mismatched double-stranded RNA - NCBI

2 downloads 0 Views 1MB Size Report
... Broad and Vine Streets, Philadelphia, PA 19102; *Department of Microbiology, Howard ..... Zar, J. H. (1974) BiostatisticalAnalysis (Prentice-Hall, Engle-.
Proc. Nati. Acad. Sci. USA Vol. 88, pp. 906-910, February 1991 Cell Biology

Cyclic AMP mediates the direct antiproliferative action of mismatched double-stranded RNA (antiproliferative effects/interferon independence/adenylate cyclase)

HOWARD R. HUBBELL*t, JOHN E. BOYER*, PHILIP ROANEO, AND RONALD M. BURCH§ *Department of Neoplastic Diseases, Hahnemann University, Broad and Vine Streets, Philadelphia, PA 19102; *Department of Microbiology, Howard University College of Medicine, Washington, DC 20059; and fLaboratory of Cell Biology, National Institutes of Mental Health, Bethesda, MD 20892

Communicated by Maurice R. Hilleman, October 31, 1990 (received for review December 29, 1989)

ABSTRACT Previous experiments have demonstrated that double-stranded RNAs (dsRNAs) can exert an antiproliferative effect on human tumor cells, independent of interferon (lEN) induction. However, the mechanism by which dsRNAs inhibit tumor growth has not been elucidated. As a first step in determining the molecular events responsible for growth arrest, we have explored the role of signal transduction through the cAMP system in the antiproliferative effect of the mismatched dsRNA, r(I),Wr(CI2,U). (Ampligen). These studies utilized the human glioma cell line A1235, which does not produce detectable levels of IFN-a, -P, or -y in response to mismatched dsRNA treatment. Treatment of A1235 cells with mismatched dsRNA in combination with either 1-(5isoquinolinesulfonyl)-2-methylpiperazine (H-7), which inhibits cAMP-dependent protein kinase and protein kinase C, or N-(2-guanidinoethyl)-5-isoquinolinesulfonamide (HA1004), which preferentially inhibits the cAMP-dependent protein kinase, yielded an antagonism of the mismatched dsRNAinduced antiproliferative effect. Measurement of adenylate cyclase activation showed a dose-dependent increase in activity at antiproliferative mismatched dsRNA concentrations, but not at lower, nonantiproliferative doses. This increase in activity was rapid, seen as early as 30 sec after initiation of treatment, and it was sustained at peak levels for 1-2 hr. Analysis of the intracellular cAMP concentration gave similar kinetics of induction. Exposure of cells to the stable cAMP analogue dibutyryl cAMP yielded dose-dependent inhibition of cell growth. The cAMP phosphodiesterase inhibitor 3-isobutyl-1methylxanthine also inhibited proliferation. In contrast, neither H-7 nor HA1004 had an effect on growth inhibition induced by human natural IFN-a treatment. In addition, antiproliferative doses of IFN-a did not increase cAMP concentrations. These results indicate that the cAMP system is utilized by mismatched dsRNA as an early signal transduction mechanism for growth control. Furthermore, the antiproliferative effects induced by mismatched dsRNA and IFN can occur by different mechanisms of action.

(7, 8), other reports have indicated that cAMP is not related to DNA synthesis or growth inhibition (9-13). Double-stranded (ds) RNAs are potent inducers of IFNs (14) and many cellular effects of dsRNAs are due to these induced IFNs. One of the cellular effects of mismatched dsRNA is the inhibition of tumor cell growth, which has been seen both in human tumor cell lines and in fresh human tumor cells in the clonogenic assay (15-19). However, recent evidence indicates that dsRNAs can inhibit cellular growth independent of IFN induction. Tumor cells show differential sensitivities to IFNs and dsRNAs, anti-IFN antibodies do not interfere with the antiproliferative effects of dsRNAs, and IFNs and dsRNA can synergistically inhibit tumor cell growth (15-18). We have utilized a human glioma cell line in which the antiproliferative effects of dsRNAs are not dependent on IFN production to study signal transduction mechanisms. We report here that a rapid increase in the intracellular concentration of cAMP is sufficient for dsRNA-induced growth inhibition. In contrast, intracellular cAMP is not increased after growth inhibition by natural human IFN-a.

