Interferon-a Treatment of Daudi Cells Down-regulates the Octamer Binding Transcription/DNA Replication Factors Oct-1 and Oct-2*

Treatment of Daudi cells with a-interferon (a-IFN) results in a considerable decrease in the levels of the octamer-binding DNA replication/transcription factors Oct-1 and Oct-2 and specifically inhibits gene expression by octamer-containing promoters. The in-hibitory effect on octamer-binding proteins also occurs after culturing cells with phorbol 12-myristate 13-acetate but it does not occur following a-IFN treatment of an a-IFN-resistant variant of the Daudi cell line or of HeLa cells. We discuss the potential role of the decreased levels of octamer-binding proteins in the inhibition of cell proliferation. The interferons a group of proteins defined by their These factors also to or repress the expression of various genes and also to inhibit the of certain cell A transformed human lymphoblastoid cell line, to a-IFN’ in a number of When these cells are a-IFN the c-fgr and c-myc inhibited. the


Treatment of Daudi cells with a-interferon (a-IFN)
results in a considerable decrease in the levels of the octamer-binding DNA replication/transcription factors Oct-1 and Oct-2 and specifically inhibits gene expression by octamer-containing promoters. The inhibitory effect on octamer-binding proteins also occurs after culturing cells with phorbol 12-myristate 13acetate but it does not occur following a-IFN treatment of an a-IFN-resistant variant of the Daudi cell line or of HeLa cells. We discuss the potential role of the decreased levels of octamer-binding proteins in the inhibition of cell proliferation.
The interferons are a group of proteins defined by their ability to inhibit virus infection (1). These factors have also been shown to induce (2) or repress (3) the expression of various genes and also to inhibit the growth of certain cell types (4). A transformed human B lymphoblastoid cell line, Daudi (5, 6 ) , responds to a-IFN' in a number of ways. When these cells are cultured with a-IFN several genes are rapidly induced (7) while the expression of others, including immunoglobulin genes (8), c-fgr (9) and c-myc (7), are inhibited. In addition, the incorporation of exogenous thymidine into DNA declines dramatically and the cell cycle is arrested in Go/GI at the expense of the S and G,/M phases (10).
The octamer binding transcription factors Oct-1 and Oct-2 are likely to play a critical role in gene expression and DNA replication in B-cells (for review, see Ref. 11). For instance, the ubiquitous transcription factor Oct-1 is required both for the constitutive transcription of the snRNA genes and the cell cycle-specific expression of the histone H2B gene (12), while the B-cell specific Oct-2 protein also plays an essential role in the expression of the immunoglobulin genes (13, 14). Moreover, Oct-1 has been shown to be identical to nuclear fraction 111, a protein which is essential for the initiation of DNA replication (15-17). It has been shown that both Oct-1 and Oct-2 are capable of stimulating DNA replication in vitro (18), indicating that they are likely to be DNA replication factors as well as transcription factors.
* This work was supported by the Cancer Research Campaign and the Kay Kendall Leukemia Trust. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The factors Oct-1 and/or Oct-2 could be involved in mediating the effects of a-IFN on gene expression or on DNA replication in Daudi cells. We therefore investigated the effects of a-IFN treatment on the levels of Oct-1 and Oct-2 activity of octamer-containing gene promoters in the presence and absence of a-IFN.

Cell Culture
Wild-type Daudi cells (Daudi) and the a-IFN-resistant mutant of the same cell line (DaudiIR (10)) were cultured at 3-7 X lo6 cells/ml in RPMI-1640 (Gibco-BRL) supplemented with 5% heat-inactivated fetal calf serum. HeLa cells were cultured as a monolayer in E4 medium, 5% fetal calf serum. Where indicated, a-JFN (Wellferon) was added to a final concentration of 300 units/ml from a stock at 3.0 X lo6 units/ml. Phorbol 12-myristate 13-acetate (PMA, Sigma) was added from a 1 mM stock in dimethyl sulfoxide to a final concentration of 100 nM. Addition of dimethylsulfoxide to control extracts had no effect on the cell cycle.

