Through ethylmethane sulfonate (EMS) mutagenesis of the CHO cell line we obtained 65 TG^r variants, 53 of which were HAT^s and the rest HAT^r. Among 36 TG^r (HAT^s) variants tested, 23 did not revert to HAT^r, 4 reverted spontaneously and with EMS, and 9 reverted only with EMS. Some of the latter were probably HPRT^- nonsense mutants because they were very stringent (had less than 2% of wild-type [³H]hypoxanthine incorporation and HPRT enzyme activity), and did not complement genetically. The introduction of a second marker (BrdUrd^r) in 7 of these strains allowed us to isolate 29 TG^r BrdUrd^r (HAT^s) double drug-resistant lines. Through one-step mutagenesis and selection in HAT medium, from two double resistant strains we could isolate HAT^r (TG^s BrdUrd^s) wild-type phenotypic revertants, each of which probably carries suppressible HPRT and TK nonsense (or missense) alleles and the corresponding supersuppressor. Our strategy could now be extended to obtain variants carrying suppressors in other cell lines.     

TGF-β Signalling Is Required for CD4+ T Cell Homeostasis But Dispensable for Regulatory T Cell Function

TGF-β is widely held to be critical for the maintenance and function of regulatory T (T_reg) cells and thus peripheral tolerance. This is highlighted by constitutive ablation of TGF-β receptor (TR) during thymic development in mice, which leads to a lethal autoimmune syndrome. Here we describe that TGF-β–driven peripheral tolerance is not regulated by TGF-β signalling on mature CD4⁺ T cells. Inducible TR2 ablation specifically on CD4⁺ T cells did not result in a lethal autoinflammation. Transfer of these TR2-deficient CD4⁺ T cells to lymphopenic recipients resulted in colitis, but not overt autoimmunity. In contrast, thymic ablation of TR2 in combination with lymphopenia led to lethal multi-organ inflammation. Interestingly, deletion of TR2 on mature CD4⁺ T cells does not result in the collapse of the T_reg cell population as observed in constitutive models. Instead, a pronounced enlargement of both regulatory and effector memory T cell pools was observed. This expansion is cell-intrinsic and seems to be caused by increased T cell receptor sensitivity independently of common gamma chain-dependent cytokine signals. The expression of Foxp3 and other regulatory T cells markers was not dependent on TGF-β signalling and the TR2–deficient T_reg cells retained their suppressive function both in vitro and in vivo. In summary, absence of TGF-β signalling on mature CD4⁺ T cells is not responsible for breakdown of peripheral tolerance, but rather controls homeostasis of mature T cells in adult mice.     

Improved salt tolerance of transgenic wheat by introducing betA gene for glycine betaine synthesis

A betA gene encoding choline dehydrogenase from Escherichia coli (E. coli) was transformed into wheat (Triticum aestivum L.) via Agrobacterium-mediated transformation. PCR amplification and Southern blotting confirmed the existence of transgene in transformed plants and their progeny. Levels of expression of the betA gene varied among the different transgenic lines based on RT-PCR analysis. Under salt stress conditions, transgenic lines L2 and L3 had higher levels of glycine betaine and chlorophyll, lower Na⁺/K⁺ ratios and solute potential, and less cell membrane damage. These lines also retained moderately higher photosynthesis rates and more vigorous growth than the wild-type line at 200 mM NaCl. In a field trial in a high salt field, transgenic lines L2 and L3 had higher germination rates, more tillers and higher grain yields in comparison to the wild-type plants. This suggested that the transgenic plants were more tolerant to salt stress and have potential for breeding salt-tolerant wheat.     

tRNASer(CGA) differentially regulates expression of wild-type and codon-modified papillomavirus L1 genes