MATERIAUS AND METHODS Reagents. The mismatched dsRNA, r(I)Dr(C12,U)n was obtained from HEM Research (Rockville, MD). The lyophilized powder was reconstituted with pyrogen-free water, heated to 650C, and cooled to room temperature to allow the homopolymer strands to reanneal. The dsRNA was then stored at -700C until use. Natural human IFN-a [specific activity, 1.7 x 106 international reference units (IRU)/mg] and IFN-3 (specific activity, 1 x 108 IRU/mg) were purchased from the New York Blood Center and the Department of Biological Resources, Roswell Park Memorial Institute (Buffalo, NY), respectively. Recombinant IFN--y (specific activity, 2.1 x 107 IRU/mg) was the gift of Biogen. All anti-IFN antibodies and control antisera were obtained from the Research Resources Branch of the National Institute of Allergy and Infectious Diseases. The protein kinase inhibitors 1-(5-isoquinolinesulfonyl)-2-methylpiperazine (H-7) and

Interferons (IFNs) have pleiotropic effects on cells in both tissue culture and in vivo, including antiviral, antiproliferative, and immunomodulatory effects (1-3). Although the biochemical mechanisms of the actions of IFN are not known, it is known that the effects are receptor mediated (4-6). Since the IFN is degraded shortly after its interaction with the receptor and internalization (4-6), receptorassociated signal transduction mechanisms may play a role in IFN-induced events. Conflicting results have been reported for the role of the cAMP signal transduction system in IFN-induced growth inhibition. Although it has been suggested that IFN-associated increases in intracellular cAMP concentration may mediate the inhibition of DNA synthesis

N-(2-guanidinoethyl)-5-isoquinolinesulfonamide (HA1004) (20) were purchased from Seikagaku America (St. Petersburg, FL). Dibutyryl cAMP and isobutylmethylxanthine (IBMX) were purchased from Sigma. Cell Lines. Cell lines were obtained as described (15, 16). The human grade IV astrocytoma cell line A1235 was maintained on Dulbecco's modified Eagle's medium (high glucose) supplemented with 10Wo heat-inactivated fetal calf serum, 2 mM glutamine, penicillin, and streptomycin. RT4, Abbreviations: dsRNA, double-stranded RNA; IFN, interferon; H-7, 1-(5-isoquinolinesulfonyl)-2-methylpiperazine; HA1004, N-(2guanidinoethyl)-5-isoquinolinesulfonamide; IBMX, 3-isobutyl-1methylxanthine; IRU, international reference unit(s); 2-SA, 2',5'-

The publication costs of this article were defrayed in part by page charge

payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

oligoadenylate. tTo whom reprint requests should be addressed.

906

Cell Biology: Hubbell et al. derived from a human transitional cell bladder carcinoma, was grown on RPMI 1640 medium supplemented as described above. These monolayer cultures were subcultured with a 0.25% trypsin/EDTA solution. Antiproliferative Assays. Cell growth inhibition assays were carried out as described (15-17), except that cell counts were determined 24 hr after the initiation of treatment. Statistical analysis of growth inhibition was performed with a two-tailed Student's t test (21). IFN Assays. A1235 cells (1 x 105 cells per dish) were treated with 200 Ag of mismatched dsRNA per ml for 2, 4, 6, or 8 hr. The mismatched dsRNA was removed and the cells were washed three times with Hanks' balanced salt solution. Fresh medium was added for 24 hr and then collected for measurement of IFN titers. IFN activity was determined by the cytopathic effect assay (22) by Biofluids (Rockville, MD). cAMP and Adenylate Cyclase Assay. Cultured cells in 35-mm Petri dishes were treated with mismatched dsRNA or natural IFN-a for various lengths of time. The medium was removed and the cAMP was solubilized with 200 ,ul of 0.1 M HCl (23). The samples were dried by evaporation under vacuum, redissolved in 50 mM sodium acetate/8 mM EDTA/ 0.1% sodium azide, pH 6.2, and assayed by RIA, using the more sensitive acetylated assay, as described by the manufacturer (New England Nuclear). Adenylate cyclase activity was assayed as described (24) by measuring the production of cAMP by cell sonicates. The A1235 cells were quick frozen on dry ice and stored at -70°C until assayed. All protein determinations were obtained by the Folin/phenol reagent method (25). In all cAMP and adenylate cyclase experiments, parallel cell counts at 24 hr were determined to verify the drug-induced antiproliferative effects.