Mobility Shift Assays
Nuclear extracts from 1-5 X lo' cells were prepared from each sample according to the method of Dignam et al. (19). Extracts were incubated with labeled oligonucleotide a (see below) which has a high affinity binding site for Oct-1 and Oct-2 and run-on nondenaturing acrylamide gels as previously described (20). Specificity of the Oct-1 and Oct-2 signals was determined by competing the labeled DNA with a 200-fold molar excess of unlabeled oligonucleotide a, with a mutant octamer (ATAATAAT) which does not bind octamer proteins (21) (data not shown), or with an Spl-binding site (22).

Cell Cycle Analysis and BrdUrd Incorporation
To determine the proportion of cells undergoing DNA synthesis, Daudi cells were cultured with 3 PM bromodeoxyuridine (BrdUrd, Sigma) in RPMI for 0-24, 24-48, and 48-72 h after addition of a-IFN. Control Daudi cells without a-IFN were cultured with BrdUrd for 24 h. Cells were fixed in 70% ethanol for 10 min, resuspended in 1 M HC1 for 10 min, then immunolabeled using anti-BrdUrd monoclonal antibody (Becton-Dickinson) and fluorescein isothiocyanate-IgG conjugate (Sigma). Cells were finally stained with propidium iodide (23) and analyzed by flow cytofluorimetry (EPICS-C, Coulter) and the proportions of cells which had incorporated BrdUrd during the 24-h labeling period were determined as were the proportions of cells in the Go/G, phase of the cell cycle.

Plasmid DNAs
Constructs a-d contained different octamer oligonucleotides cloned upstream of the herpes simplex virus (HSV) thymidine kinase promoter driving the chloramphenicol acetyltransferase (CAT) gene in the vector pBLzCAT (24). Construct d contains a consensus octamer motif, ATGCAAATAA, which is found in immunoglobulin gene promoters and enhancers (21, 25). Construct b contains the octamerrelated TAATGARAT motif, GCGGTAATGAGAT, which is found in the HSV immediate-early 4/5 gene promoter (26). Constructs a and c contain overlapping octamer/TAATGARAT motifs of the type found in the HSV immediate-early 1 gene (27). Construct a has the sequence ATGCTAATGAGAT, while that in c is identical except for the substitution of T for G at the 11th position. The Rous sarcoma virus-CAT construct which does not contain an octamer sequence in its promoter is as described by Gorman et al. (28).

Transfection into Daudi Cells
10 pg of each DNA construct was transfected into 8 X lo6 Daudi cells by electroporation. Exponentially growing cells were pelleted and the conditioned medium was retained. The cells for each transfection were resuspended in 0.25 ml of RPMI 1640, HEPES, 5% fetal calf serum and pipetted into a 0.4-cm electroporation cuvette (Bio-Rad Gene Pulser). The DNA was added to the cells which were then electroporated a t 0.25 V, 960 pF ( T = 33-38). These cells were then put into a flask containing 8 ml of the retained medium with a 1-ml wide-necked pipette and then split into 4-ml aliquots in each of two flasks. Cells were left for 4-6 h and then a-IFN (300 units/ml) was added to one of the flasks. The cells were harvested after a further 48 h in culture and cell extracts were equalized according to protein content (29) and then assayed for CAT activity as described by Gorman (30). The activity was quantified by collecting the radioactive areas from the thin layer chromatography plates which were then subjected to scintillation counting. The data is presented as percent conversion of starting material to monoacetylated product.

