Differential Effects of Retroviral Long Terminal Repeats on Interleukin-3 Gene Expression and Autocrine Transformation X-Y Wang1 and JA Mccubrey1,2

Differential effects of retroviral long terminal repeats on interleukin-3 gene expression and autocrine transformation X-Y Wang1 and JA McCubrey1,2

1Department of Microbiology and Immunology and 2Leo Jenkins Cancer Center, East Carolina University School of Medicine, Greenville, North Carolina 27858, USA

Previously we documented the transposition of an intracister- gration include: promoter insertion,4,5 enhancer insertion,6,7 nal A particle (IAP) provirus to the interleukin 3 (IL-3) locus inactivation of a gene,8,9 and disruption of a mRNA stability which resulted in autocrine transformation. In the present 10 study, the effects of different long terminal repeats (LTRs) on region. IL-3 gene expression and autocrine transformation were inves- Tumors induced by insertional mutagenesis mechanisms tigated. LTRs from defective IAPs, and replication competent have also been observed after retrotransposon transpositions, Moloney murine leukemia virus (MoMuLV), human T cell leuke- such as murine intracisternal A particles (IAPs).11–13 Murine mia (HTLV), and immunodeficiency (HIV) viruses, were inserted IAPs are defective retroviruses which have deletions in their ′ 5 of the IL-3 promoter region, and their transforming abilities env genes, thus they do not appear to leave the cells. There determined. Addition of the lymphocyte specific (LS) IAP-LTR are approximately 1000 copies of IAP proviral elements in a to the germline IL-3 (gIL3) gene, the IAP-LTR present in the pre- 14 viously described transposition, resulted in a modified IL-3 haploid murine genome. Most IAP transpositions were dis- gene that only infrequently transformed IL-3-dependent cells. covered because they affected the function of genes at the In contrast, addition of plasmacytoma (PC) IAP-LTRs to the target sites which resulted in traits that were readily detected, gIL3 gene, which were isolated from IAPs expressed in plasma- eg ability of the cells to grow in the absence of a previously cytomas, resulted in modified IL-3 genes that transformed IL-3- required growth factor.11–13 In addition to IAP elements, six dependent cells more readily. The MoMuLV-LTR and the TCR␦ enhancer also stimulated high levels of IL-3 expression and other families of murine retrotransposons have been autocrine transformation. In contrast, the HTLV-I, HTLV-II and described. They are VL30, ETn, GLN, MuRRS, MrRVY, and 15 HIV LTRs did not induce significant IL-3 synthesis or autocrine MYS. The transposed elements range from a full length retro- transformation. Consistent with these results, higher levels of transposon or just a part of a LTR sequence. The transposed CAT expression were observed in cells transiently transfected retrotransposons, including IAP elements, can act as inser- with PC-IAP-LTR or a TCR enhancer compared with LS-IAP and tional mutagens as was described above for chronic retro- HTLV LTRs. In summary, the rank order for the effects of different LTRs on IL-3 expression and cell transformation is: TCR␦- viruses. IAP proviral insertions alter the expression of many enhancer Ϸ MoMuLV-LTR Ͼ PC-IAP-LTRs À LS-IAP-LTR À different classes of genes, including: protooncogenes, devel- HTLV-LTRs Ϸ HIV-LTR. These results indicate that the LS-IAP- opmental, growth factor and receptor genes.11–13,16–19 LTR is very weak at inducing IL-3 gene transcription and Like other retroviral proviruses, IAP elements contain LTRs additional genetic mutations may be necessary for LS-IAPs to at both the 5′ and 3′ flanking regions. These regulatory LTR induce autocrine transformation of hematopoietic cells. In con- sequences consist of U3, R, and U5 regions. Several protein trast, the enhancers contained in PC-IAP-LTRs and TCR enhancers may be more effective in inducing abnormal gene binding sites within the U3 region have been described. The expression and malignant transformation. important sites with regard to the following studies are: a sim- Keywords: IL-3; autocrine transformation; retroviral LTRs ian virus 40 core enhancer sequence (Enh1), a transcriptional enhancer Enh2 binding site, and a cyclic AMP response element or the homologous ATF binding site (ATF/CRE).20–25 Introduction By the interaction of specific proteins to these and other regulatory sequences contained within the LTRs, IAP expression Replication-competent retroviral genomes contain gag, pol can be controlled by alterations in methylation status, onco- 14,24–26 and env coding sequences which encode the functional pro- gene expression and hormonal modulations. IAP gene teins that are required for viral replication.1 A long terminal expression and particle formation have been most often docu- repeat (LTR) flanks the coding sequences at both the 5′ and mented in actively dividing cells, such as, preimplantation 14 3′ ends of the retroviral provirus. Retroviruses which induce embryos, immature thymocytes, or in tumor cells. IAP tumors are called oncogenic or transforming retroviruses. expression has been less frequently observed in normal mouse Acute oncogenic retroviruses, which contain an oncogene, tissues. Organs and cells which express low levels of IAPs cause cancer after a short period of time post-infection of a include: spleen, muscle, pancreas, ovary, thymus, placenta, 14,27 susceptible host, ranging from one to a few months. Chronic macrophage and fibroblast cells. IAP expression is under oncogenic retroviruses do not normally encode oncogenes, genetic control in non-neoplastic cells. This results in differ- and a malignancy only occurs many months or years post- ences in the amount of IAP RNAs detected in certain cells as infection. Chronic oncogenic retroviruses can cause cancer well as the proportion of genomic and subgenomic mRNA 27–29 via proviral insertional activation of flanking cellular genes.2,3 transcripts. The novel insertion of the viral sequences into host cell DNAs Murine plasmacytoma cells (B plasma immunoglobulin- can have profound effects on the regulation of the adjacent secreting tumor cells) generally express higher levels of IAP 14 genes and lead to malignant transformation. Mechanisms mRNA transcripts than normal activated B or T lymphocytes. which result in abnormal gene expression after proviral inte- Moreover, there are sequence differences between the LTR cDNAs isolated from the plasmacytoma cells and normal lymphocytes.14,26 These observations gave rise to the nomencla- Correspondence: JA McCubrey ture plasmacytoma (PC) IAP and lymphocyte specific (LS) Received 11 March 1997; accepted 6 June 1997 IAP.26 Furthermore, there are two classes of PC-IAPs (PCI-IAP Effects of LTRs on IL-3 expression X-Y Wang and JA McCubrey 1712 and PCII-IAP), whose DNA sequences differ primarily in the Materials and methods Enh2 domain. The PC-IAP and LS-IAP LTR sequences vary considerably in the Enh1, Enh2, and ATF/CRE DNA Cell culture sequences.26 In addition, the LTRs present in PC-IAPs are dif- ferentially hypomethylated when compared to the IAP-LTRs 26 Cells were maintained in a humidified 5% CO2 incubator with expressed in the normal lymphocytes. These variations in Dulbecco’s modified Eagle’s medium (DMEM) containing 5% DNA sequences and methylation status in the enhancer iron-supplemented bovine calf serum (CS; Hyclone, Logan, regions of the PC and LS IAP-LTRs have been proposed to UT, USA). In all experiments, medium (DMEM) contained 5% 26 result in different patterns of expression of these IAPs. CS and is referred to as DMEM. The IL-3-dependent lymphoid The type I human T cell leukemia virus (HTLV-I) is the pri- line, FL5.12,59,60 and the factor-dependent myeloid FDC-P161 mary cause of adult T cell leukemia/lymphoma (ATLL).30,31 cell line were maintained in DMEM supplemented with clari- Infection by HTLV-I normally results in a slow onset of clinical fied 20% supernatant prepared from the WEHI-3B cell line disease which manifests itself after a long latency period rang- (WEHI conditioned medium, WCM) which served as a source ing from 20 to 30 years.32,33 HTLV-I infection usually targets of IL-3. CD4+ T cells and is capable of transforming these cells.30,31 The viral protein product p40tax has been shown to immor- talize T cells34,35 and transform rat fibroblasts.36,37 This protein Genetic modifications of the germline IL-3 gene by is essential for the transforming effects induced by HTLV-I adding enhancers infection. Tax exerts its effects by transactivation of both viral38–41 and cellular genes, including cytokines (IL-2) and cytokine receptor (IL-2 receptor) genes.42–45 HTLV-II was orig- The germline IL-3 gene (gIL3) was subcloned into the 61 46,47 pSV2neo expression vector as described (Figure 1). To facili- inally isolated from patients with hairy cell leukemia. ′ HTLV-II is closely related to HTLV-I, and also encodes a tax tate further manipulations, a 5 end unique EcoRI site was created which resulted in p5′EcoRI-gIL3.62 Subsequent manipu- gene.48–50 Although HTLV-II has been shown to associate with lations were performed on the p5′EcoRI-gIL3 construct. The a variety of lymphoproliferative disorders, there is no evidence lymphocyte-specific (LS) IAP-LTR,13 and the Moloney murine indicating a direct contribution of HTLV-II in lymphoprolifer- leukemia virus (MoMuLV)-LTR63 sequences were inserted at ative malignancies.51–54 Moreover, the function of the Tax the 5′ end of IL-3 gene by using the unique EcoRI site of protein in HTLV-II is different from that in HTLV-I.55 p5′EcoRI-gIL3 (Figure 1). The plasmacytoma (PC) IAP-LTRs Human immunodeficiency virus type 1 (HIV-1) is the (PCI-IAP, PCII-IAP),26 which were provided by Dr KK Lueders etiologic agent of acquired immunodeficiency syndrome 56,57 (NIH, Bethesda, MD, USA), were cleaved from (AIDS). However, HIV infection can also lead to certain pLTR(PC45)CAT or pLTR(PC17)CAT26 by HindIII digestion neoplasia including non-Hodgkin’s lymphoma, Kaposi’s sar- and inserted into the blunted EcoRI site in p5′EcoRI-gIL3 vec- 58 coma and cervical carcinoma. Like other retroviruses, the tor after filling in the HindIII ends (Figure 1). Human T lym- integrated HIV provirus also contains flanking LTR sequences. photropic virus-LTRs (HTLV-I, HTLV-II), human immunodefi- With the increasing development of beneficial AIDS treat- ciency virus (HIV)-LTR,64 were provided by Dr K Shimotohno ments and the prolonged survival of AIDS patients, the long- (National Cancer Institute of Japan, Tokyo, Japan), which were term tumorigenic effects associated with HIV infection will cut out of pHTLV-I-CAT, pHTLV-II-CAT or pHIVLTR-luciferase become a more and more important health care issue. by HindIII and inserted into the p5′EcoRI-gIL3 plasmid as In present study, the effects of cis-acting elements contained described for PC-IAP-LTRs (Figure 1). The TCR␦ enhancer65 in different retroviral LTRs on IL-3 gene expression and cell was provided by Dr MS Krangel (Duke University, Durham, transformation were investigated in order to understand how NC, USA). The enhancer sequence was released by EcoRI and proviral transpositions and insertional mutagenesis can affect BamHI digestion from pBluescript KS+ vector, which was then the expression of cytokine genes and result in autocrine trans- blunt-ended and inserted into p5′EcoRI-gIL3 blunted EcoRI formation of hematopoietic cells. In this study, the effects of site (Figure 1). All the LTRs and TCR␦ enhancer DNA these different retroviral LTRs on the expression of a given sequences were inserted in both 5′→3′ and 3′→5′ orien- gene (IL-3) on the cytokine dependency and tumorigenicity tations. Plasmids containing positive constructs were verified of a single hematopoietic cell line have been unequivocally by restriction endonuclease digestions. compared. LTRs isolated from IAPs which are expressed in normal lymphocytes were weak in stimulating sufficient IL-3 expression to result in autocrine transformation. This suggests Electroporation and isolation of stable transfectants that insertional mutagenesis promoted by these IAPs is associated with additional events. In contrast, LTRs from IAPs Cell electroporation and stable transfectant isolation were per- which are expressed at high levels in tumor cells were better formed as described previously.62,66 Following electropor- able to stimulate sufficient IL-3 expression to result in ation, the cells were plated in 100-mm culture dishes contain- autocrine transformation. This was also the case with the ing DMEM + 20% CS + IL-3 for 2 days to allow the transfected MoMuLV-LTR and TCR␦ enhancer, as they effectively induced DNAs to become stably incorporated into the chromosomes. autocrine transformation when ligated to the murine IL-3 Then the cells were transferred into 96-well microtiter plates gene. In contrast, the HTLV and HIV LTRs did not stimulate with normal DMEM media containing the antibiotic G418 (2 sufficient IL-3 expression to induce autocrine transformation. mg/ml, Geneticin; Sigma, St Louis, MO, USA), or under the Thus there are distinct differences in these LTRs to promote same conditions but supplemented with IL-3. The plates were cytokine gene expression and in some cases malignant trans- subsequently fed every 3 days with fresh media containing formation of hematopoietic cells. G418. After 2 weeks in culture, the number of wells containing G418-resistant cells were enumerated. Effects of LTRs on IL-3 expression X-Y Wang and JA McCubrey 1713