RESULTS dsRNA Does Not Induce IFN Production in A1235 Cells. To interpret the results of the antiproliferative effect of mismatched dsRNA alone, it was necessary to determine whether this dsRNA could induce IFN production in the A1235 cells. We have previously reported (16) that IFN-a and IFN-f3 can directly inhibit the growth of A1235 cells. Cells were plated in 35-mm Petri dishes by the same protocol used for antiproliferative assays. dsRNA treatment of A1235 cells for 2, 4, 6, or 8 hr followed by a 24-hr incubation in tissue culture medium yielded no detectable IFN antiviral activity (limit of detection, 5 IRU/ml) in the medium. IFN activity was assayed on both human foreskin fibroblasts and WISH human amnion cells, which can preferentially detect IFN-a and IFN-f3, or IFN-y, respectively. As a further indication of the lack of involvement of IFNs in the antiproliferative effect of mismatched dsRNA, A1235

Proc. Natl. Acad. Sci. USA 88 (1991)

907

cells were incubated with neutralizing polyclonal antibodies to IFNs along with mismatched dsRNA and the percentage

of control growth was determined 24 hr later. As shown in Table 1, the anti-IFN antibodies did not inhibit the antiproliferative effects of mismatched dsRNA (P < 0.001 compared to antibody treatment alone), similar to previously reported results (15). Treatment of cells with IFN-a, -f3, and -r was used to demonstrate the effectiveness ofthe antibodies in this system. Since A1235 cells are not sensitive to IFN-y (16), RT4 cells, which are sensitive to IFN-y (17), were used to show neutralization by anti-IFN-y antibody. Mismatched dsRNA-Associated Signal Transduction Mechanisms. As a first test of a possible role of signal transduction mechanisms in the antiproliferative effect of mismatched dsRNA, A1235 cells were treated with mismatched dsRNA in the presence of H-7 and HA1004. H-7 inhibits both protein kinase C and cAMP-dependent protein kinase, while HA1004 preferentially inhibits the cAMP-dependent protein kinase (20). As shown in Table 2, both H-7 and HA1004 antagonized the antiproliferative effect of two growth-inhibiting concentrations of mismatched dsRNA (P < 0.05 compared to mismatched dsRNA treatment alone). The slight decrease in percentage control growth seen in A1235 cells treated with 25 ,uM H-7 may be due to the antiproliferative effects of this agent, since higher concentrations of H-7 (50 or 100 ,uM) can cause significant growth inhibition. These results suggest that the cAMP-dependent protein kinase may play a role in A1235 cell growth inhibition. Antiproliferative Doses of Mismatched dsRNA Induce Adenylate Cyclase and Increase Intracellular cAMP Concentration. To assess the role of adenylate cyclase in the antiproliferative effect of mismatched dsRNA, A1235 cells were treated with antiproliferative doses (25 and 200 ,ug/ml) or a non-growth-inhibitory dose (1 ,ug/ml) of mismatched dsRNA. Induction of adenylate cyclase activity occurred in a dose-dependent manner, with no significant increase in activity seen in the cells treated with the nonantiproliferative dose, as compared to untreated controls (Fig. 1). The induction was rapid, with measurable increases seen after 30 sec of treatment. The increase in adenylate cyclase activity peaked at 2-10 min and remained constant for 1-2 hr, followed by a decay in activity to approximately baseline levels over the course of 24 hr. The rapid induction of adenylate cyclase activity implies that intracellular cAMP concentrations should be similarly regulated. In independent experiments, intracellular cAMP concentration was measured by using the same dosage and time course regimen that was used in the adenylate cyclase assays. Treatment of A1235 cells with antiproliferative doses of mismatched dsRNA induced dose-dependent increases in cAMP concentration (Fig. 2). The intracellular cAMP concentration was significantly increased at 30 sec with a peak at