PCR from RNA
cDNA Synthesis-Total RNA (1 pg) was denatured a t 75 "C for 2 min and then chilled on ice. PCR buffer, 1 mM deoxynucleotide triphosphates, 1 unitlpl of RNasin, 100 pmol of random hexamers, and 10 units/pl of Moloney murine leukemia virus reverse transcriptase was then added to a final volume of 20 pl and incubated a t 37 "C for 1 h.
PCR-80 p1 of PCR buffer containing 50 pmol of each upstream (1 below) and downstream (2 below) primer and 0.05 unit/pl of Taq polymerase (Cetus) was then added. 21 PCR cycles were carried out, each cycle being: 30 s a t 94 "C, 30 s a t 54 "C, 30 s a t 72 "C. Then 50 pl of each sample was removed for Southern blot analysis and the remaining sample underwent a further 5 cycles.

Oct-1 and Oct-2 Binding with DNA: Mobility Shift Assays-
After culturing Daudi cells with a-IFN for 48 h a clear decrease was observed in the level of the protein-DNA complex formed by Oct-1 and a much larger decrease in the smaller complex formed by Oct-2. This effect was observed in several different extract preparations, two of which are illustrated in Fig. 1, panels a and b. No down-regulation of Oct-1 or Oct-2 binding occurred following treatment with a-IFN for up to 24 h, suggesting that this effect requires relatively prolonged exposure to a-IFN. The specificity of the signals for the octamer-binding proteins was demonstrated by competition experiments using the identical untreated extract and an excess of unlabeled competitor containing the octamer motif or the unrelated binding site for the Spl transcription factor (Fig. 1, panel d ) . In these experiments no effect of a-IFN treatment was observed on the levels of other transcription factors, for example, AP1, TFIIIC, ATF (AP1 is shown in Fig. 1,panel c), which confirmed that this effect was specific to the octamer-binding proteins.
Interferon-a can have many different effects on cell physiology, including the induction of an anti-viral state and the inhibition of cell proliferation. We wished to determine whether the changes in octamer binding were related to any resistant to the antiproliferative effects of a-IFN but which has functional a-IFN receptors and responds to treatment with a-IFN by the induction of IFN-induced genes (7). Interferon-a treatment of these cells did not result in any reduction in the levels of octamer-binding proteins (Fig. 2, panel A ) . Similarly no effect of a-IFN treatment on the level of Oct-1 was observed in extracts of HeLa cells (Fig. 2, panel B ) which similarly continue to proliferate in the presence of a-IFN and which respond to a-IFN by the induction of IFN-inducible genes (data not shown).
As expected these cells did not contain Oct-2 which is found only in B cells and some neuronal cells (13, 31).
The cell cycle of Daudi cells can also be inhibited in Go/GI by culturing with PMA (32). This agent is not known to induce an antiviral state or to induce the same subset of genes as a-IFN (data not shown). As with a-IFN a down-regulation of Oct-1 and, to a lesser extent, of Oct-2 binding was observed after culturing Daudi cells with PMA (Fig. 1).