Figure 1 Structure of the modified IL-3 gene constructs and their abilities to abrogate the factor dependency of IL-3-dependent cells. The unique EcoRI site at the 3′ end of gIL3 gene was destroyed by a fill-in reaction and a unique EcoRI site was created by insertion of a linker to facilitate the construction of the modified gIL3 gene constructs. LS-IAP-LTR, PCI-IAP-LTR, PCII-IAP-LTR, TCR␦ enhancer, HIV-LTR, HTLV-I-LTR, HTLV-II-LTR, and MoMuLV-LTR were inserted 1.5 kb 5′ of IL-3 transcription initiation site. All the enhancer elements were inserted in both 5→3′ and 3′→5′ orientations. E, EcoRI; B, BamHI; H, HindIII. The bars around a restriction site indicated that the site was destroyed. DMEM + G418 indicated that the cells were plated in 2 mg/ml G418 with the absence of exogenous IL-3. IL-3 + G418 indicated that the cells were plated in 2 mg/ml G418 with the presence of exogenous IL-3. The fractions refer to number of wells positive for growth over total number of wells examined. Since there were no significant difference in transforming efficiency whether the enhancers were inserted in either 5′→3′ or 3′→5′ orientation, the results for the same enhancer in both orientations were combined. Division of number of wells by 96 equals the number of times the transfection with a particular DNA was performed. DMEM/IL-3 in last column indicated the percentage of growth in the absence vs the presence of IL-3.

DNA preparation and Southern blot analysis RNA isolation and northern blot analysis

Total cellular RNAs were isolated by lysis in TRIzol reagent High molecular weight genomic DNAs were isolated from the (Gibco/BRL) as described.66 Twenty micrograms RNAs were 60,62 cell lines as described earlier. Ten micrograms of DNA denatured with formaldehyde/formamide and electrophoresed was digested with restriction endonucleases (Gibco/BRL, Gai- in 1.2% agarose gels containing formaldehyde. RNAs were thersburg, MD, USA). The digested DNAs were electrophor- transferred to Hybond-N membranes. Hybridization was per- esed through 0.8% agarose gels, transferred to Hybond-N formed with radiolabeled probes as described.66,67 membranes (Amersham, Arlington Heights, IL, USA) and then hybridized with radiolabeled probes as described.60,62,66,67 Ribonuclease protection assay

The RNA probes were synthesized by using the T7/T3 MAXI- Limiting dilution assay script kit (Ambion, Austin, TX, USA). An antisense RNA probe specific for IL-3 was synthesized by T7 polymerase after lin- earizing the IL-3 pGEM vector with HindIII, which gave a 364 The cloning efficiencies of the cells were measured by plating nucleotide protected probe. An RNA probe for ␤-actin was the cells at different concentrations in 96-well microtiter prepared by the same method and gave a 250 nucleotide pro- plates containing either DMEM or DMEM + IL-3 media. The tected fragment. The ribonuclease protection assay was per- plates were fed every 3 days with fresh media. Microtiter wells formed as described,68 by using RPA II ribonuclease protec- negative for growth were determined after 2 weeks of culture. tion assay kit (Ambion). The samples were electrophoresed on The cloning frequencies of the cells were estimated based on a 5% polyacrylamide (acrylamide/bisacrylamide ratio the Poisson distribution. Some factor-independent subclones 37.5:1)/8 m urea gel. The protected fragments were detected were isolated and further characterized. by autoradiography and quantified by densitometric scanning. Effects of LTRs on IL-3 expression X-Y Wang and JA McCubrey 1714 Nuclear run-on transcription assays The HTLV-LTR-CAT constructs were provided by Dr K Shimo- tohno.64 The TCR␣-CAT construct was provided by Dr JM Nuclei isolation, in vitro transcription reaction and hybridiz- Leiden (University of Chicago, Chicago, IL, USA).74 ation were performed as described previously.13,62 IL-3 transcription levels were quantified by PhosphorImager scanning (Molecular Dynamics, Sunnyvale, CA, USA). The signal inten- Tumorigenicity assays sities were normalized to genomic DNA controls. 1 × 106 cells were injected s.c. in the subgastric region of BALB/c nude mice and tumor growth was observed as Nucleic acid probes described.62,66 At the appearance of a tumor 1-cm in diameter, the mice were sacrificed. In some cases, cells were recovered The IL-3-specific probe69 was provided by Dr I Young from the tumors and grown in medium containing G418. (Australian National University, Canberra, Australia). The murine macrophage growth factor-specific probe (MGF)70 was provided by Dr S Lyman (Immunex, Seattle, WA, USA). The c- Results myc-specific probe71 was provided by Dr MD Cole (Princeton University, Princeton, NJ, USA). Different abilities of enhancer modified IL-3 genes to transform factor-dependent cells