Table 1. Antibodies to IFNs do not inhibit mismatched dsRNA-induced antiproliferation in A1235 cells Antibody (240 neutralizing units/ml) Treatment None Anti-IFN-a Anti-IFN-y Anti-IFN-P A1235 cells None 91.8 ± 5.9 90.8 ± 7.4 87.8 ± 3.1 Mismatched dsRNA (200 pg/ml) 59.9 ± 4.6 55.7 ± 2.1* 50.8 ± 1.7* 56.5 ± 2.0* IFN-a (100 IRU/ml) 41.4 ± 4.6 91.8 ± 3.0 ND ND 40.9 ± 2.3 ND 84.9 ± 3.9 ND IFN-,8 (100 IRU/ml) RT4 cells None ND ND 108.9 ± 5.6 70.2 ± 3.8 IFN-y (25 IRU/ml) ND ND 99.1 ± 7.1 Cell growth inhibition assays were performed as described (15). Results are expressed as % control growth at 24 hr + SD. % control growth = {[treated cells (24 hr) - control cells (0 hr)]/[control cells (24 hr) - control cells (0 hr)]} x 100. ND, not done. *P < 0.001 compared to antibody treatment alone.

908

Cell Biology: Hubbell et al.

Proc. Natl. Acad. Sci. USA 88 (1991)

Table 2. Antagonism of mismatched dsRNA antitumor effect by H-7 and HA1004 on A1235 cells Mismatched dsRNA 0 ,g/ml 25 ,g/ml 200 Itg/ml H-7, ,uM 0 66.7 ± 12.6 48.2 ± 5.4 5 103.3 ± 13.6 108.3 ± 15.5* 75.2 ± 4.5* 25 87.5 ± 6.7 89.7 ± 11.0 72.8 ± 7.9* HA1004, AM 5 112.7 ± 16.2 94.1 ± 13.4* 113.4 ± 3.2* 117.2 ± 16.7* 25 113.2 ± 14.3 101.7 ± 6.8* Results are expressed as % control growth at 24 hr ± SD. *P < 0.05 compared to mismatched dsRNA treatment alone.

0.28-

c 0.24 -

s

a0.20-

0.1 0.16 0.12

cE

600

A

-

>

Un *

-

0

E

0.08

u.u04 -lmit ofdetection ............. ................................................................

01

5-10 min, which remained constant for =1 hr. The cAMP

concentration then returned to baseline levels by the end of the 24-hr time course. Cells were also treated with 1 ,ug of mismatched dsRNA per ml, but the intracellular cAMP concentration did not rise above the level of detection of the RIA assay (0.02 pmol ofcAMP per ,ug of protein) at any point during the time course studied. Untreated control cells also had cAMP concentrations below the level of detection. Dibutyryl cAMP and IBMX Inhibit A1235 Cell Growth. If the cAMP system is involved in mismatched dsRNA-induced growth regulation, then other agents that raise intracellular cAMP concentrations should also inhibit cell growth. A1235

H

2

5

lb

3

Time (min)

0.28

B

0.24

.s 0.2 a.

0.20 0.16 0.12

0.08 0.04

0.04l lm Ot ... . ....... . . ::::::::::::: :

500 detection .CI 8 .124 o

400

0

300

0

200

00E

0.5

100

QL .1 2

0:5

.E

Ex

5

10

30

i_

400-

200 0 -

a

0.

FIG. 2. Induction of intracellular cAMP levels in mismatched dsRNA-treated A1235 cells. Cells were treated with 25 (A) or 200 (m) ,ug of mismatched dsRNA per ml for various amounts oftime and the intracellular cAMP was solubilized with 0.1 M HCL. Untreated control cells and cells treated with 1 pg of mismatched dsRNA per ml were also assayed for cAMP concentration but were found to be consistently below the level of detection for this assay (0.02 pmol of cAMP per ,ug of protein). (A) Time course of activity from 0 to 30 min. (B) Time scale is changed to show results obtained by 0.5-24 hr of treatment.