Oct-1 and Oct-2 mRNA Leoels-In order to determine whether the effects we observed on Oct-1 and Oct-2 levels in mobility shift assays were mediated by changes in the corresponding mRNAs, we prepared mRNA from a-IFN and PMA Daudi cells. This mRNA was used to prepare cDNA and the cDNA was then used as a template in a PCR with either Oct-1 or Oct-2 specific primers. The results of this experiment (Fig. 3) showed that 48 h of a-IFN treatment produced a large decrease in the level of Oct-2 mRNA in Daudi cells, although no detectable fall in Oct-1 mRNA levels was observed.
A smaller but detectable fall in the level of Oct-2 mRNA was also observed after 20 h of a-IFN treatment (data not shown). As expected, no effect on the level of either Oct-1 or Oct-2 mRNAs was observed upon a-IFN treatment of interferonresistant Daudi" cells. In contrast, PMA treatment produced a large fall in the Oct-1 mRNA level in Daudi cells without apparently affecting Oct-2 mRNA levels. All samples showed similar levels of mRNA when primers specific for the actin transcript were used in the PCR (data not shown).
These data therefore parallel the results of the mobility shift assays in which a-IFN produced a more dramatic effect on Oct-2 than Oct-1 levels, while PMA had the opposite effect. It may be that only gross changes in mRNA levels can be observed by the PCR method and that smaller changes in the mRNA level of Oct-1 in a-IFN-treated cells and of Oct-2 in PMA-treated cells were not detected in our experiments. However, the data would indicate that at least the most dramatic effects of these agents on the levels of octamerbinding proteins are mediated by corresponding changes at the mRNA level.
Cell Cycle and DNA Synthesis-The majority (87% f 5.6 S.D.) of actively dividing Daudi cells which were not cultured with a-IFN incorporated BrdUrd over a 24-h period (Fig. 4) -IFN (Fig. 4).
Gene Expression and Octamer-containing Promoters-To investigate the effect of a-IFN treatment on octamer-mediated gene expression in Daudi cells, we made use of a series of constructs in which the octamer motif is cloned upstream of the HSV thymidine kinase promoter in the vector pBL2CAT (24). Daudi cells were transfected with these constructs and CAT activity was assayed 48 h after the addition of a-IFN. As shown in Fig. 5 and 6, a-IFN treatment resulted in a 3-5-fold decrease in the activity of each of the octamercontaining promoters (constructs a-d). In contrast no effect of a-IFN treatment was observed on the activity of the parental pBL,CAT vector (Fig. 6, u ) or on that of an Rous sarcoma virus-CAT construct (28) (Fig. 5 , r ) indicating that this effect is specific to octamer-containing promoters. The vector pBL2CAT contains a truncated thymidine kinase pro-  0-24, 24-48, and 48-72  h. The percentage of cells in G , , / G , ( 0 ) at 0. 24, 48. and 72 h  moter which lacks octamer-binding sites (24), but it does contain two sites for the transcription factor Spl (33). Even though the expression is low, the data shown in Fig. 6, panel  u, indicates that a promoter containing Spl sites is unaffected by a-IFN treatment. Hence the decline in Oct-1 and Oct-2 levels following treatment of Daudi cells with a-IFN is associated with reduced activity of octamer-containing promoters.