Proliferation assays Different retroviral LTRs were inserted 5′ of the gIL3 gene to compare the abilities of endogenous as well as exogenous The levels of IL-3 secretion by the different transfected cells retroviral enhancers to induce autocrine transformation of were assayed on the IL-3/GM-CSF-dependent FDC-P1 cells as hematopoietic cells (Figure 1). All of these enhancers induce described previously.62 In the MoAb neutralization experi- malignant transformation under certain circumstances, albeit ments, supernatants were incubated with neutralizing ␣IL-3 or at varying frequencies and by different mechanisms. The ␣GM-CSF MoAb for 1 h at 37°C prior to the addition of FDC- hematopoietic FL5.12 IL-3-dependent cell line is a good P1 cells. Neutralizing monoclonal antibodies to murine IL-3 recipient to measure the oncological effects of these and GM-CSF were purchased from Genzyme (Boston, MA, enhancers since it is absolutely dependent upon IL-3 for USA). growth in vitro and it does not form tumors upon injection into immunocompromised mice.13,59,61,62,66 The murine IAP- LTRs (LS-IAP-LTR, PCI-IAP-LTR, and PCII-IAP-LTR), the Chloramphenicol acetyltransferase assay (CAT) MoMuLV-LTR, and the human HTLV-I, HTLV-II, HIV retroviral LTRs were all inserted approximately 1.5 kb upstream of the Plasmids containing CAT reporter genes were transiently transcription start site of the germline IL-3 gene (Figure 1). As transfected into FL5.12 cells by electroporation. 1 × 107 cells a control, the TCR␦ enhancer was also inserted in the gIL3 at logarithmic stage were washed and electroporated with gene construct. The TCR␦ enhancer is a cellular enhancer supercoiled plasmid DNA containing 20 ␮g CAT construct which has similar transcription factor binding sites as the and 4 ␮g CMV ␤-galactosidase plasmid (Clontech Labora- MoMuLV-LTR,75 and is translocated in certain T cell leuke- tories, Palo Alto, CA, USA) at 250 V, 960 ␮FD. The cells were mias.76–78 These modified IL-3 constructs allow comparison of collected by centrifugation 48 h after electroporation and the effects of different LTRs inserted into a single site on the washed twice with ice cold PBS, which were then resus- induction of cytokine-independence and leukemia. pended in 300 ␮l of 250 mm Tris (pH 7.8) and subjected to The modified gIL3 constructs were transfected into IL-3- three cycles of freeze (ethanol/dry ice, 10 min)/thaw (37°C, dependent FL5.12 cells. Stable transfectants were isolated 10 min) treatment. The cell debris was pelleted by centrifug- after selection in G418 in the presence or absence of exogen- ation at 12 000 g for 5 min. The supernatants were collected ous IL-3. The drug selection conditions were sufficient for the and heated at 65°C for 10 min to inactivate the deacetylases. isolation of stably transfected cells, as no G418r cultures were The CAT assay reaction was performed following previously observed in mock-transfected cells (Figure 1). Furthermore, described procedure.72 One hundred microliter cell extract inheritance of the empty vector (pSV2neo) did not transform was incubated in a reaction mixture containing 1.6 mm chlor- the cells to growth factor independence. The data is presented amphenicol (Sigma, St Louis, MO, USA), 150 mm Tris (pH as the number of G418r wells growing in the presence or 7.8), 100 ␮m acetyl CoA (Sigma) and 0.1 ␮Ci [14C]-acetyl CoA absence of exogenous IL-3 over the total number of wells (NEN, Boston, MA, USA) in a 200 ␮l total volume. The incu- examined. This data has been further compared as the ratios bation was carried out at 37°C for 2 h. After the incubation, of wells growing in the absence of exogenous IL-3 vs the num- 350 ␮l cold ethyl acetate (Fisher, Raleigh, NC, USA) was ber of wells recovered when IL-3 was added (G418 was added to the reaction mixture, and mixed by vigorous vor- present in both conditions). The inheritance of transfected texing. The layers were separated by centrifugation at 12 000 genes was verified by Southern blot analysis and the individ- g for 5 min. Three hundred microliters of the top aqueous ual clones usually contained one to two copies of transgenes phase was removed and counted on a Beckman LS 7000 (see Figure 9 below). (Sunnydale, CA, USA) scintillation counter. The analysis for As controls, FL5.12 cells were also transfected with the gIL3 each construct was performed three times. The results were and gIL3 + LS-IAP-LTR transgenes. Previous studies have normalized by the ␤-galactosidase activity. The ␤-galacto- shown that these constructs did not efficiently abrogate the sidase assay was performed as described.73 The ␤-galacto- cytokine dependency of FL5.12 cells.62,66 Essentially no sidase activity was measured by reading the optical density at growth factor-independent cells were isolated after transfec- 420 nm on a spectrophotometer. The LS-IAP-LTR-CAT, PC- tion with the gIL3 gene (less than 1%), indicating that the IAP-LTR-CAT constructs were provided by Dr KK Lueders.26 inheritance of another gIL3 gene did not readily abrogate IL- Effects of LTRs on IL-3 expression X-Y Wang and JA McCubrey 1715 3 dependency. Moreover, addition of a LS-IAP-LTR to the gIL3 to cytokine-independent cells, as the pools of transfected cells gene resulted in some factor-independent cells but again at a (presented in Figure 1). The MoMuLV-LTR and TCR␦ enhancer relatively low frequency (8%) (Figure 1). were better in inducing cytokine-independent growth than the When the PC-IAP-LTR modified gIL3 gene constructs were PC-IAP-LTRs, which in turn were more effective than the LS- introduced into FL5.12 cells, a four-fold higher transforming IAP-LTR. Thus it is unlikely that there were only rare trans- frequency was observed as compared with LS-IAP-LTR modi- fected cells which had inherited multiple copies of the IL-3 fied gIL3 gene (34 vs 8%, Figure 1). The two types of PC-IAP- transgenes or peculiar integration fragments which gave rise LTRs (PCI-IAP-LTR and PCII-IAP-LTR) differ primarily in DNA to all the cytokine-independent cells. sequences contained in the Enh2 domains of the U3 region.26 However, these two types of PC-IAP-LTRs had similar effects on abrogating the cytokine dependency of IL-3-dependent IL-3 mRNA expression in the transfected cells cells when ligated to the gIL3 gene (Figure 1). Thus both PC- IAP-LTRs activated IL-3 expression better than the LS-IAP-LTR. To determine whether the recovery of factor-independent cells The MoMuLV-LTR-modified gIL3 gene was also more effec- correlated with the amount of IL-3 mRNA transcripts which tive in transforming factor-dependent cells than the LS-IAP- accumulated in the cells, Northern blot analysis was perfor- LTR, as eight-fold more factor-independent cells were reco- med (Figure 3). mRNAs were extracted from pools selected in vered (Figure 1). The TCR␦ enhancer, which has similar tran- IL-3- and G418-containing medium and there was no prior scription factor binding motifs as MoMuLV-LTR,75 was also selection for factor-independent growth. IL-3 mRNA tran- eight-fold more effective in inducing factor-independence scripts were not detected in the parental FL5.12 cells, as well than the LS-IAP-LTR, when linked to the gIL3 gene (Figure 1). as cells transfected with gIL3 genes containing the LS-IAP-LTR, None of HIV-LTR, HTLV-I-LTR, or HTLV-II-LTR modified HTLV-I-LTR, HTLV-II-LTR, and HIV-LTR (Figure 3, lanes 1–2, gIL3 genes could efficiently transform IL-3-dependent cells 4–6). However, IL-3 mRNA transcripts were readily detected (Figure 1). This indicated that these human retroviral LTRs, in the cells transfected with gIL3 genes containing the PCI- by themselves, were not sufficient to stimulate adequate IL-3 IAP-LTR, MoMuLV-LTR, and the TCR␦ enhancer (Figure 3, expression to relieve the cytokine dependency of murine IL- lanes 3, 7 and 8). Therefore higher levels of IL-3 mRNA were 3-dependent cells. Thus the different retroviral LTRs varied in observed in those pools which more frequently gave rise to their abilities to abrogate the cytokine dependency of IL-3-independent cells. FL5.12 cells.

Addition of transcriptional enhancers did not alter IL-3 Isolation of IL-3-independent subclones from clones mRNA stability transfected with the LTR-modified gIL3 genes To determine whether the addition of different transcriptional To determine whether factor-independent subclones could be enhancers had any effects on IL-3 mRNA stability, the IL-3 isolated from individual clones that were first isolated in the mRNA half-lives in some IL-3-independent clones were exam- presence of IL-3 and G418, limiting dilution experiments were ined by ribonuclease protection assays (Figure 4). mRNAs performed. The cells were plated at different cell concen- were isolated from the transfected cells that had been incu- trations in the presence and absence of exogenous IL-3. These bated with the transcriptional inhibitor actinomycin D. As clones had been isolated in the presence of IL-3 and G418 as shown in Figure 4, the IL-3 mRNAs in the transfected cells described in Figure 1, and there had been no selection for exhibited short half-lives (t. = 0.5 to 1 h), similar to the half- factor-independent growth. The cloning efficiencies obtained life of wild-type germline IL-3 mRNA documented earlier.62,66 with the individual clones that were obtained after transfec- In contrast, the ␤-actin mRNAs had long half-lives (Ͼ8 h). tion with a given transgene (n = 3 to 6 for each type of trans- These results indicated that the manipulations of the gIL3 gene fected clone) were similar and have been combined in Figure by adding enhancer elements did not affect the IL-3 mRNA 2 to present data containing means and standard deviations. stability. Factor-independent subclones were isolated rarely from the gIL3 + LS-IAP-LTR gene transfected cells as they were only recovered when у1000 cells/well were plated (Figure 2a). In IL-3 mRNA transcription rates were increased in cells contrast, IL-3-independent subclones were isolated much transfected with gIL3 + LTRs more frequently from the clones which inherited the gIL3 gene plus either the MoMuLV-LTR or the TCR␦ enhancer (Figure To determine whether the transcription rate of the IL-3 gene 2b and data not presented). The gIL3 + MoMuLV clones had was increased in the stable transfectants which inherited the a nearly equal plating efficiency when grown in the presence modified IL-3 genes, nuclear run-on assays were performed. and absence of exogenous IL-3, indicating that essentially The levels of IL-3 transcripts were quantified by Phosphor- every transfected cell gave rise to a factor-independent sub- Imager scanning. As shown in Figure 5, no IL-3 transcription clone. Factor-independent subclones were isolated less fre- was detected in the parental IL-3-dependent FL5.12 cells (lane quently from clones transfected with the gIL3 + PC-IAP-LTR 1). Similar levels of IL-3 nascent transcription were detected construct as only one in 90 cells would give rise to a subclone in IL-3-dependent gIL-3 transfected cells (data not presented). that would proliferate in the absence of exogenous IL-3 (Figure The LS-IAP-LTR modified gIL3 gene transfected cells exhibited 2c and d). Factor-independent subclones were essentially not increased levels of IL-3 transcription rate (an average of four- isolated from the transfected cells containing the gIL3 plus fold higher than the base line observed with FL5.12 cells, human retroviral LTRs genes (data not presented). lanes 2 and 3). As a negative control, the transcription rate of These results indicate that the individual clones recovered MGF (macrophage growth factor, also known as c-kit ligand) after transfection with the LTR modified gIL3 genes displayed was examined and no transcription of this gene was observed similar growth properties, in terms of the ability to give rise in the cell lines (Figure 5). As a positive control, the levels of Effects of LTRs on IL-3 expression X-Y Wang and JA McCubrey 1716

Figure 2 Limiting dilution analysis of FL5.12 cells transfected with modiﬁed IL-3 genes. Limiting dilution analysis was performed on individual clones. These results obtained with the individual clones transfected with the same construct were similar so they were averaged together. (a) FL-gIL3 + LS-IAP-LTR (n = 6); (b) FL-gIL3 + MoMuLV-LTR (n = 4); (c) FL-gIL3 + PCII-IAP-LTR (n = 3); and (d) FL-gIL3 + PCI-IAP-LTR (n = 4). The dotted line is at 37% wells negative for growth from which the cloning efﬁciency can be estimated by the Poisson distribution.

Figure 3 Expression of the IL-3 mRNA in modiﬁed IL-3 gene transfected cells. Northern blot analysis was performed with mRNAs extracted from the following modiﬁed IL-3 gene transfected FL5.12 pools which were isolated in IL-3 + G418 containing medium. Lane 1, FL5.12 parental cells; lane 2, FL-gIL3 + LS-IAP-LTR; lane 3, FL-gIL3 + PCI-IAP-LTR; lane 4, FL-gIL3 + HTLV-I-LTR; lane 5, FL-gIL3 + HTLV-II-LTR; lane 6, FL- gIL3 + HIV-LTR; lane 7, FL-gIL3 + MoMuLV-LTR; lane 8, FL-gIL3 + TCR␦. Effects of LTRs on IL-3 expression X-Y Wang and JA McCubrey 1717