Exogenous transfer RNAs (tRNAs) favor translation of bovine papillomavirus 1 wild-type (wt) L1 mRNA in in vitro translation systems (Zhou et al. 1999, J. Virol., 73, 4972–4982). We, therefore, investigated whether papillomavirus (PV) wt L1 protein expression could be enhanced in eukaryotic cells following exogenous tRNA supplementation. Both Chinese hamster ovary (CHO) and Cos1 cells, transfected with PV1 wt L1 genes, effectively transcribed the genes but did not translate them. However, L1 protein translation was demonstrated following co-transfection with the L1 gene and a gene expressing tRNA^Ser(CGA). Cell lines, stably transfected with a bovine papillomavirus 1 (BPV1) wt L1 expression construct, produced L1 protein after the transfection of the tRNA^Ser(CGA) gene, but not following the transfection with basal vectors, suggesting that tRNA^Ser(CGA) gene enhanced wt L1 translation as a result of endogenous tRNA alterations and phosphorylation of translation initiation factors elF4E and elF2α in the tRNA^Ser(CGA) transfected L1 cell lines. The tRNA^Ser(CGA) gene expression significantly reduced translation of L1 proteins expressed from codon-modified (HB) PV L1 genes utilizing mammalian preferred codons, but had variable effects on translation of green fluorescent proteins (GFPs) expressed from six serine GFP variants. The changes of tRNA pools appear to match the codon composition of PV wt and HB L1 genes and serine GFP variants to regulate translation of their mRNAs. These findings demonstrate for the first time in eukaryotic cells that translation of the target genes can be differentially influenced by the provision of a single tRNA expression construct.     

Defect of Mitotic Vimentin Phosphorylation Causes Microophthalmia and Cataract via Aneuploidy and Senescence in Lens Epithelial Cells

Vimentin, a type III intermediate filament (IF) protein, is phosphorylated predominantly in mitosis. The expression of a phosphorylation-compromised vimentin mutant in T24 cultured cells leads to cytokinetic failure, resulting in binucleation (multinucleation). The physiological significance of intermediate filament phosphorylation during mitosis for organogenesis and tissue homeostasis was uncertain. Here, we generated knock-in mice expressing vimentin that have had the serine sites phosphorylated during mitosis substituted by alanine residues. Homozygotic mice (VIM^SA/SA) presented with microophthalmia and cataracts in the lens, whereas heterozygotic mice (VIM^WT/SA) were indistinguishable from WT (VIM^WT/WT) mice. In VIM^SA/SA mice, lens epithelial cell number was not only reduced but the cells also exhibited chromosomal instability, including binucleation and aneuploidy. Electron microscopy revealed fiber membranes that were disorganized in the lenses of VIM^SA/SA, reminiscent of similar characteristic changes seen in age-related cataracts. Because the mRNA level of the senescence (aging)-related gene was significantly elevated in samples from VIM^SA/SA, the lens phenotype suggests a possible causal relationship between chromosomal instability and premature aging.     

Podophyllotoxin resistance: A codominant selection system for quantitative mutagenesis studies in mammalian cells

Mutants resistant to the microtubule inhibitor podophyllotoxin (Pod^R), a codominant marker, can be readily selected in various mammalian cell lines such as, CHO, HeLa, mouse L cells, Syrian hamster cells (BHK₂₁) and a mouse teratocarcinoma cell line OC15. In CHO cells, the recovery of Pod^R mutants is not affected by cell density (up to 1 × 10⁶ cells per 100-mm diameter dish), and after treatment with the mutagen ethyl methanesulfonate maximum mutagenic effect is achieved after a relatively short expression time (40–48 h). The frequency of Pod^R mutants in various cell lines increased in a dose-dependent manner in response to treatment with the mutagens ethyl methanesulfonate and N-methyl-N′-nitro-N-nitrosoguanidine. The Pod^R selection system thus provides a new genetic marker which should prove useful in studies of quantitative mutagenesis in mammalian cells.     

Elevated ATP via enhanced miRNA-30b, 30c,and 30e downregulates the expression of CD73 in CD8+ T cells of HIV-infected individuals

CD8⁺ T cells play a crucial role against chronic viral infections, however, their effector functions are influenced by the expression of co-stimulatory/inhibitory receptors. For example, CD73 works with CD39 to convert highly inflammatory ATP to adenosine. However, its expression on T cells in the context of viral infections has not been well defined. Here, we analyzed the expression of CD73 on human T cells in a cohort of 102 HIV-infected individuals including those on antiretroviral therapy (ART), ART-naïve, and long-term non-progressors who were not on ART. We found that the frequency of CD73⁺ T cells was markedly lower among T cell subsets (e.g. naïve, effector or memory) in the peripheral blood of all HIV-infected individuals. Notably, CD73 was decreased at the cell surface, intracellular and gene levels. Functionally, CD8⁺CD73⁺ T cells exhibited decreased cytokine expression (TNF-α, IFN-γ and IL-2) upon global or antigen-specific stimulation and impaired expression of cytolytic molecules at the gene and protein levels. In contrast, CD8⁺CD73⁺ T cells expressed elevated levels of homing receptors such as CCR7, α4β7 integrin, which suggests a migratory advantage for these cells as observed in vitro. We also observed significant migration of CD73⁺CD8⁺ T cells into the cerebrospinal fluids of multiple sclerosis (MS) patients at the time of disease relapse. Moreover, we found that elevated levels of ATP in the plasma of HIV-infected individuals upregulates the expression of miRNA30b-e in T cells in vitro. In turn, inhibition of miRNAs (30b, 30c and 30e) resulted in significant upregulation of CD73 mRNA in CD8⁺ T cells. Therefore, we provide a novel mechanism for the downregulation of CD73 via ATP-induced upregulation of miRNA30b, 30c and 30e in HIV infection. Finally, these observations imply that ATP-mediated downregulation of CD73 mainly occurs via its receptor, P2X1/P2RX1. Our results may in part explain why HIV-infected individuals have reduced risk of developing MS considering the role of CD73 for efficient T cell entry into the central nervous system.     