D

300-

E

Time (hr)

cells were treated with dibutyryl cAMP and growth inhibition was measured 24 hr after the initiation of treatment. The dibutyryl cAMP caused a dose-dependent antiproliferative effect (Table 3). Cells were also treated with IBMX, which inhibits cAMP phosphodiesterase activity (26), yielding an accumulation of cAMP. IBMX inhibited cell growth in a 5Esdose-dependent and time-dependent manner (Table 3) with as

600-

c 5Q0 _ 0

Time(mi)

24

n

Table 3. Treatment of A1235 cells with dibutyryl cAMP or IBMX inhibits cell growth Treatment time

100-

0

o512 4

8

24

0.5 Time (hr)

FIG. 1. Induction of adenylate cyclase activity in mismatched dsRNA-treated A1235 cells. Cells were treated with 1 (o), 25 (A), or 200 (-) j.g of mismatched dsRNA per ml for various amounts of time and then quick frozen on dry ice. Adenylate cyclase activity was measured as pmol of cAMP produced in 15 min with sonicated cell preparations. (A) Time course of activity from 0 to 30 min. (B) Time scale is changed to show results obtained by 0.5-24 hr of treatment.

cAMP, ,uM Dibutyryl 25 100 250 500 1000 IBMX, zM 100 500

1 hr

24 hr

ND ND ND ND ND

87.2±4.0 77.5 55.8 47.8 39.3

± ± ± ±

1.5 9.5 7.2 8.0

75.8 ± 12.2 48.5 ± 2.3 67.0 ± 8.2 31.2 ± 4.4 Results are expressed as % control growth at 24 hr ± SD. ND, not done.

Cell

Biology: Hubbell et al.

Proc. Natl. Acad. Sci. USA 88 (1991)

little as 1 hr of treatment time necessary to induce antiproliferation. Analysis of intracellular dibutyryl cAMP concentrations, after treatment of A1235 cells with dibutyryl cAMP or IBMX, showed increases in cAMP levels similar to those induced by mismatched dsRNA (data not shown). Natural Human IFN-a Does Not Inhibit Cells Through Activation of the cAMP-Dependent Protein Kinase. The growth of A1235 cells can be inhibited by natural human IFN-a in a dose-dependent manner (16). Although dsRNA does not induce IFN-a in A1235 cells, the possibility exists that IFN-a may also work through the cAMP system. A1235 cells were treated with antiproliferative doses of IFN-a in combination with H-7 and HA1004, and cell counts were obtained after 24 hr of treatment. Neither H-7 nor HA1004 antagonized the antiproliferative effect of IFN-a (Table 4). In some IFN-a/H-7 combination treatments, significant differences were seen compared to IFN-a treatment alone (P < 0.05). However, these differences appear to represent the experimental variability seen in this biological system. As a further indication that cAMP is not involved in growth inhibition by IFN-a, intracellular cAMP concentrations were measured in A1235 cells treated with two antiproliferative doses (250 and 1000 IRU/ml) of IFN-a (16). At time points ranging from 1 min to 24 hr after initiation of IFN-a treatment, the intracellular cAMP concentration remained below the level of detection of the assay (data not shown).

DISCUSSION The direct antiproliferative effects of mismatched dsRNA can be exerted by both IFN-associated and IFN-independent mechanisms (15). The A1235 cells used here are inhibited by mismatched dsRNA in an IFN-independent manner. These studies demonstrated that mismatched dsRNA treatment can rapidly activate adenylate cyclase, leading to a similarly rapid increase in intracellular cAMP concentration. Treatment of cells with mismatched dsRNA for as little as 10 min can inhibit the growth of A1235 cells (data not shown). This short treatment time corresponds approximately to the amount of time needed to generate peak levels of cAMP. Other methods of elevating cAMP concentrations (e.g., dibutyryl cAMP, IBMX) can also inhibit the growth of A1235 cells in a dose-dependent manner. In addition, blocking cAMPdependent protein kinase activity inhibits the antiproliferative effects of mismatched dsRNA. Our results indicate that the increase in intracellular cAMP concentration is sufficient to inhibit the growth of A1235 cells. These results also imply that tumor cells that are resistant to the IFN-independent antiproliferative effects of mismatched dsRNA (15, 16) may be deficient in a component of the cAMP system. cAMP regulates gene transcription through the phosphorylation of proteins that bind to specific promoter/enhancer sequences known as cAMP-responsive Table 4. H-7 and HA1004 do not antagonize the antitumor effect of natural human IFN-a on A1235 cells IFN-a 0