Figure 4 Half-life of IL-3 mRNA in transfected lines. The half-life of IL-3 mRNA was determined by ribonuclease protection assays. The Figure 5 IL-3 mRNA transcription rate detected by nuclear run-on RNAs were isolated at different time-points after actinomycin D (Act analyses. Nascent IL-3 transcription was analyzed in certain modified D) treatment of the following factor-independent cell lines: (a) FL-gIL3 IL-3-inherited clones. Lane 1, FL5.12; lane 2, FL-gIL3 + LS-IAP-LTR + ␦ ′→ ′ + ␦ ′→ ′ + TCR 5 3 C1; (b) FL-gIL3 TCR 3 5 C1; (c) FL-gIL3 LS-IAP- 3′→5′ C2.1; lane 3, FL-gIL3 + LS-IAP-LTR 3′→5′ C1.1. These clones ′→ ′ LTR 3 5 C1.1. were isolated in the presence of DMEM as described in Figure 2. Cul- ture of parental cell line in the absence of exogenous IL-3 did not result in IL-3 nascent transcription and the levels of IL-3 nascent tran- c-myc nascent transcription were examined and was detected scription in gIL3-transfected cells was similar to that observed in at similar levels in all the cell lines (Figure 5). Thus the FL5.12 cells (Refs 13, 62, 66, and data not presented). addition of the IAP-LTR to the gIL3 gene specifically increased the IL-3 mRNA transcription rate. Supernatants were isolated from these cultures and tested for cytokine bioactivity with IL-3/GM-CSF dependent FDC-P1 IL-3 mRNA transcription results in growth factor cells (Figure 6). secretion The highest levels of growth factor activity were detected in the cells transfected with the gIL3 gene plus either the To determine whether the IL-3 mRNA expression led to func- MoMuLV-LTR or the TCR␦ enhancer. This level of IL-3 tional IL-3 protein, the presence of the IL-3 protein was ana- secretion is designated ++++ or 100% (see legend to Table 1). lyzed in the different transfected cells. The same pools were This was confirmed to be IL-3 by neutralization experiments used in these experiments as in the Northern blot analyses. with monoclonal antibodies (MoAb) to murine IL-3 and The pools of IL-3-transfected cells were first cultured in the GM-CSF. presence of exogenous IL-3 at a concentration of 1 × 106 The level of cytokine activity detected in the cells which cells/ml, washed with PBS to remove exogenous IL-3, and cul- inherited the PCI-IAP-LTR modified IL-3 gene was also high, tured in DMEM without exogenous IL-3 for another 24 h. but lower than that recovered from cells transfected with the Effects of LTRs on IL-3 expression X-Y Wang and JA McCubrey 1718 or + depending on the amount detected in individual clones or subclones (see Figure 7). Finally, no growth factor activity was observed in the cells transfected with the gIL3 gene plus HIV, HTLV-I, or HTLV-II LTRs (Figure 6 and data not presented). IL-3 secretion was also examined in some factor-independent clones, which were isolated in the absence of exogenous IL-3. High levels of cytokine activities were observed in factor-independent cells transfected with the gIL3 gene plus MoMuLV-LTR. In contrast, the LS-IAP-LTR-modified

Figure 6 Growth factor activity in modified IL-3 gene transfected cells. Stable transfectants, which were isolated in IL-3 + G418-containing medium, were grown in the presence of IL-3 to a concentration of 1 × 106 cells/ml, then washed with PBS and plated in DMEM + BCS growth media. After another 24 h, the supernatants from the cells were isolated and clarified. Serial three-fold dilutions of the extracts were incubated with factor-dependent FDC-P1 cells for 24 h before the pulse of [3H]-thymidine for 6 h. Supernatants were isolated from the cells which inherited the following enhancer modified gIL3 genes: , FL-gIL3 + PCI-IAP-LTR; b, FL-gIL3 + MoMuLV-LTR; í, FL- gIL3 + TCR␦; ̆, FL-gIL3 + HTLV-I-LTR; v, FL-gIL3 + LS-IAP-LTR; ć, FL-gIL3 + HIV-LTR.

gIL3 plus either the MoMuLV-LTR or the TCR␦ enhancer. This Figure 7 Growth factor activity in factor-independent subclones. level of IL-3 secretion is designated +++. Similar results were Growth factor activities of the supernatants isolated from different factor-independent lines were quantiﬁed by incubating factor-dependent observed in supernatants from the cells transfected with the 3 gIL3 gene plus the PCII-IAP-LTR (data not presented). FDC-P1 cells in the presence of [ H]-thymidine. Supernatants were isolated from the following cell lines: í, FL-gIL3 + LS-IAP-LTR 3′→5′ Supernatants recovered from cells transfected with the gIL3 C1.1; ̆, FL-gIL3 + LS-IAP-LTR 3′→5′ C2.1; v, FL-gIL3 + LS-IAP-LTR + LS-IAP-LTR construct contained even lower levels of IL-3 3′→5′ C2.2; ., FL-gIL3 + MoMuLV-LTR 3′→5′ C1.1; X, FL-gIL3 + (Figure 6). This level of IL-3 secretion is designated either ++ MoMuLV-LTR 3′→5′ C2.1; b, WEHI-3B.

Table 1 Tumorigenicity of transfected cell linesa

Cell line Requirement of Secretion of IL-3b No. of mice with Latency period exogenous tumors/No. of mice (weeks) IL-3 for growth examined

FL5.12 + ND 0/4 Ͼ20 FL-gIL3-LS-IAP-LTR 5′→3′C1 + ND 0/4 Ͼ20 FL-gIL3-LS-IAP-LTR 5′→3′C1.1 −++4/4 3–4 FL-gIL3-LS-IAP-LTR 5′→3′C1.2 −++4/4 3–4 FL-gIL3-LS-IAP-LTR 5′→3′C2 + ND 0/4 Ͼ20 FL-gIL3-LS-IAP-LTR 5′→3′C2.1 + ND 0/4 Ͼ20 FL-gIL3-LS-IAP-LTR 5′→3′C2.2 + ND 0/4 Ͼ20 FL-gIL3-LS-IAP-LTR 5′→3′C2.3 −+0/4 Ͼ20 FL-gIL3-LS-IAP-LTR 5′→3′C2.4 −++4/4 6 FL-gIL3-LS-IAP-LTR 5′→3′C3 + ND 0/4 Ͼ20 FL-gIL3-LS-IAP-LTR 5′→3′C3.1 + ND 0/4 Ͼ20 FL-gIL3-LS-IAP-LTR 5′→3′C3.2 + ND 0/4 Ͼ20 FL-gIL3-LS-IAP-LTR 5′→3′C3.3 −++4/4 4–9 FL-gIL3-LS-IAP-LTR 3′→5′C1 + ND 0/4 Ͼ20 FL-gIL3-LS-IAP-LTR 3′→5′C1.1 −++4/4 4 FL-gIL3-LS-IAP-LTR 3′→5′C2 + ND 0/8 Ͼ20 FL-gIL3-LS-IAP-LTR 3′→5′C2.1 −+0/4 Ͼ20 FL-gIL3-LS-IAP-LTR 3′→5′C2.2 −+0/4 Ͼ20 FL-gIL3-LS-IAP-LTR 3′→5′C2.3 −+0/4 Ͼ20

aMice were injected as described in Materials and methods. bCell lines have been classified into the levels of IL-3 they secreted. Determined by comparison to the level of IL-3 in FL-gIL3 + MoMuLV- LTR3′→5′c1.1 as it is the highest level of IL-3 detected by any transfected cell line and is designated ++++ = 100% (+++ = 10 to 99%, eg FL-gIL3 + PCI-IAP-LTR5′→3′c1.1); (++ = 1 to 9%, eg FL-gIL3-LS-IAP-LTR 5′→3′ C1.1); and (+=0.1 to 0.9% eg FL-gIL3-LS-IAP-LTR 5′→3′ C2.3). ND, not determined. Effects of LTRs on IL-3 expression X-Y Wang and JA McCubrey 1719 IL-3 gene transfectants exhibited lower levels of cytokine transfected cells showed a high level of CAT activity (lane 3), activities. Representative data are presented in Figure 7. In this about 16-fold higher than the base line. The LS-IAP-LTR, particular experiment, the levels of [3H]-thymidine uptake HTLV-I-LTR, and HTLV-II-LTR-CAT constructs all exhibited were higher than in the previous experiment. These experi- about two-fold more CAT activity than the negative control ments were performed on different days with different batches (lanes 4, 7 and 8), but were still quite low. In contrast, the of FDC-P1 cells. However, the qualitative differences between PCI-IAP-LTR and PCII-IAP-LTR-CAT constructs exhibited the levels of IL-3 present in the different LTR transfected cells higher levels of CAT activity (lanes 5 and 6), which were is still comparable within the individual experiments. These about four-fold higher than that observed with the LS-IAP-LTR- results indicate that IL-3 was secreted in all the IL-3-inde- CAT construct. The TCR␣ enhancer was used in these experi- pendent clones. However, the level of IL-3 secretion was ments as a positive control since it has similar transcription related to which LTR was ligated to the gIL3 gene. factor binding sites as the TCR␦ enhancer and the MoMuLV These results indicated the rank order for IL-3 bioactivity in LTR enhancers.74,75 The TCR␣-CAT construct transfected cells different transfectants: TCR␦ enhancer Ϸ MoMuLV-LTR Ͼ PC- also showed high CAT activity which was approximately five- IAP-LTRs À LS-IAP-LTR À HIV-LTR Ϸ HTLV-LTRs, which is fold higher than the LS-IAP-LTR (lane 9). Thus these different consistent with the abilities to abrogate cytokine dependency LTRs and enhancers displayed varying abilities to stimulate and the levels of steady-state IL-3 mRNA detected in trans- CAT expression in transient transfection assays as well as IL- fected cells containing these enhancer-modified IL-3 genes. 3 transcription in stably transformed cells.