SPARC Expression Is Selectively Suppressed in Tumor Initiating Urospheres Isolated from As+3- and Cd+2-Transformed Human Urothelial Cells (UROtsa) Stably Transfected with SPARC

Background

This laboratory previously analyzed the expression of SPARC in the parental UROtsa cells, their arsenite (As⁺³) and cadmium (Cd⁺²)-transformed cell lines, and tumor transplants generated from the transformed cells. It was demonstrated that SPARC expression was down-regulated to background levels in Cd⁺²-and As⁺³-transformed UROtsa cells and tumor transplants compared to parental cells. In the present study, the transformed cell lines were stably transfected with a SPARC expression vector to determine the effect of SPARC expression on the ability of the cells to form tumors in immune-compromised mice.

Methods

Real time PCR, western blotting, immunohistochemistry, and immunofluorescence were used to define the expression of SPARC in the As⁺³-and Cd⁺²-transformed cell lines, and urospheres isolated from these cell lines, following their stable transfection with an expression vector containing the SPARC open reading frame (ORF). Transplantation of the cultured cells into immune-compromised mice by subcutaneous injection was used to assess the effect of SPARC expression on tumors generated from the above cell lines and urospheres.

Results

It was shown that the As⁺³-and Cd⁺²-transformed UROtsa cells could undergo stable transfection with a SPARC expression vector and that the transfected cells expressed both SPARC mRNA and secreted protein. Tumors formed from these SPARC-transfected cells were shown to have no expression of SPARC. Urospheres isolated from cultures of the SPARC-transfected As⁺³-and Cd⁺²-transformed cell lines were shown to have only background expression of SPARC. Urospheres from both the non-transfected and SPARC-transfected cell lines were tumorigenic and thus fit the definition for a population of tumor initiating cells.

Conclusions

Tumor initiating cells isolated from SPARC-transfected As⁺³-and Cd⁺²-transformed cell lines have an inherent mechanism to suppress the expression of SPARC mRNA.     

Design of a statistical method for the analysis of mutagenesis at the hypoxanthine-guanine phosphoribosyl transferase locus of cultured Chinese hamster ovary cells

Mutagenesis data collected in the mammalian cell CHO/HGPRT assay were analyzed to study the distribution of the experimental errors associated with the test. The data neither followed the widely assumed Poisson distribution nor satisfied the usual statistical assumptions of normality and hompgeneous variance of experimental errors. We transformed the data by using the power formula Y = (X + A)^λ where X is the observed mutation frequency, Y is the transformed frequency, and A and λ are constants determined by the procedure of Box and Cox. Setting A = 1 and λ = 0.15 we produced transformed values for which the assumptions of homogeneous variance and normal distribution were satisfied. This transformation enables one to properly use Student's t-test and dose-response analysis of variance to analyze CHO/HGPRT results. The experimental design for CHO/HGPRT mutagenesis assays is also discussed.     

Transfer of anchorage independence by isolated metaphase chromosomes in hamster cells

The cellular property of being able to grow on agar (aga⁺) or to show anchorage independence has been transferred by means of metaphase chromosomes from CHO cells to BHK and other permanent transformed hamster lines unable to grow on agar. As with other genetic markers, the transferents are unstable when grown under non-selective conditions. The aga⁺ transferents are tumorigenic, providing further evidence for the association between the ability to grow in agar or anchorage independence and tumorigenicity. Evidence has been obtained in these experiments for the existence of at least two discrete events in the transformation of normal into tumorigenic cells. The ability to transfer and select for the aga⁺ marker in recipient cells indicates that tumorigenicity behaves dominantly phenotypically.