100

250

1000

IRU/ml

IRU/ml

IRU/ml

IRU/ml

H7, AM 0 5 25

68.1 ± 8.8 50.1 6.1 43.7 5.1 78.4 ± 4.5 45.0 ± 2.5* 48.6 ± 2.6 56.2 ± 3.5* 70.3 ± 7.6 52.8 ± 1.6* 53.7 ± 7.5 50.0 ± 1.8

HA1004, ,M 5 92.8 ± 1.8 67.7 + 1.9 58.4 ± 4.5 54.2 + 5.3 25 70.3 + 4.1 57.2 ± 3.2 60.5 ± 5.0 51.7 ± 5.3 Results are expressed as % control growth at 24 hr ± SD. *P < 0.05 compared to IFN-a treatment alone.

909

elements. Genes responsive to cAMP can be divided into two types, based on the time frame of the response (27). Several genes have been shown to respond rapidly (within minutes) to cAMP (28-31). These genes have cAMP regulatory elements of the basic sequence TGACGTCA that lie within the first 150 base pairs of the 5'-flanking region of the gene (27). In addition, a second sequence has been identified, known as the activation protein 2 binding site, which acts as an inducible promoter site (32). This element has the basic sequence CCCCAGGC and can be rapidly induced by either cAMP or activation of protein kinase C. A second set of slow cAMPresponding (over several hours) genes has been identified (33, 34). The transcription of these genes requires protein synthesis (35). We are presently studying the effect of mismatched dsRNA treatment on the activation of cAMPresponsive elements in sensitive and resistant tumor cells. Recent evidence has similarly indicated that dsRNA can inhibit tumor cell growth by an IFN-independent mechanism (18). In this study, growth inhibition is associated with the induction of dsRNA-dependent 2',5'-oligoadenylate (2-SA) synthetase and subsequent degradation of rRNA, presumably through the 2-5A-dependent RNase L. The induction of 2-SA synthetase occurs >4 hr after the initiation of dsRNA treatment. Epinephrine and theophylline, which elevate intracellular cAMP concentrations, can also increase 2-SA synthetase activity and inhibit 2'-phosphodiesterase, which degrades 2-5As (36). Such results suggest that dsRNAinduced growth inhibition of tumor cells may utilize the 2-5A synthetase/RNase L system through the cAMP signal transduction mechanism. Previous studies have suggested that the dsRNA, poly(I)poly(C), may exert anticellular effects after cellular internalization (18, 37). However, the studies reported here argue against the internalization of mismatched dsRNA and implicate receptor-mediated events that lead to an antiproliferative effect. First, the activation of adenylate cyclase is known to be mediated through a variety of membrane receptors coupled to the guanine nucleotide-binding protein system (38-40). Second, the rapid induction of adenylate cyclase activity seen here, within 30 sec, suggests a more direct activation of the enzyme than dsRNA internalization followed by other intracellular events leading to the production of cAMP. Specific receptors for poly(I)-poly(C) have been described by using RK-13 cells and a poly(I)-poly(C)-resistant variant of RK-13, PR-RK cells (41). The data derived from studies of this pair of cells suggest that a specific poly(I)poly(C) receptor exists; however, nonspecific adsorption of poly(I)-poly(C) occurs on the cell membrane. Further work with these cells indicates that a 60-kDa protein can be detected that is a specific receptor for poly(I)-poly(C) (42). Specific binding of poly(I)-poly(C) has also been reported on murine B cells (43). However, Scatchard analysis of specific binding was not done in either of these studies. Therefore, it is not clear whether a specific dsRNA receptor has been identified. Combinations of IFN with mismatched dsRNA can synergistically inhibit tumor growth by direct antiproliferative mechanisms (15, 16). Similarly, the antitumor effect of IFN-f3 is potentiated by the cAMP phosphodiesterase inhibitor mopidamole (44). These synergistic effects suggest that IFN does not interact with the cAMP system at the same level as dsRNA or additive antitumor effects would be expected. Our results directly confirm the prediction that IFN and mismatched dsRNA utilize different antiproliferative mechanisms. Various studies have related the antitumor effects of IFN to diacylglycerol metabolism (45), dsRNA-dependent 2-SA synthetase (46-48), or dsRNA-dependent protein kinase (49). The potential relationship between the cAMP and 2-5A synthetase systems (36) are of particular interest. We are presently exploring the potential interactions of these