Transcription stimulating activities of different Differential tumorigenicity of transfected cells retroviral LTRs To determine whether the transfected cells which contained To determine directly the transcription stimulating effects of the modified IL-3 genes were tumorigenic, the cells were the different enhancer elements in IL-3-dependent hematopoi- injected into immunocompromised nude mice. As shown in etic cells, CAT assays were performed (Figure 8). Different Table 1, the factor-dependent parental cells (FL5.12) and enhancer-CAT reporter constructs were transiently transfected clones which were isolated in the presence of IL-3 and G418 into FL5.12 cells. Cell extracts were isolated from these cells (FL-gIL3-IAP-LTR 5′→3′C1, C2, C3, and FL-gIL3-IAP-LTR and CAT activities determined. In the mock-transfected cells 3′→5′ C1, C2) were not tumorigenic, as no tumors were and the cells transfected with the negative control construct detected up to 20 weeks. pSV0CAT, the CAT activities were low and considered to be The tumorigenicity of some IL-3-dependent and inde- the basal levels (lanes 1 and 2). The positive control pSV2CAT pendent subclones were examined. These subclones were isolated by limiting dilution experiments from stably transfected clones as shown in Figure 2. For example, the FL-gIL3-LS-IAP- LTR 5′→3′ C1.1 and C1.2 are two IL-3-independent subclones isolated from the FL-gIL3-LS-IAP-LTR 5′→3′ C1 clone by limiting dilution. The subclones which were examined for their tumorigenicity were divided into different classes depending on the level of IL-3 secretion (see legend to Table 1). The levels of IL-3 secretion in the factor-independent cells were determined as described in Figure 7. The factor-dependent subclones (FL-gIL3-LS-IAP-LTR 5′→3′ C2.1, C2.2, C3.1 and C3.2) were not tumorigenic. The factor- independent subclones which secreted low levels of IL-3 (designated IL-3 secretion class +, (FL-gIL3-LS-IAP-LTR 5′→3′ C2.3, FL-gIL3-LS-IAP-LTR 3′→5′ C2.1, C2.2 and C2.3) did not form tumors upon injection into mice. The factor-independent subclones which secreted higher levels of IL-3 (designated IL-3 secretion class ++, FL-gIL3-LS- IAP-LTR 5′→3′ C1.1, C1.2, C2.4 and C3.3, and FL-gIL3-LS- IAP-LTR 3′→5′ C1.1) were tumorigenic. Moreover, factor- independent cells which secrete high amounts of IL-3 (eg FL- gIL3 + MoMuLV-LTR3′→5′c1.1) rapidly induced tumors upon injection into immunocompromised mice (data not presented). Overall, these results indicated that the tumorigenicity of autocrine factor-independent cells corresponded posi- tively with the amount of IL-3 they secreted. Figure 8 Transcriptional activity of different LTRs detected in FL5.12 cells. Twenty micrograms of each CAT-containing plasmid and 4 ␮g of CMV ␤-galactosidase plasmid were transfected into FL5.12 cells, cell extracts were isolated 48 h later, and CAT assays were per- Inheritance of the IL-3 transgene in the factor- formed on each cell extract. The results for CAT activities were nor- independent subclones malized to the levels of ␤-galactosidase enzymatic activity detected. Cell extracts were isolated from cells transfected with the following To determine whether the factor-independent subclones con- constructs: (1) Mock; (2) pSV0CAT; (3) pSV2CAT; (4) LS-IAP-LTR-CAT; (5) PCI-IAP-LTR-CAT; (6) PCII-IAP-LTR-CAT; (7) HTLV-I-LTR-CAT; (8) tained the same IL-3 transgene integration fragments as the HTLV-II-LTR-CAT; (9) TCR␣-CAT. The analysis for each construct was parental IL-3-dependent clone from which they were derived, performed three times. Standard deviations are indicated by error bars. the structure of the IL-3 locus was examined by Southern blot- Effects of LTRs on IL-3 expression X-Y Wang and JA McCubrey 1720 ting. The factor-independent subclones all contained the same integration fragments as the parental factor-dependent clone (Figure 9). There did not appear to be any gene rearrange- ments or amplifications of the IL-3 transgene in these subclones. Similar results were observed in four other clone/subclone families (data not presented). Factor-independent cells were recovered from some tumors and grown in culture. These contained the same IL-3 integration fragments as the parental IL-3-dependent clones (data not presented). Thus abrogation of IL-3 dependency in the factor- independent subclones did not appear to be the result of a gross modification of the IL-3 transgene.

Increased expression of IL-3 in the factor-independent Figure 10 Nascent transcriptional analysis of IL-3 expression in IL- subclones is due to a transcriptional mechanism 3-dependent clones and IL-3 independent subclones. Nuclei were prepared from lane: (1) FL-gIL3-LS-IAP-LTR 3′→5′C1; (2) FL-gIL3-LS- ′→ ′ ′→ ′ To determine the mechanism responsible for the increase in IAP-LTR 3 5 C1.1; (3) FL-gIL3-LS-IAP-LTR 3 5 C1.2. IL-3 expression in the factor-independent subclones, the rate of IL-3 nascent transcription was compared in the factor- dependent clones and the factor-independent subclones. A low level of IL-3 nascent transcription in the IL-3-dependent FL-gIL3-LS-IAP-LTR 3′→5′C1 was observed (lane 1). The transcription rate of the IL-3 gene was increased about 2–5-fold in the factor-independent subclones (lanes 2 and 3) (Figure 10). Nascent RNA transcripts encoding c-myc were detected in IL-3-dependent and IL-3-independent cells. The level of nascent c-myc transcription was higher in the IL-3-dependent clone, probably because it was grown in relatively high levels of exogenous IL-3. In contrast MGF transcripts were not detected in any of the cell lines. The levels of nascent transcription observed when genomic DNA was used as a probe were similar among the different cell lines. Therefore the increase in IL-3 nascent transcription observed in the IL-3- independent subclones was speciﬁc and associated with the conversion to growth in the absence of exogenous IL-3. To determine whether the stability of the IL-3 transcripts was altered in the factor-independent subclones, the half-life of the IL-3 mRNA was examined by Northern blot analysis (Figure 11). The half-life of the IL-3 mRNA was roughly . h,