DISCUSSION
We have shown here that culturing Daudi B-cells with a-IFN results in a decrease in the levels of the Oct-1 and Oct-2 transcription factors and a decrease in the transcriptional efficiency of octamer-containing promoters. The Oct-1 and Oct-2 proteins have been shown to be required for the transcription of certain genes (11-14), but they also play a role in DNA replication (15-18). Therefore the down-regulation of Oct-1 and Oct-2 caused by a-IFN could affect either process.
It is known that binding Oct-2 to the octamer motif in immunoglobulin gene promoters and enhancers plays a critical role in their B-cell-specific expression. The DNA sequences involved were identified by transfection studies with simple octamer-containing promoters (34) and it was shown that Oct-2 could activate the immunoglobulin promoter in non-B-cells in which it is normally silent (35). Thus the decrease in Oct-2 reported here should affect immunoglobulin gene expression in Daudi cells. There are conflicting reports as to whether there is increased (36) or decreased (8) immunoglobulin synthesis in Daudi cells following a-IFN treatment. However, the down-regulation of Oct-2 shown here probably does not account for the decrease in IgM heavy chain (p-chain) brought about by a-IFN in the short term, since nuclear run-on experiments have suggested that this effect is not mediated via decreased transcription of p-chain mRNA (8). Furthermore, Oct-2.DNA complex formation is down-regulated between 24 and 48 h after the addition of a-IFN (Fig. l), well after the synthesis of the p-chain has ceased (8). Similarly, since both c-fgr (9) and c-myc (Ref. 7 and data not shown) are down-regulated rapidly after the addition of a-IFN, the Oct-1/-2 proteins probably do not play a part in these processes either. However, as we have shown here that a-IFN inhibits transcription from simple octamer containing constructs (Figs. 5 and 6), the down-regulation of the Oct-l/ -2 transcription factors must have long term effects on gene expression in Daudi cells, although the genes affected have yet to be identified.
The ubiquitous Oct-1 protein plays a critical role both in the transactivation of the HSV immediate-early genes during lytic infection (37,38). Transactivation of the immediateearly genes is dependent on the formation of a complex between Oct-1 and the virion protein Vmw65, which binds to the octamer-related TAATGARAT motif in the immediateearly promoters (37,39). This transactivation event is inhib-ited in a-IFN-treated cells, resulting in a lack of immediateearly gene expression and a consequent failure of the viral lytic cycle (40,41). It is possible therefore that the downregulation of Oct-1 we have observed may be responsible for this effect. In agreement with this idea, three of the constructs (a, b, and c) whose down-regulation we observed in a-IFNtreated cells contain TAATGARAT sequences found in the immediate-early promoters (26,27) either alone ( b ) or overlapping an octamer motif ( a and c). In addition, a construct in which the HSV immediate-early 3 promoter drives the CAT gene is also down-regulated following a-IFN treatment of Daudi cells (data not shown). However, Oct-1 is not decreased in HeLa or DaudiIR cells after culturing with a-IFN, even though these cells have a-IFN receptors and respond by the induction of IFN-inducible genes (7, 42, 43, and data not shown). Therefore the down-regulation of Oct-1 by a-IFN which occurs in Daudi cells may not in general be part of the antiviral mechanisms induced by a-IFN.
Interferon-a not only regulates gene expression in Daudi cells but also causes the inhibition of cell-cycle progression by arresting cells in Go/G1 (10). However, culturing the DaudiIR or HeLa cell lines with a-IFN does not lead to the inhibition of the cell cycle and these cells continue to proliferate normally. The decline in Oct-1/-2 levels brought about by a-IFN occurs in Daudi but not in DaudiIR or HeLa cell lines ( Figs. 1 and 2). The timing of this effect correlates with the inhibition of cell cycle progression and DNA synthesis (Fig. 4). In addition, PMA also down-regulates Oct-1/-2 levels (Fig. 1). This agent is not known to induce an antiviral state and it does not induce the same subset of genes as a-IFN. However, culturing Daudi cells with PMA causes an inhibition of proliferation and cells accumulate in Go/G1 (32). Therefore, the decline in complex formation between Oct-1/-2 and DNA may be part of the mechanism by which a-IFN inhibits DNA replication.
Interestingly, while a-IFN produced a more dramatic decline in Oct-2 levels compared to Oct-l, the reverse effect was observed with a PMA where Oct-1 levels were more reduced. These effects were paralleled at the RNA level where the levels of Oct-2 mRNA in a-IFN-treated cells and of Oct-1 mRNA in PMA-treated cells were greatly reduced compared to those observed in untreated cells. Hence the most dramatic effects of these agents on the levels of octamer-binding protein observed in DNA mobility shift assays are probably produced by a decline in the levels of their corresponding mRNAs rather than, for example, by a change in the DNA binding ability of the protein. Our inability to observe any change in the Oct-1 mRNA level following a-IFN treatment or in the Oct-2 mRNA level following PMA treatment may reflect the inability of our PCR procedure to detect relatively small alterations in mRNA levels. Alternatively, the smaller changes in protein binding observed in these cases may be produced by changes in the translation of pre-existing mRNA or in the DNA binding ability of pre-existing protein without any alteration in mRNA level.
The declining levels of octamer-binding proteins which we have observed could affect cell division either directly or indirectly. An indirect mechanism might involve inhibiting histone H2B synthesis, which is regulated by Oct-1 (12), since cells with artificially lowered levels of this histone have been shown to grow more slowly than normal cells (44). Alternatively, since both Oct-1 and Oct-2 can act as DNA replication factors (18), it is possible that the decline in their abundance directly affects DNA replication. Another octamer-binding factor, Oct-3, has been shown to be required for DNA replication to occur in mouse embryonic cells (45). Therefore,