910

Cell Biology: Hubbell et al.

systems with the cAMP cascade to determine the mechanism of this potentiated growth inhibition. The authors thank Biogen Research for the gift of the recombinant IFN-yand Dee Valley for help in preparation of this manuscript. This work was supported in part by U.S. Public Health Service Grant 1 P01 CA29545 from the National Cancer Institute and by a grant from HEM Research. 1. Taylor-Papadimitriou, J. & Balkwill, F. R. (1982) Biochim. Biophys. Acta 695, 49-67. 2. Gresser, I. & Tovey, M. G. (1978) Biochim. Biophys. Acta 516, 231-247. 3. Taylor, J. L., Sabran, J. L. & Grossberg, S. E. (1984) Handb. Exp. Pharmakol. 71, 169-204. 4. Branca, A. A., Faltynek, C. R., D'Alessandro, S. B. & Baglioni, C. (1982) J. Biol. Chem. 257, 13291-132%. 5. Zoon, K. C., Arnheiter, H., Zur Nedden, D., Fitzgerald, D. J. & Willingham, M. C. (1983) Virology 130, 195-203. 6. Anderson, P., Yip, Y. K. & Vilcek, J. (1983) J. Biol. Chem. 258, 6497-6502. 7. Fuse, A. & Kuwata, T. (1978) J. Natl. Cancer Inst. 60, 1227-1232. 8. Panniers, L. R. V. & Clemens, M. J. (1981) J. Cell Sci. 48,

259-279. 9. Tovey, M. G., Rochette-Egly, C. & Costagna, M. (1979) Proc. Natl. Acad. Sci. USA 76, 3890-3893. 10. Schneck, J., Rager-Zisman, B., Rosen, 0. M. & Bloom, B. B. (1982) Proc. Natl. Acad. Sci. USA 79, 1879-1883. 11. Banejee, D. K., Baksi, K. & Gottesman, M. M. (1983) Virology 129, 230-238. 12. Ebsworth, N. M., Taylor-Papadimitriou, J. & Rosengurt, E. (1984) J. Cell. Physiol. 120, 146-150. 13. Fukumoto, Y., Kawahara, Y., Kariya, K., Araki, S., Fukuzaki, H. & Takai, Y. (1988) Biochem. Biophys. Res. Commun. 157, 337-345. 14. Marcus, P. I. & Sekellick, M. J. (1977) Nature (London) 266, 378-393. 15. Hubbell, H. R., Liu, R.-S. & Maxwell, B. L. (1984) Cancer Res. 44, 3252-3257. 16. Dick, R. S. & Hubbell, H. R. (1987) J. Neuro-Oncol. 5, 331-338. 17. Hubbell, H. R. (1986) Int. J. Cancer 37, 359-365. 18. Chapekar, M. S., Knode, M. C. & Glazer, R. I. (1988) Mol. Pharmacol. 34, 461-466. 19. Strayer, D. R., Weisband, J., Carter, W. A., Black, P., Nidzgorski, F. & Cook, A. W. (1987) Am. J. Clin. Oncol. 10, 281-284. 20. Hidaka, H., Inagaki, M., Kawamoto, S. & Sasaki, Y. (1984) Biochemistry 23, 5036-5041. 21. Zar, J. H. (1974) Biostatistical Analysis (Prentice-Hall, Englewood Cliffs, NJ), pp. 105-107. 22. Finter, N. B. (1969) J. Gen. Virol. 5, 419-427.