Figure 9 Structure of the IL-3 transgene in an IL-3-dependent clone and IL-3-independent derivative subclones. DNAs were isolated and cleaved with the restriction enzyme BamHI, and Southern blot analysis was performed with an IL-3 cDNA probe. DNAs were prepared from lane: (1) FL5.12; (2) FL-IL3-R2; (3) FL-gIL3-LS-IAP-LTR 3′→5′C1; (4) FL-gIL3-LS-IAP-LTR 3′→5′C1.1; (5) FL-gIL3-LS-IAP-LTR 3′→5′C1.2; (6) FL-gIL3-LS-IAP-LTR 3′→5′C1.3; (7) FL-gIL3-LS-IAP-LTR 3′→5′C1.4; and (8) FL-gIL3-LS-IAP-LTR 3′→5′C1.5. Similar results were obtained when the DNAs were cleaved with EcoRI. Similar transgene integration fragment studies were performed with the IL-3- dependent FL-gIL3-LS-IAP-LTR 5′→3′C1, FL-gIL3-LS-IAP-LTR Figure 11 IL-3 mRNA half-life in an IL-3-independent subclone. 5′→3′C2, FL-gIL3-LS-IAP-LTR 5′→3′C3 and FL-gIL3-LS-IAP-LTR 3′→5′ mRNA was prepared from FL-gIL3-LS-IAP-LTR 3′→5′C1.1 cells grown and 15 IL-3-independent subclones and the subclones contained the in the presence of actinomycin D for varying period of time and then same IL-3 integration fragments as the parental IL-3-dependent clones. Northern blot analysis was performed with IL-3 and ␤-actin probes. Effects of LTRs on IL-3 expression X-Y Wang and JA McCubrey 1721 similar to that observed previously in Figure 4. Thus the a greater ability of PC-IAP-LTRs to stimulate gene transcription increase in IL-3 expression in the factor-independent sub- than the LS-IAP-LTR, which is consistent with the results clones was due to a transcriptional mechanism and did not observed with the transformation to cytokine-independence. appear to result from a post-transcriptional mechanism or The PC-IAP-LTRs differ considerably from the LS-IAP-LTR in rearrangement of the IL-3 transgene. the DNA sequences contained in the Enh1, Enh2, ATF/CRE, and TATA sites as well as the DNA sequences surrounding the TATA site.14,24,26 Plasmacytoma IAP-LTRs contain a typical Discussion ATF sequence, while the normal lymphocytes IAP-LTRs contain an altered ATF sequence.14,24,26 The LTRs present in LS- In the present study, we have investigated the effects of certain IAPs were relatively weak in stimulating sufficient IL-3 LTRs on activating the transcription of a germline IL-3 gene. expression and LS-IAPs transpositions may only transform In some cases, addition of the retroviral LTRs to the gIL3 gene hematopoietic cells when they interfere with another regulat- led to the autocrine transformation of IL-3-dependent cells. ory function of the affected gene. In the case of the LS-IAP This is a model for elucidating some of the steps in tumor transposition in the FL-IL3-R cells, the transposition also affec- progression of hematopoietic cells as the factor-dependent ted mRNA stability. cell lines did not induce tumors upon injection into nude mice When the MoMuLV-LTR was inserted 5′ to the IL-3 pro- whereas certain autocrine transformants which secrete IL-3 moter region, it could efficiently transform IL-3-dependent readily did.62,66 cells to IL-3-independent growth. This is not surprising, since Inheritance of an additional germline IL-3 gene was not nor- MoMuLV is a replication-competent exogenous retrovirus, mally sufficient to abrogate the cytokine dependency of IL-3- and can cause leukemia by insertional mutagenesis mech- dependent cells. The LS-IAP-LTR contained in endogenous anisms.79–83 The MoMuLV-LTR can serve as a cis-acting IAP retroviral genomes could abrogate the cytokine depen- element to induce gene expression when it is inserted a long dency of FL5.12 cells when linked to a gIL3 gene, but only at distance away from the affected gene. It has been observed a low frequency. This IAP-LTR did not normally stimulate high to induce evi-1 expression when it is inserted 90 kb away levels of IL-3 expression in transfected FL5.12 cells. Moreover, from the evi-1 gene locus in certain hematopoietic cell lines.84 it did not induce high levels of transcription in transient CAT Moreover, a MoMuLV-LTR can induce c-myc expression in T assays. Thus this IAP-LTR was relatively weak in stimulating cell lymphomas when it was inserted 30 to 270 kb away from sufficient IL-3 expression to induce autocrine transformation. c-myc locus.85 Rare factor-independent subclones could be recovered from The effects of the MoMuLV-LTR on the transcription rate of factor-dependent gIL3-LS-IAP-LTR transfected cells. These the IL-3 gene may be complex. Recent studies have shown subclones expressed relatively low levels of IL-3. However that the MoMuLV-LTR also encodes a trans-activator which certain subclones which expressed higher levels of IL-3 could induces the transcription and expression of the class I major induce tumors in immunocompromised mice whereas the histocompatibility locus as well as certain cytokine genes.86–88 cytokine-dependent clones from which they were derived did These genes often have Ap-1 sites in their promoter regions not. A transcriptional mechanism was responsible for the IL- and the MoMuLV-LTR increases cellular Ap-1 DNA binding 3 expression in the factor-independent subclones as the tran- activity.86–88 The IL-3 gene has an Ap-1 site in its promoter scription rate of the IL-3 gene was increased. A post-transcrip- region.89 Hence, the MoMuLV-LTR could stimulate IL-3 tional mechanism was ruled out since the cells expressed IL- expression by inducing Ap-1 activity. In addition, the 3 mRNA with a normal half-life. There did not appear to be MoMuLV-LTR also contains transcription factor binding sites any rearrangement of the IL-3 gene in these cells and the for the c-Ets proto-oncogene and a core site that binds the structure of other genes associated with signal transduction in polyoma virus enhancer-binding protein 2.86–88 These may be these cells appeared normal (eg IL-3R␣, IL-3R␤, Jak2, Stat-5, some of the important reasons why the MoMuLV LTR was one data not presented). The mechanism which results in the of the strongest enhancers with regard to abrogation of growth increase in transcription is unknown. It could be due to subtle factor dependency of IL-3-dependent hematopoietic cells. point mutations in the IAP-LTR or IL-3 promoter region. Alter- The TCR␦ enhancer, when ligated to the gIL3 gene, could natively there could be subtle mutations in oncogenes or other also induce autocrine transformation of IL-3-dependent genes involved in signal transduction and the prevention of FL5.12 cells. The TCR␦ gene is located within the TCR␣ locus, apoptosis. Additional experiments need to be performed to and has its own enhancer element (TCR␦ enhancer).64,90–94 identify whether there are mutations which can be correlated The core binding site within the TCR␦ enhancer, which is with the conversion to IL-3 expression and tumorigenicity. required for its transcription induction effect, shares similar In contrast, the PC-IAP-LTRs, when linked to a gIL3 gene, nucleotide sequence as the MoMuLV-LTR, and binds the same could transform cells to factor-independent growth at higher nuclear factors.74,91–93 Translocations involving TCR␦ chain frequencies than that observed with the gIL3 + LS-IAP-LTR are found frequently in certain leukemias, which result in transgene. LS-IAPs are normally expressed in lipopolysaccha- abnormal activation of adjacent genes.76–78 ride-stimulated normal B cells, while PC-IAPs are expressed Factor-independent cells were very infrequently recovered in tumor cells.14,26 Transpositions of PC-IAPs normally affect after transfection of IL-3-dependent cells with the gIL3 con- the expression of the adjacent genes more profoundly than taining the HIV, and HTLV LTRs. These results indicated that LS-IAPs. For example, in WEHI-3B cells, the transposed PCI- these human retroviral cis-acting elements are not sufficient IAP in the 5′ untranslated region of IL-3 gene resulted in a for transforming murine IL-3-dependent cells. This is consist- high level of IL-3 expression.11 Likewise, in the mast cell line ent with previous observations which indicated that retrovir- PB-3c, the transposed PCI-IAP also led to a high level of IL-3 ally encoded proteins, such as Tax, Rex, Tat, Nef, and Rev, expression.12 In contrast, the transposition of a LS-IAP to the play important roles in cell transformation and pathogenesis vicinity of IL-3 gene in FL-IL3-R cells only generated a moder- induced by these human retroviruses.37,94–98 Recent experi- ate level of IL-3 expression.13 Our transient transfection CAT ments undertaken to compare the effects of different regions assays with PC and LS IAP-LTRs in FL5.12 cells also indicated of the HIV and MoMuLV LTRs on HIV expression and patho- Effects of LTRs on IL-3 expression X-Y Wang and JA McCubrey 1722 genesis have revealed interesting results. The studies were per- ences also affect the frequency of transposition and expression formed by constructing recombinant retroviruses by substitut- of the IAPs. ing regions of the MoMuLV and HIV enhancers. The modified The WEHI-3B cell line contains multiple IAP transpo- retrovirus which contained the MoMuLV/HIV chimeric LTR, sitions.11,18 IAP transpositions have been detected in both the in which the MoMuLV-LTR enhancer replaced the NF␬B- IL-3 and Hox2.4 genes as well as other genes in WEHI-3B binding motifs, was transcriptionally active in murine cells cells.11,18 Preliminary observations indicate that there is a high that lack the tax gene.99 The transgenic mice which had the level of IAP expression in this cell line when compared to the retrovirus containing MoMuLV/HIV chimeric LTR expressed FL5.12 line (data not presented). These events may result in higher levels of virus than transgenic mice containing wild- the enhanced tumorigenicity of WEHI-3B cells. However, the type HIV.99 The MoMuLV/HIV chimeric LTR mice came down primary event responsible for initiating the expression and with disease faster than the wild-type HIV transgenic mice. It transposition of the PC-IAPs is unknown. appeared that the tissue specificities of the wild-type and Additional events such as disruption of mRNA stability chimeric LTRs in the mice were similar, with the possible sequences13,62,66 may be necessary for LS-IAPs to transform exception of lymphoid cells. This may be related to the differ- cells. This is the case of the LS-IAP transposition in the FL-IL3- ent tropisms of the two viruses. The HIV-LTR which contains R cells, as the IAP transposition also affected mRNA stability.13 the NF-␬B-binding sites may promote more expression in These autocrine-transformed cell lines proliferated in the macrophage cells,99 whereas the MoMuLV/HIV chimeric LTR absence of exogenous IL-3 to a maximal cell density that was may induce more transcription in lymphoid cells. decreased compared with either the parental or v-abl-transfor- It may be difficult to define precisely the nature of which med cells.100,101 The v-abl oncoprotein transformed FL5.12 enhancer sequences are important for gene expression since cells by a nonautocrine mechanism.60 The expression of the they may vary in cells of different lineages or differentiation IL-3R␣ and IL-3R␤ chain genes was also examined in these states. For example, certain cells may or may not express the cells.100,101 The expression of both the IL-3R␣ and ␤ chains NF-␬B, c-Ets, Ap-1, c-Myb, or other transcription factors. was increased in the autocrine transformants when compared Enhancer sequences which are important for LTR function in with the parental cell line.100–102 Thus in the autocrine-trans- macrophage cells may not be that critical for LTR function in formed FL-IL3-R line, there was an additional change involved lymphoid cells. There may also be functional redundancies, in the abrogation of cytokine dependency which was an up- as certain enhancer binding sites may be occupied in some regulation of cognate receptor gene expression. types of cells but not others. Additional sites may be occupied IAP proviral transpositions into the regulatory regions of in other cells which results in either activation or inactivation. cytokine genes have been documented and speculated to be A more prudent way to examine the effects of different involved in oncogenesis.11–13,17–19,102 However, in most cases enhancers on gene expression in a particular model system is the studies ceased at this point, since the parental cells lacking to transfect the cells with a relevant target gene and different the particular IAP transposition were not available. Our stud- enhancer elements. This is what has been presented in our ies directly demonstrate the important effects that IAP transpo- studies. We have examined the ability of various viral and sitions can have on induction of gene transcription. These cell cellular LTRs to stimulate IL-3 gene expression in lymphoid lines provide a model to determine how the aberrant IL-3-dependent cells. Similar studies comparing the expression of cytokine genes can result in leukemogenesis. efficiencies of autocrine transformation induced by LS-IAP, PC-IAP, MoMuLV and human retroviral LTRs in lymphoid cells have not been published previously. Acknowledgements Overall, our results indicated that the LS-IAP-LTR is much weaker than either the PC-IAP-LTR, MoMuLV-LTR and TCR␦ We appreciate the intellectual and technical assistance of PA enhancer in inducing sufficient IL-3 transcription for autocrine Algate, PE Hoyle, MW Mayo, KD Prevost, and LS Steelman. transformation. While these studies do not pinpoint the pre- We thank the following investigators for their gifts of plasmid cise sequences in the different LTRs which are important with DNAs: Drs MD Cole, MS Krangel, KK Lueders, S Lyman, K regard to abrogation of cytokine dependency, they do provide Shimotohno and I Young. This work was supported in part by a guide to the differences between the LTRs which are a grant (R01CA51025) from the National Institutes of Health important with regard to how IAPs and exogenous retroviruses to JAM. can cause insertional activation of cytokine genes and lead to autocrine transformation of IL-3-dependent hematopoietic cells. References Tumor progression is a process which involves the accumu- lation of mutations in multiple genes which permit the normal 1 Varmus H. Retroviruses. Science 1988; 240: 1427–1435. cell to escape the checks and balances that regulate and limit 2 Shih CC, Stoye JP, Coffin JM. Highly preferred targets for retro- cell growth. This report demonstrates that PC-IAP-LTRs had a virus integration. Cell 1988; 53: 531–537. considerably higher tendency to transform factor-dependent 3 Peters G. Oncogenes at viral integration sites. Cell Growth Diff cells than the LS-IAP-LTR when they were inserted in the 1990; 1: 503–510. 4 Hayward WS, Neel BG, Astrin SM. Activation of a cellular onc vicinity of a critical gene, in this case the germline IL-3 gene. gene by promoter insertion in ALV-induced lymphoid leukosis. Differences in nucleotide sequence have been found upon Nature 1981; 290: 475–480. comparison of PC vs LS IAP-LTRs.26 However, this may not 5 Kabrun N, Bumstead N, Hayman MJ, Enrietto PJ. Characteriz- be the sole factor responsible for the difference in the tran- ation of a novel promoter insertion in the c-rel locus. Mol Cell scription stimulation. Different methylation patterns have also Biol 1990; 10: 4788–4794. been observed between PC-IAP-LTRs and LS-IAP-LTR in cer- 6 van Ooyen A, Nusse R. Structure and nucleotide sequence of the 26 putative mammary oncogene int-1: proviral insertions leave the tain tumor cells. It is possible that diverse methylation pat- protein-encoding domain intact. Cell 1984; 39: 233–240. terns play a role in the differential transformation effects 7 Nusse R, van Ooyen A, Rijsewijk F, van Lohuizen M, Schuring, exerted by PC and LS IAP-LTRs. It is likely that these differ- E, van’t Veer L. Retroviral insertional mutagenesis in murine Effects of LTRs on IL-3 expression X-Y Wang and JA McCubrey 1723 mammary cancer. Proc Roy Soc Lond Ser B: Biol Sci 1985; 226: murine IAP-promoted placenta-expressed gene. Nucleic Acids 3–13. Res 1991; 19: 3667–3672. 8 Rovinski B, Munroe D, Peacock J, Mowat M, Berstein A, Benchi- 30 Poiesz BF, Ruscitti FW, Gazdar AF, Bunn PA, Minna JD, Gallo mol S. Deletion of 5′-coding sequences of the cellular p53 gene RC. Detection and isolation of type C retrovirus particles from in mouse erythroleukemia: a novel mechanism of oncogene fresh cultured lymphocytes of a patient with cutaneous T cell regulation. Mol Cell Biol 1987; 7: 847–853. lymphoma. Proc Natl Acad Sci USA 1980; 77: 7415–7419. 9 Hicks GG, Mowat M. Integration of Friend murine leukemia virus 31 Yoshida M, Miyoshi I, Hinuma Y. Isolation and characterization into both alleles of the p53 oncogene in an erythroleukemic cell of retrovirus from cell lines of human adult T cell leukemia and line. J Virol 1988; 62: 4752–4755. its implication in the disease. Proc Natl Acad Sci USA 1982; 79: 10 Selten G, Cuypers HT, Berns A. Proviral activation of the putative 2031–2035. oncogene Pim-1 in MuLV induced T cell lymphomas. EMBO J 32 Hinuma Y, Komoda H, Chosa T, Kondo T, Kohakura M, Tak- 1985; 4: 1793–1798. enaka T, Kikuchi M, Ichimaru M, Yunoki K, Sato I, Matsuo R, 11 Ymer S, Tucker WQJ, Sanderson CJ, Hapel AJ, Campbell HD, Takiuchi Y, Uchino H, Hanaoka M. Antibodies to adult T cell Young IG. Constitutive synthesis of interleukin-3 by leukaemia leukemia-virus-associated antigen (ATLA) in sera from patients cell line WEHI-3B is due to retroviral insertion near the gene. with ATL and controls in Japan: a nationwide sero-epidemiologic Nature 1985; 317: 255–258. study. Int J Cancer 1982; 29: 631–635. 12 Hirsch HH, Nair APK, Moroni C. Suppressible and nonsuppres- 33 Murphy EL, Hanchard B, Figueroa JP, Gibbs WN, Lofters WS, sible autocrine mast cell tumors are distinguished by insertion of Campbell M, Goedert JJ, Blattner WA. Modeling the risk of adult an endogenous retroviral element (IAP) into the interleukin 3 T cell leukemia/lymphoma in persons infected with human T gene. J Exp Med 1993; 178: 403–411. lymphotropic virus type I. Int J Cancer 1989; 43: 250–253. 13 Algate PA, McCubrey JA. Autocrine transformation of hemopo- 34 Grassmann R, Berchtold S, Radant I, Alt M, Fleckenstein B, Sod- ietic cells resulting from cytokine message stabilization after roski JG, Haseltine WA, Ramstedt U. Role of human T cell leuke- intracisternal A particle transposition. Oncogene 1993; 8: mia virus type I X-region proteins in immortalization of primary 1221–1232. human lymphocytes in culture. J Virol 1992; 66: 4570–4575. 14 Kuff EL, Lueders KK. The intracisternal A-particle gene family: 35 Grassmann R, Dengler C, Muller-Fleckenstein I, Fleckenstein B, structure and functional aspects. Adv Cancer Res 1988; 51: McGuire K, Dokhelar MC, Sodroski JG, Haseltine WA. Trans- 183–276. formation to continuous growth of primary human T lymphocytes 15 Keshet E, Schiff R, Itin A. Mouse retrotransposons: a cellular res- by human T cell leukemia virus type I X-region genes transduced ervoir of long terminal repeat (LTR) elements with diverse tran- by a Herpesvirus saimiri vector. Proc Natl Acad Sci USA 1989; scriptional specificities. Adv Cancer Res 1991; 56: 215–251. 86: 3351–3355. 16 Hawley RG, Shulman MJ, Hozumi N. Transposition of two differ- 36 Tanaka A, Takahashi C, Yamaoka S, Nosaka T, Maki M, Hat- ent intracisternal A particle elements into an immunoglobulin anaka M. Oncogenic transformation by the tax gene of human kappa-chain gene. Mol Cell Biol 1984; 4: 2565–2572. T cell leukemia virus type I in vitro. Proc Natl Acad Sci USA 17 Horowitz M, Luria S, Rechavi G, Givol D. Mechanism of acti- 1990; 87: 1071–1075. vation of the mouse c-mos oncogene by the LTR of an intracister- 37 Smith MR, Green WC. Type I human T cell leukemia virus tax nal A-particle gene. EMBO J 1984; 3: 2937–2941. protein transforms rat fibroblasts through the cyclic adenosine 18 Kongsuwan K, Allen J, Adams JM. Expression of Hox-2.4 homeo- monophosphate response element binding protein/activating box gene directed by proviral insertion in a myeloid leukemia. transcription factor pathway. J Clin Invest 1991; 88: 1038–1042. Nucleic Acids Res 1989; 17: 1881–1892. 38 Felber BK, Paskalis H, Kleinmann-Ewing C, Wong-Staal F, Pav- 19 Blatt C, Aberdam D, Schwartz R, Sachs L. DNA rearrangement lakis GN. The px protein of HILV-I is a transcriptional activator of a homeobox gene in myeloid leukaemic cells. EMBO J 1988; of its long terminal repeats. Science 1985; 229: 675–680. 7: 4283–4290. 39 Fujisawa J, Seiki M, Kiyokawa T, Yoshida M. Functional acti- 20 Lamb BT, Satymoorthy K, Li L, Solter D, Howe CC. CpG methyl- vation of the human T cell leukemia virus type I transacting fac- ation of an endogenous retroviral enhancer inhibits transcription tor. Proc Natl Acad Sci USA 1985; 82: 2277–2281. factor binding and activity. Gene Expression 1991; 1: 185–196. 40 Seiki M, Inoue JI, Takeda T, Yoshida M. Direct evidence that 21 Lamb BT, Satyamoorthy K, Solter D, Basu A, Xu MQ, Weinmann p40x of human T cell leukemia virus type I is a transacting tran- R, Howe CC. A DNA element that regulates expression of an scriptional activator. EMBO J 1986; 5: 561–565. endogenous retrovirus during F9 cell differentiation is E1A- 41 Sodroski JG, Rosen CA, Haseltine WA. Trans-acting transcrip- dependent. Mol Cell Biol 1992; 12: 4824–4833. tional activation of the long terminal repeat of human T lympho- 22 Satyamoorthy K, Park K, Atchison ML, Howe CC. The intracister- cytes viruses in infected cells. Science 1984; 226: 177–179. nal A-particle upstream element interacts with transcription fac- 42 Ballard DW, Bohnlein E, Lowenthal JW, Wano Y, Franza BR, tor YY1 to activate transcription: pleiotropic effects of YY1 on Greene WC. HTLV-I tax induces cellular proteins that activate distinct DNA promoter elements. Mol Cell Biol 1993; 13: the ␬B element in the IL-2 receptor ␣ gene. Science 1988; 241: 6621–6628. 1652–1655. 23 Falzon M, Kuff EL. Isolation and characterization of a protein 43 Leung K, Nabel GJ. Interaction of a common cellular transcrip- fraction that binds to enhancer core sequences in intracisternal tion factor, ATF, with regulatory elements in both E1a- and cyclic A-particle long terminal repeats. J Biol Chem 1989; 264: AMP-inducible promoters. Proc Natl Acad Sci USA 1988; 85: 21915–21922. 3396–3400. 24 Falzon M, Kuff EL. Binding of the transcription factor EBP-80 44 Ruben SM, Potent H, Tan TH, Kawakami K, Roeder R, Haseltine mediates the methylation response of an intracisternal A-particle W, Rosen C. Cellular transcription factors and regulation of IL-2 long terminal repeat promoter. Mol Cell Biol 1991; 11: 117–125. receptor gene expression by HTLV-I tax gene product. Science 25 Kamat JP, Basu A, Satyamoorthy K, Xu MQ, Showe L, Howe CC. 1988; 241: 89–92. IPEB transcription factor regulating the intracisternal A particle 45 Siekevitz M, Feinberg MB, Holbrook N, Wong-Staal F, Greene gene during F9 cell differentiation is expressed at sites of lymph- WC. Activation of the interleukin-2 and interleukin-2 receptor oid development. Mol Reprod Dev 1995; 41: 8–15. (Tac) promoter expression by the transactivator (tat) gene product 26 Lueders KK, Fewell JW, Morozov VE, Kuff EL. Selective of human T cell leukemia virus type I. Proc Natl Acad Sci USA expression of intracisternal A-particle genes in established mouse 1987; 84: 5389–5393. plasmacytomas. Mol Cell Biol 1993; 13: 7439–7446. 