Proc. Natl. Acad. Sci. USA 88 (1991) 23. Burch, R. M., White, M. F. & Connor, J. R. (1989) J. Cell. Physiol. 139, 29-33. 24. Young, W. S., III, Shepard, E. A. & Burch, R. M. (1987) Mol. Endocrinol. 1, 884-888. 25. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J.

(1951) J. Biol. Chem. 193, 265-275. 26. Beavo, J. A., Rogers, N. L., Crofford, 0. 1)., Hardman, J. G., Sutherland, E. W. & Newman, E. V. (1970) Mol. Pharmacol. 6, 597-603. 27. Roesler, W. J., Vandenbark, G. R. & Hanson, R. W. (1988) J. Biol. Chem. 263, 9063-9066. 28. Silver, B. J., Bokar, J. A., Virgin, J. B., Vallen, E. A., Milstead, A. & Nilson, J. H. (1987) Proc. Natl. Acad. Sci. USA 84, 2198-2202. 29. Tsudaka, F., Fink, J. S., Mandel, G. & Goodman, R. H. (1987) J. Biol. Chem. 262, 8743-8747. 30. Sassone-Corsi, P., Visvader, J., Ferland, L., Mellon, P. L. & Verma, I. M. (1988) Genes Dev. 2, 1529-1538. 31. Comb, M., Birnberg, N. C., Seasholtz, A., Herbert, E. & Goodman, H. M. (1986) Nature (London) 323, 353-356. 32. Mitchell, P. J., Wang, C. & Tjian, R. (1987) Cell 50, 847-861. 33. Milstead, A., Cox, R. P. & Nilson, J. H. (1987) DNA 6, 213-219. 34. Jungmann, R. A., Kelley, D. C., Miles, M. F. & Milkowski, D. M. (1983) J. Biol. Chem. 258, 5312-5318. 35. Andersen, B., Milstead, A., Kennedy, G. & Nilson, J. H. (1988) J. Biol. Chem. 263, 15578-15583. 36. Itkes, A. V., Turpaev, K. T., Kartesheva, 0. N., Kafiani, C. A. & Severin, E. S. (1984) Mol. Cell. Biochem. 58,165-171. 37. Milhaud, P. G., Silhol, M., Salehzada, T. & LeBleu, B. (1987) J. Gen. Virol. 68, 1125-1134. 38. Ross, E. M. & Gilman, A. G. (1980) Annu. Rev. Biochem. 495, 533-564. 39. Lefkowitz, R. J., Stadel, J. M. & Caron, M. G. (1983) Annu. Rev. Biochem. 52, 159-186. 40. Smigel, M. D., Fergusin, J. M. & Gilman, A. G. (1985) Adv. Cyclic Nucleotide Protein Phosphorylation Res. 19, 103-111. 41. Yoshida, I. & Azuma, M. (1985) J. Interferon Res. 5, 1-10. 42. Azuma, M., Yoshida, I. & Kawai, H. (1988) J. Interferon Res. 8, S90. 43. Diamantstein, T. & Blitstein-Willinger, E. (1978) Eur. J. Immunol. 8, 896-899. 44. Biddle, W., Leong, S. S., Horoszewicz, J. & Ambrus, J. L. (1984) Proc. Soc. Exp. Biol. Med. 177, 487-490. 45. Yap, W. H., Teo, T. S., McCoy, E. & Tan, Y. H. (1986) Proc. Natl. Acad. Sci. USA 83, 7765-7769. 46. Kimchi, A., Shure, H. & Revel, M. (1981) Eur. J. Biochem. 114, 5-10. 47. Jacobsen, H., Krause, D., Friedman, R. M. & Silverman, R. H. (1983) Proc. Natl. Acad. Sci. USA 80, 4954-4958. 48. Wells, M. & Mallucci, L. (1985) Exp. Cell Res. 159, 27-36. 49. Petryshyn, R., Chen, J.-J. & London, I. M. (1984) J. Biol. Chem. 259, 14736-14742.