46 Rosenblatt JD, Golde DW, Wachsman W, Jacobs A, Schmidt G, 27 Kuff EL, Fewell JW. Intracisternal A-particle gene expression in Quan S, Gasson JC, Chen ISY. A second HTLV-II isolate associa- normal mouse thymus tissue: gene products and strain-related ted with atypical hairy cell leukemia. New Engl J Med 1986; 315: variability. Mol Cell Biol 1985; 5: 474–483. 372–375. 28 Leiter EH, Kuff EL. Immunocytochemical demonstration of 47 Kalyanaraman VS, Sarngadharan MG, Robert-Guroff M, Miyoshi increased antigen (p73) in genetically diabetic mice. Am J Pathol I, Golde D, Gallo RC. A new subtype of human T cell leukemia 1984; 114: 46–55. virus (HTLV-II) associated with a T cell variant of hairy cell leuke- 29 Chang-Yeh A, Mold DE, Huang RCC. Identification of a novel mia. Science 1982; 218: 571–573. Effects of LTRs on IL-3 expression X-Y Wang and JA McCubrey 1724 48 Haseltine WA, Sodroski JG, Patarca R. Structure and function of bol esters induce different patterns of immediate–early gene the genome of HTLV. Curr Top Microbiol Immunol 1985; 115: expression in an interleukin-3-dependent cell line. Oncogene 177–209. Res 1991; 6: 1–12. 49 Shimotohno K, Wachsman W, Takahashi Y, Golde DW, Miwa 68 Melton DA, Krieg PA, Rebagliati MR, Maniatis T, Zinn K, Green M, Sugimura T, Chen ISY. Nucleotide sequence of the 3′ region MR. Efficient in vitro synthesis of biologically active RNA and of an infectious human T cell leukemia virus type II genome. RNA hybridization probes from plasmids containing a bacterio- Proc Natl Acad Sci USA 1984, 81: 6657–6661. phage SP6 promoter. Nucleic Acids Res 1984; 12: 7035–7056. 50 Shimotohno K, Takahashi Y, Shimizu N, Gojobori T, Golde DW, 69 Fung MC, Hapel AJ, Ymer S, Cohen DR, Johnson RM, Campbell Chen IS, Miwa M, Sugimura T. Complete nucleotide sequence HD, Young IG. Molecular cloning of cDNA for murine interleu- of an infectious clone of human T cell leukemia virus type II: an kin-3. Nature 1984; 307: 233–237. open reading frame for the protease gene. Proc Natl Acad Sci 70 Williams DE, Eisenman J, Baird A, Rauch C, Van Ness K, March USA 1985; 82: 3101–3105. CJ, Park LS, Martin U, Mochizuki DY, Boswell HS, Burgess GS, 51 Hagan KW, Ruiz-Echevarria MJ, Quan Y, Peltz SW. Characteriz- Cosman D, Lyman SD. Identification of a ligand for the c-kit ation of cis-acting sequences and decay intermediates involved proto-oncogene. Cell 1990; 63: 167–174. in nonsense-mediated mRNA turnover. Mol Cell Biol 1995; 15: 71 Sheng-Ong GLC, Keith EJ, Piccoli SP, Cole MD. Novel myc onco- 809–823. gene RNA from abortive immunoglobulin-gene recombination in 52 Hjelle B, Mills R, Swenson S, Mertz G, Key C, Allen S. Incidence mouse plasmacytomas. Cell 1982; 31: 443–452. of hairy cell leukemia, mycosis fungoides, and chronic lympho- 72 Sleigh MJ. A nonchromatographic assay for expression of the cytic leukemia in first known HTLV-II-endemic population. J chloramphenicol acetyltransferase gene in eucaryotic cells. Anal Infect Dis 1991; 163: 435–440. Biochem 1986; 156: 251–256. 53 Hjelle B, Chaney R. Sequence variation of functional HTLV-II 73 Rosenthal N. Identification of regulatory elements of cloned tax alleles among isolates from an endemic population: lack of genes with functional assays. Meth Enzymol 1987; 152: 704– evidence for oncogenic determinant in tax. J Med Virol 1992; 720. 36: 136–141. 74 Ho I-C, Yang L-H, Morle G, Leiden JM. A T cell-specific tran- 54 Martin MP, Biggar RJ, Hamlin-Green G, Staal S, Mann D. Large scriptional enhancer element 3′ of C␣ in the human T cell recep- granular lymphocytosis in a patient infected with HTLV-II. AIDS tor ␣ locus. Proc Natl Acad Sci USA 1989; 86: 6714–6718. Res Hum Retrovir 1993; 9: 715–719. 75 Redondo JM, Pfohl JL, Hernandez-Munain C, Wang S, Speck NA, 55 Semmes OJ, Majone F, Cantemir C, Turchetto L, Hjelle B, Jeang Krangel MS. Indistinguishable nuclear factor binding to func- K-T. HTLV-I and HTLV-II tax: difference in induction of tional core sites of the T cell receptor ␦ and murine leukemia micronuclei in cells and transcription activation of viral LTRs. virus enhancers. Mol Cell Biol 1992; 12: 4817–4823. Virol 1996; 217: 373–379. 76 Kasai M, Maziarz RT, Aoki K, Macintyre E, Strominger JL. Mol- 56 Barre-Sinoussi F, Chermann JC, Rey F, Nugeyre MT, Chamaret ecular involvement of the pvt-1 locus in a ␥/␦ T cell leukemia S, Gruest J, Dauguet C, Axler-Blin C, Vezinet-Brun F, Rouzioux bearing a variant t(8;14)(q24;q11) translocation. Mol Cell Biol C, Rozenbaum W, Montagnier L. Isolation of a T lymphocyte 1992; 12: 4751–4757. retrovirus from a patient at risk for acquired immune deficiency 77 Bernard O, Barin C, Charrin C, Mathieu-Mahul D, Berger R. syndrome (AIDS). Science 1983; 220: 868–871. Characterization of translocation t(1;14)(p32;q11) in a T and a B 57 Gallo RC, Salahuddin SZ, Popovic M, Shearer GM, Kaplan M, acute leukemia. Leukemia 1993; 7: 1509–1513. Haynes BF, Palker TJ, Redfield R, Oleske J, Safai B, White G, 78 Duro D, Bernard O, Della Valle V, Berger R, Larsen CJ. A new Foster P, Markham PD. Frequent detection and isolation of cyto- type of p16INK4/MTS1 gene transcript expressed in B cell malig- pathic retroviruses (HTLV-III) from patients with AIDS and at risk nancies. Oncogene 1995; 11: 21–29. of AIDS. Science 1984; 224: 500–503. 79 Gonda TJ, Cory S, Sobieszczuk P, Holtzman D, Adams JM. Gen- 58 Boyle MJ, Goldstein DA, Frazer IH, Sculley TB. How HIV pro- eration of altered transcripts by retroviral insertion within the c- motes malignancies. Med J Aust 1996; 234: 47–52. myb gene in two murine monocytic leukemias. J Virol 1987; 61: 59 McKearn JP, McCubrey JA, Fagg B. Enrichment of hematopoietic 2754–2763. precursor cells and cloning of multipotential B lymphocyte pre- 80 Selten G, Guypers HT, Zijlstra M, Melief C, Berns A. Involvement cursors. Proc Natl Acad Sci USA 1985; 82: 7414–7418. of c-myc in MuLV-induced T cell lymphomas in mice: frequency 60 McCubrey JA, Holland G, McKearn J, Risser R. Abrogation of and mechanism of activation. EMBO J 1984; 3: 3215–3222. factor-dependence in two IL-3-dependent cell lines can occur by 81 Shen-Ong GLC, Morse HC III, Potter M, Mushinski JF. Two two distinct mechanisms. Oncogene Res 1989; 4: 97–109. modes of c-myb activation in virus-induced mouse myeloid 61 Dexter TM, Garland J, Scott D, Scolnick E, Metcalf D. Growth tumors. Mol Cell Biol 1986; 6: 380–392. of factor-dependent hemopoietic precursor cell lines. J Exp Med 82 Weinstein Y, Cleveland JS, Askew DS, Rapp UR, Ihle JN. Inser- 1980; 152: 1036–1047. tion and truncation of c-myb by murine leukemia virus in a 62 Mayo MW, Wang XY, Algate PA, Arana GF, Hoyle PE, Steelman myeloid cell line derived from cultures of normal hematopoietic LS, McCubrey JA. Synergy between AUUUA motif disruption and enhancer insertion results in autocrine transformation of interleu- cell. J Virol 1987; 61: 2339–2343. kin-3-dependent hematopoietic cells. Blood 1995; 86: 3139– 83 Christy RJ, Huang RCC. Functional analysis of the long terminal 3150. repeats of intracisternal A-particle genes: sequences within the 63 Van Beveren C, Goddard JD, Berns A, Verma IM. Structure of U3 region determine both the efficiency and direction of pro- Moloney murine leukemia viral DNA: nucleotide sequence of moter activity. Mol Cell Biol 1988; 8: 1093–1102. the 5′ long terminal repeat and adjacent cellular sequences. Proc 84 Bartholomew C, Ihle JN. Retroviral insertions 90 kilobases proxi- Natl Acad Sci USA 1980; 77: 3307–3311. mal to the Evi-1 myeloid transforming gene activate transcription 64 Shimotohno K, Takano M, Teruuchi T, Miwa M. Requirement of from the normal promoter. Mol Cell Biol 1991; 11: 1820–1828. multiple copies of a 21-nucleotide sequence in the U3 regions 85 Tsichlis PN, Lee JS, Bear SE, Lazo PA, Patriotis C, Gustafson E, of human T cell leukemia virus type I and type II long terminal Shinton S, Jenkins NA, Copelan NG, Huebner K, Croce C, Levan repeats for trans-acting activation of transcription. Proc Natl Acad G, Hanson C. Activation of multiple genes by provirus integration Sci USA 1986; 83: 8112–8116. in the M1vi-4 locus in T cell lymphomas induced by Moloney 65 Redondo JM, Pfohl JL, Krangel MS. Identification of an essential murine leukemia virus. J Virol 1990; 64: 2236–2244. site for transcriptional activation within the human T cell receptor 86 Wang XY, McCubrey JA. Regulation of interleukin 3 expression ␦ enhancer. Mol Cell Biol 1991; 11: 5671–5680. in normal and autocrine transformed hematopoietic cells. Int J 66 Wang XY, McCubrey JA. Malignant transformation induced by Oncol 1997; 10: 989–1001. cytokine genes: a comparison of the abilities of germline and 87 Choi SY, Faller DV. The long terminal repeats of a murine retro- mutated interleukin 3 genes to transform hematopoietic cells by virus encode a trans-activator for cellular genes. J Biol Chem transcriptional and post-transcriptional mechanisms. Cell Growth 1994; 269: 19691–19694. Diff 1996; 7: 487–500. 88 Weng H, Choi SY, Faller DV. The Moloney leukemia retroviral 67 McCubrey JA, Steelman LS, McKearn JP. Interleukin-3 and phor- long terminal repeat trans-activates Ap-1 inducible genes and Effects of LTRs on IL-3 expression X-Y Wang and JA McCubrey 1725 Ap-1 transcription factor binding. J Biol Chem 1995; 270: MM, Wigdahl B. Neuroglial-specific factors and the regulation of 13637–13644. retrovirus transcription. Adv Neuroimmunol 1994; 4: 305–318. 89 Choi SY, Faller DV. A transcript from the long terminal repeats 97 Seeler JS, Muchart C, Podar M, Gaynor RB. Regulatory elements of a murine retrovirus associated with transactivation of cellular involved in tax-mediated transactivation of the HTLV-I LTR. genes. J Virol 1995; 69: 7054–7060. Virology 1993; 196: 442–450. 90 Redondo JM, Hata S, Brocklehurst C, Krangel MS. A T cell-spe- 98 Holzmeister J, Ludewig B, Pauli G, Simon D. Sequence-specific cific transcriptional enhancer within the human T cell receptor binding of the transcription factor c-Ets1 to the human immuno- ␦ locus. Science 1990; 247: 1225–1229. deficiency virus type I long terminal repeat. Biochem Biophy Res 91 Sun W, Graves BJ, Speck NA. Transactivation of the Moloney Comm 1993; 197: 1229–1233. murine leukemia virus and T cell receptor beta-chain enhancers 99 Dickie P, Gazzinelli R, Chang LJ. Models of HIV type 1 proviral by cbf and ets requires intact binding sites for both proteins. J gene expression in wild-type HIV and MLV/HIV transgenic mice. Virol 1995; 69: 4941–4949. AIDS Res Hum Retrovir 1996; 12: 1103–1116. 92 Hsiang YH, Spencer D, Wang S, Speck NA, Raulet DH. The role 100 Algate PA, Steelman LS, Mayo MW, Miyajima A, McCubrey JA. of viral enhancer ‘core’ motif-related sequences in regulating T Regulation of the interleukin-3 (IL-3) receptor by IL-3 in the fetal cell receptor-gamma and -delta gene expression. J Immunol liver-derived FL5.12 cell line. Blood 1994; 83: 2450–2468. 1993; 150: 3905–3916. 101 Steelman LS, Algate PA, Blalock WL, Wang XY, Prevost KD, 93 Wang S, Wang Q, Crute BE, Melnikova IN, Keller SR, Speck NA. Hoyle PE, McCubrey JA. Oncogenic effects of overexpression of Cloning and characterization of the subunits of the T cell receptor the interleukin-3 receptor on hematopoietic cells. Leukemia and murine leukemia virus enhancer core-binding factor. Mol 1996; 10: 528–542. Cell Biol 1993; 13: 3324–3339. 94 Reddy P, Dasgupta P. Regulation of HIV-1 gene expression by 102 Wang X-Y, Steelman LS, McCubrey JA. Abnormal activation of cellular transcription factor. Pathobiology 1992; 60: 219–224. cytokine gene expression by intracisternal type A particle trans- 95 Verdin E, Van Lint C. Internal transcriptional regulator elements position: effects of mutations which result in autocrine growth in HIV and other retroviruses. Cell Mol Biol 1995; 41: 365–369. stimulation and malignant transformation. Cytok Cell Molec Ther 96 Tillmann M, Krebs FC, Wessner R, Pomeroy SM, Goodenow 1997; 3: 3–19.